文本方向分类优化&更新版面分析模型

.env

@@ -3,5 +3,4 @@ POSTGRESQL_PORT=
 POSTGRESQL_USERNAME=
 POSTGRESQL_PASSWORD=
 POSTGRESQL_DATABASE=
-
-VISUAL=0
+POSTGRESQL_SCHEMA=

.env.dev

@@ -3,5 +3,4 @@ POSTGRESQL_PORT=54321
 POSTGRESQL_USERNAME=postgres
 POSTGRESQL_PASSWORD=123456
 POSTGRESQL_DATABASE=pdf-qa
-
-VISUAL=1
+POSTGRESQL_SCHEMA=pre_train

helper/constants.py

@@ -0,0 +1,25 @@
+import enum
+
+
+class PageDetectionEnum(enum.Enum):
+    # 注意：使用训练后的权重时，id需要+1，TEXT从1开始，使用预训练权重时，TEXT从0开始
+    TEXT = 1
+    TITLE = 2
+    FIGURE = 3
+    FIGURE_CAPTION = 4
+    TABLE = 5
+    TABLE_CAPTION = 6
+    HEADER = 7
+    FOOTER = 8
+    REFERENCE = 9
+    EQUATION = 10
+    SCANNED_DOCUMENT = 11
+
+    @property
+    def label(self):
+        return self.name.lower()
+
+
+class PDFAnalysisStatus(enum.Enum):
+    SUCCESS = 2
+    FAIL = 3

helper/content_recognition/main.py

@@ -1,8 +1,9 @@
 from typing import List
-import cv2
-from .utils import scanning_document_classify, table_rec, scanning_document_rec, markdown_rec, assign_tables_to_titles, remove_watermark
+from .utils import table_rec, scanning_document_rec, markdown_rec, assign_tables_to_titles
 from tqdm import tqdm
 from ..image_helper import text_rec
+from ..page_detection.utils import PageDetectionResult
+from ..constants import PageDetectionEnum as E
 
 
 class LayoutRecognitionResult(object):
@@ -17,10 +18,7 @@ class LayoutRecognitionResult(object):
         return f"[{self.clsid}] {self.content}"
 
 
-expand_pixel = 10
-
-
-def rec(page_detection_results, tmp_dir) -> List[List[LayoutRecognitionResult]]:
+def rec(page_detection_results: List[PageDetectionResult]) -> List[List[LayoutRecognitionResult]]:
     page_recognition_results = []
 
     for page_idx in tqdm(range(len(page_detection_results)), '文本识别'):
@@ -29,48 +27,25 @@ def rec(page_detection_results, tmp_dir) -> List[List[LayoutRecognitionResult]]:
             page_recognition_results.append([])
             continue
 
-        img = cv2.imread(results.image_path)
-        h, w = img.shape[:2]
-
-        for layout in results.boxes:
-            # box往外扩一点便于ocr
-            layout.pos[0] -= expand_pixel
-            layout.pos[1] -= expand_pixel
-            layout.pos[2] += expand_pixel
-            layout.pos[3] += expand_pixel
-
-            layout.pos[0] = max(0, layout.pos[0])
-            layout.pos[1] = max(0, layout.pos[1])
-            layout.pos[2] = min(w, layout.pos[2])
-            layout.pos[3] = min(h, layout.pos[3])
+        img = results.image
 
         outputs = []
 
-        is_scanning_document = False
         for layout in results.boxes:
             x1, y1, x2, y2 = layout.pos
             x1, y1, x2, y2 = int(x1), int(y1), int(x2), int(y2)
             layout_img = img[y1: y2, x1: x2]
             content = None
-            if layout.clsid == 0:
+            if layout.clsid == E.TEXT.value:
                 # text
                 content = markdown_rec(layout_img)
-            elif layout.clsid == 2:
-                # figure
-                if scanning_document_classify(layout_img):
-                    # 扫描件
-                    is_scanning_document = True
-                    content, layout_img = scanning_document_rec(layout_img)
-                    source_page_unwatermarked_img = remove_watermark(cv2.imread(f'{tmp_dir}/{page_idx + 1}.jpg'))
-            elif layout.clsid == 4:
+            elif layout.clsid == E.TABLE.value:
                 # table
-                if scanning_document_classify(layout_img):
-                    is_scanning_document = True
-                    content, layout_img = scanning_document_rec(layout_img)
-                    source_page_unwatermarked_img = remove_watermark(cv2.imread(f'{tmp_dir}/{page_idx + 1}.jpg'))
-                else:
-                    content = table_rec(layout_img)
-            elif layout.clsid == 5:
+                content = table_rec(layout_img)
+            elif layout.clsid == E.SCANNED_DOCUMENT.value:
+                # scanned document
+                content = scanning_document_rec(layout_img)
+            elif layout.clsid == E.TABLE_CAPTION.value:
                 # table caption
                 _, ocr_results, _ = text_rec(layout_img)
                 content = ''
@@ -86,32 +61,33 @@ def rec(page_detection_results, tmp_dir) -> List[List[LayoutRecognitionResult]]:
             result = LayoutRecognitionResult(layout.clsid, content, layout.pos)
             outputs.append(result)
 
-        if is_scanning_document and len(outputs) == 1:
-            # 扫描件额外提取标题
-            h, w = source_page_unwatermarked_img.shape[:2]
-            if h > w:
-                title_img = source_page_unwatermarked_img[:360, :w, ...]
-
-                # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}.jpg', title_img)
-                # vis = cv2.rectangle(source_page_unwatermarked_img.copy(), (0, 0), (w, 360), (255, 255, 0), 3)
-                # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}-vis.jpg', vis)
-            else:
-                title_img = source_page_unwatermarked_img[:410, :w, ...]
-
-                # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}.jpg', title_img)
-                # vis = cv2.rectangle(source_page_unwatermarked_img.copy(), (0, 310), (w, 410), (255, 255, 0), 3)
-                # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}-vis.jpg', vis)
-            _, title, _ = text_rec(title_img)
-            outputs[0].table_title = '\n'.join(title)
-        else:
-            # 自动给表格分配距离它最近的标题
-            assign_tables_to_titles(outputs)
+        # if is_scanning_document and len(outputs) == 1:
+        #     # 扫描件额外提取标题
+        #     h, w = layout_img.shape[:2]
+        #     if h > w:
+        #         title_img = layout_img[:360, :w, ...]
+
+        #         # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}.jpg', title_img)
+        #         # vis = cv2.rectangle(layout_img.copy(), (0, 0), (w, 360), (255, 255, 0), 3)
+        #         # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}-vis.jpg', vis)
+        #     else:
+        #         title_img = layout_img[:410, :w, ...]
+
+        #         # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}.jpg', title_img)
+        #         # vis = cv2.rectangle(layout_img.copy(), (0, 310), (w, 410), (255, 255, 0), 3)
+        #         # cv2.imwrite(f'/mnt/pdf2markdown/temp/{page_idx + 1}-vis.jpg', vis)
+        #     _, title, _ = text_rec(title_img)
+        #     outputs[0].table_title = '\n'.join(title)
+        # else:
+            
+        # 自动给表格分配距离它最近的标题
+        assign_tables_to_titles(outputs)
 
         # 表格标题可以删掉了
-        outputs = [_ for _ in outputs if _.clsid != 5]
-        # 将2-图片 和 4-表格转为数据库中的枚举 1-表格
+        outputs = [_ for _ in outputs if _.clsid != E.TABLE_CAPTION.value]
+        # 将表格转为数据库中的枚举 1-表格
         for o in outputs:
-            if o.clsid == 2 or o.clsid == 4:
+            if o.clsid == E.TABLE.value:
                 o.clsid = 1
         page_recognition_results.append(outputs)
         

helper/content_recognition/utils.py

@@ -1,7 +1,6 @@
 import os
 import tempfile
 import cv2
-import numpy as np
 from marker.converters.table import TableConverter
 from marker.models import create_model_dict
 from marker.output import text_from_rendered
@@ -11,38 +10,7 @@ from magic_pdf.data.read_api import read_local_images
 from markdownify import markdownify as md
 import re
 from ..image_helper import text_rec
-
-
-def scanning_document_classify(image):
-    # 判断是否是扫描件
-
-    # 将图像从BGR颜色空间转换到HSV颜色空间
-    hsv = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)
-
-    # 定义红色的HSV范围
-    lower_red1 = np.array([0, 70, 50])
-    upper_red1 = np.array([10, 255, 255])
-    lower_red2 = np.array([170, 70, 50])
-    upper_red2 = np.array([180, 255, 255])
-
-    # 创建两个掩码，一个用于低色调的红色，一个用于高色调的红色
-    mask1 = cv2.inRange(hsv, lower_red1, upper_red1)
-    mask2 = cv2.inRange(hsv, lower_red2, upper_red2)
-
-    # 将两个掩码合并
-    mask = cv2.bitwise_or(mask1, mask2)
-
-    # 计算红色区域的非零像素数量
-    non_zero_pixels = cv2.countNonZero(mask)
-    return 1 < non_zero_pixels < 1000
-
-
-def remove_watermark(image):
-    # 去除红色印章
-    _, _, r_channel = cv2.split(image)
-    r_channel[r_channel > 210] = 255
-    r_channel = cv2.cvtColor(r_channel, cv2.COLOR_GRAY2BGR)
-    return r_channel
+from ..constants import PageDetectionEnum as E
 
 
 def html2md(html_content):
@@ -68,6 +36,8 @@ def markdown_rec(image):
 
 def table_rec(image):
     boxes, texts, conficences = text_rec(image)
+    if not texts:
+        return None
     ocr_result = list(zip(boxes, texts, conficences))
     return table2md_pipeline(image, ocr_result)
 
@@ -75,18 +45,15 @@ def table_rec(image):
 table_converter = TableConverter(artifact_dict=create_model_dict())
 
 
-def scanning_document_rec(image):
+def scanning_document_rec(scanning_document_image):
     tmp_image_path = f'{tempfile.mktemp()}.jpg'
-
     try:
-        unwatermarked_image = remove_watermark(image)
-        cv2.imwrite(tmp_image_path, unwatermarked_image)
-
+        cv2.imwrite(tmp_image_path, scanning_document_image)
         rendered = table_converter(tmp_image_path)
         text, _, _ = text_from_rendered(rendered)
     finally:
         os.remove(tmp_image_path)
-    return text, unwatermarked_image
+    return text
 
 
 def compute_box_distance(box1, box2):
@@ -126,8 +93,8 @@ def compute_box_distance(box1, box2):
 
 
 def assign_tables_to_titles(layout_results, max_distance=200):
-    tables = [_ for _ in layout_results if _.clsid == 4]
-    titles = [_ for _ in layout_results if _.clsid == 5]
+    tables = [_ for _ in layout_results if _.clsid in (E.TABLE.value, E.SCANNED_DOCUMENT.value)]
+    titles = [_ for _ in layout_results if _.clsid == E.TABLE_CAPTION.value]
 
     table_to_title = {}
     title_to_table = {}
@@ -184,11 +151,3 @@ def assign_tables_to_titles(layout_results, max_distance=200):
             table.table_title = title.content
         else:
             table.table_title = None
-
-
-if __name__ == '__main__':
-    # content = text_rec('/mnt/research/PaddleOCR/pdf2md_pipeline/s4_content_recognition/all_layouts/5.jpg')
-    # content = markdown_rec('/mnt/research/PaddleOCR/pdf2md_pipeline/s4_content_recognition/all_layouts/3.jpg')
-    # content = table_rec('/mnt/research/PaddleOCR/pdf2md_pipeline/s4_content_recognition/all_layouts/6.jpg')
-    content = scanning_document_rec('/mnt/research/PaddleOCR/pdf2md_pipeline/s4_content_recognition/all_layouts/103.jpg')
-    print(content)

helper/db_helper.py

@@ -13,7 +13,8 @@ def create_connection():
             user=os.environ['POSTGRESQL_USERNAME'],
             password=os.environ['POSTGRESQL_PASSWORD'],
             host=os.environ['POSTGRESQL_HOST'],
-            port=os.environ['POSTGRESQL_PORT']
+            port=os.environ['POSTGRESQL_PORT'],
+            options=f'-c search_path={os.environ["POSTGRESQL_SCHEMA"]}'
         )
         conn.autocommit = False
         with conn.cursor() as cur:

helper/image_helper.py

@@ -1,54 +1,236 @@
 from typing import List
+from typing_extensions import deprecated
+import numpy as np
 from pdf2image import convert_from_path
 import os
-import paddleclas
 import cv2
 from .page_detection.utils import PageDetectionResult
 from paddleocr import PaddleOCR
+from .constants import PageDetectionEnum as E
+from paddlex import create_model
 
 
-paddle_clas_model = paddleclas.PaddleClas(model_name="text_image_orientation")
+ocr = PaddleOCR(use_angle_cls=False, lang='ch', use_gpu=True, show_log=False)
+text_orient_model = create_model(model_name="PP-LCNet_x1_0_doc_ori")
+
 
-def pdf2image(pdf_path, output_dir):
-    if not os.path.isdir(output_dir):
-        os.makedirs(output_dir)
+def pdf2image(pdf_path: str) -> List[np.ndarray]:
     images = convert_from_path(pdf_path)
-    for i, image in enumerate(images):
-        image.save(f'{output_dir}/{i + 1}.jpg')
+    images = [cv2.cvtColor(np.array(_.convert('RGB')), cv2.COLOR_RGB2BGR) for _ in images]
+    return images
+
+
+def boxes_to_heatmap(boxes, image_shape):
+    """生成热力图：每个像素表示被多少个box覆盖"""
+    heatmap = np.zeros(image_shape, dtype=np.float32)
+    for xmin, ymin, xmax, ymax in boxes:
+        xmin, ymin, xmax, ymax = map(int, (xmin, ymin, xmax, ymax))
+        heatmap[ymin:ymax, xmin:xmax] += 1
+    return heatmap
+
+
+def compute_integral_image(heatmap):
+    """计算积分图"""
+    return heatmap.cumsum(axis=0).cumsum(axis=1)
+
+
+def get_sum_in_rect(integral, x1, y1, x2, y2):
+    """使用积分图获取矩形区域的和"""
+    total = integral[y2, x2]
+    if x1 > 0:
+        total -= integral[y2, x1 - 1]
+    if y1 > 0:
+        total -= integral[y1 - 1, x2]
+    if x1 > 0 and y1 > 0:
+        total += integral[y1 - 1, x1 - 1]
+    return total
+
+
+def find_max_area_square(boxes, image_width, image_height, window_size=224):
+    heatmap = boxes_to_heatmap(boxes, (image_height, image_width))
+    integral = compute_integral_image(heatmap)
+
+    h, w = heatmap.shape
+    H = h - window_size + 1
+    W = w - window_size + 1
+
+    # 创建矩阵索引
+    ys, xs = np.meshgrid(np.arange(H), np.arange(W), indexing='ij')
+    x1 = xs
+    y1 = ys
+    x2 = x1 + window_size - 1
+    y2 = y1 + window_size - 1
+
+    # 用矢量化方式计算所有区域的总和
+    total = (
+        integral[y2, x2]
+        - np.where(x1 > 0, integral[y2, x1 - 1], 0)
+        - np.where(y1 > 0, integral[y1 - 1, x2], 0)
+        + np.where((x1 > 0) & (y1 > 0), integral[y1 - 1, x1 - 1], 0)
+    )
+
+    max_idx = np.unravel_index(np.argmax(total), total.shape)
+    top_left_y, top_left_x = max_idx
+    best_window = [int(top_left_x), int(top_left_y),
+                   int(top_left_x + window_size), int(top_left_y + window_size)]
+    best_score = total[max_idx]
+    return best_window, best_score
+
+
+def image_orient_cls(images):
+    if isinstance(images, np.ndarray):
+        images = [images]
+
+    angles = []
+    for img in images:
+        h, w = img.shape[:2]
+        det_res = ocr.ocr(img, det=True, rec=False, cls=False)[0]
+        boxes = []
+        for r in det_res:
+            x1, y1, x2, y2 = int(r[0][0]), int(r[0][1]), int(r[2][0]), int(r[2][1])
+            boxes.append((x1, y1, x2, y2))
+        square_box, _ = find_max_area_square(boxes, w, h, 224)
+        x1, y1, x2, y2 = square_box
+        x1 = max(0, x1 - 16)
+        y1 = max(0, y1 - 16)
+        x2 = min(w, x2 + 16)
+        y2 = min(h, y2 + 16)
+        _img = img[y1:y2, x1:x2]
+
+        output = text_orient_model.predict(_img,  batch_size=1)
+        angle = int(next(output)['label_names'][0])
+        angles.append(angle)
+
+    return angles
+
+
+@deprecated('Low precision classification of scanned documents')
+def scanning_document_classify(image):
+    # 判断是否是扫描件
+
+    # 将图像从BGR颜色空间转换到HSV颜色空间
+    hsv = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)
+
+    # 定义红色的HSV范围
+    lower_red1 = np.array([0, 70, 50])
+    upper_red1 = np.array([10, 255, 255])
+    lower_red2 = np.array([170, 70, 50])
+    upper_red2 = np.array([180, 255, 255])
+
+    # 创建两个掩码，一个用于低色调的红色，一个用于高色调的红色
+    mask1 = cv2.inRange(hsv, lower_red1, upper_red1)
+    mask2 = cv2.inRange(hsv, lower_red2, upper_red2)
 
+    # 将两个掩码合并
+    mask = cv2.bitwise_or(mask1, mask2)
 
-def image_orient_cls(input_data):
-    return paddle_clas_model.predict(input_data)
+    # 计算红色区域的非零像素数量
+    non_zero_pixels = cv2.countNonZero(mask)
+    return 1 < non_zero_pixels < 1000
+
+
+def remove_watermark(image):
+    # 去除红色印章
+    _, _, r_channel = cv2.split(image)
+    r_channel[r_channel > 210] = 255
+    r_channel = cv2.cvtColor(r_channel, cv2.COLOR_GRAY2BGR)
+    return r_channel
+
+
+def overlay_rectangle(image, top_left, bottom_right, color, alpha=0.5):
+    """
+    在图像的矩形区域内覆盖颜色并设置透明度。
+
+    参数：
+    - image: 输入图像（numpy数组，BGR格式）
+    - top_left: 左上角坐标 (x, y)
+    - bottom_right: 右下角坐标 (x, y)
+    - color: 填充颜色，BGR 格式，默认绿色
+    - alpha: 透明度，0.0（完全透明）到 1.0（完全不透明）
+
+    返回：
+    - 带有矩形覆盖效果的图像副本
+    """
+    # 确保 alpha 合法
+    alpha = np.clip(alpha, 0, 1)
+
+    # 复制图像用于输出
+    output = image.copy()
+
+    # 创建一个矩形区域的遮罩图层
+    overlay = image.copy()
+    cv2.rectangle(overlay, top_left, bottom_right, color, thickness=-1)
+
+    # 将 overlay 融合到原图上
+    cv2.addWeighted(overlay, alpha, output, 1 - alpha, 0, output)
+
+    return output
+
+
+def overlay_rectangle(image, top_left, bottom_right, text, color, alpha=0.5, font_scale=1.0, thickness=1):
+    """
+    在图像上绘制带透明填充的矩形框和位于框外部的带背景标签。
+
+    返回：
+    - 带可视化效果的图像副本
+    """
+    alpha = np.clip(alpha, 0, 1)
+    output = image.copy()
+    overlay = image.copy()
+
+    # 1. 绘制目标区域透明填充
+    cv2.rectangle(overlay, top_left, bottom_right, color, thickness=-1)
+
+    x, y = top_left
+    # 获取文字尺寸
+    (text_width, text_height), baseline = cv2.getTextSize(text, cv2.FONT_HERSHEY_TRIPLEX, font_scale, thickness)
+    text_origin = (x, y)  # 文字左下角坐标
+    text_topleft = (x, y - text_height - baseline)
+    text_bottomright = (x + text_width, y)
+
+    # 防止文字背景区域越界
+    text_topleft = (max(0, text_topleft[0]), max(0, text_topleft[1]))
+
+    # 绘制文字背景到 overlay（同样使用透明度）
+    cv2.rectangle(overlay, text_topleft, text_bottomright, color, thickness=-1)
+
+    # 将文字写到 output 图像上
+    cv2.putText(output, text, text_origin, cv2.FONT_HERSHEY_TRIPLEX, font_scale, (0, 0, 0), thickness)
+
+    # 3. 合成透明效果
+    cv2.addWeighted(overlay, alpha, output, 1 - alpha, 0, output)
+
+    return output
 
 
 def page_detection_visual(page_detection_result: PageDetectionResult):
-    img = cv2.imread(page_detection_result.image_path)
+    img = page_detection_result.image
     for box in page_detection_result.boxes:
         pos = box.pos
         clsid = box.clsid
         confidence = box.confidence
-        if clsid == 0:
-            color = (0, 0, 0)
-            text = 'text'
-        elif clsid == 1:
-            color = (255, 0, 0)
-            text = 'title'
-        elif clsid == 2:
-            color = (0, 255, 0)
-            text = 'figure'
-        elif clsid == 4:
-            color = (0, 0, 255)
-            text = 'table'
-        if clsid == 5:
-            color = (255, 0, 255)
-            text = 'table caption'
-        text = f'{text} {confidence}'
-        img = cv2.rectangle(img, (int(pos[0]), int(pos[1])), (int(pos[2]), int(pos[3])), color, 2)
-        cv2.putText(img, text, (int(pos[0]), int(pos[1])), cv2.FONT_HERSHEY_TRIPLEX, 1, color, 2)
-    return img
+        if clsid == E.TEXT.value:
+            text = E.TEXT.label
+            color = (177, 216, 178)
+        elif clsid == E.TITLE.value:
+            text = E.TITLE.label
+            color = (181, 136, 145)
+        elif clsid == E.TABLE.value:
+            text = E.TABLE.label
+            color = (255, 97, 176)
+        elif clsid == E.TABLE_CAPTION.value:
+            text = E.TABLE_CAPTION.label
+            color = (66, 73, 255)
+        elif clsid == E.SCANNED_DOCUMENT.value:
+            text = E.SCANNED_DOCUMENT.label
+            color = (255, 239, 145)
+        else:
+            continue
+        text = f'{text} {confidence:.2f}'
+        img = overlay_rectangle(img, (int(pos[0]), int(pos[1])), (int(pos[2]), int(pos[3])), text, color, font_scale=0.6)
 
+    return img
 
-ocr = PaddleOCR(use_angle_cls=False, lang='ch', use_gpu=True, show_log=False)
 
 def text_rec(image):
     result = ocr.ocr(image, cls=False)

helper/page_detection/main.py

@@ -1,7 +1,10 @@
 from typing import List
+
 from .pdf_detection import Pipeline
 from utils import non_max_suppression, merge_text_and_title_boxes, LayoutBox, PageDetectionResult
 from tqdm import tqdm
+from ..constants import PageDetectionEnum as E
+from ..image_helper import remove_watermark
 
 
 """
@@ -15,25 +18,40 @@ from tqdm import tqdm
     7 - Footer
     8 - Reference
     9 - Equation
+    使用训练后的权重时，id需要+1，即TEXT从1开始
 """
-pipeline = Pipeline('./models/PaddleDetection/inference_model/picodet_lcnet_x1_0_fgd_layout_cdla_infer')
+pipeline = Pipeline('./models/PaddleDetection/inference_model/picodet_lcnet_x1_0_fgd_layout_cdla_infer_v2')
 
-effective_labels = [0, 1, 2, 4, 5]
-# nms优先级，索引越低优先级越低，box重叠时优先保留表格
-label_scores = [1, 5, 0, 2, 4]
+effective_labels = [E.TEXT.value, E.TITLE.value, E.TABLE.value, E.TABLE_CAPTION.value, E.SCANNED_DOCUMENT.value]
+# nms优先级，索引越低优先级越低
+label_scores = [E.TITLE.value, E.TABLE_CAPTION.value, E.TEXT.value, E.TABLE.value, E.SCANNED_DOCUMENT.value]
 expand_pixel = 10
 
 
-def layout_analysis(image_paths) -> List[PageDetectionResult]:
+def layout_analysis(images) -> List[PageDetectionResult]:
     layout_analysis_results = []
-    for image_path in tqdm(image_paths, '版面分析'):
-        page_detecion_outputs = pipeline(image_path)
+    for image in tqdm(images, '版面分析'):
+        page_detecion_outputs = pipeline(image)
+
         layout_boxes = []
-        for i in range(len(page_detecion_outputs)):
-            clsid, box, confidence = page_detecion_outputs[i]
+        is_scanned_document = False
+        for o in page_detecion_outputs:
+            clsid, box, confidence = o
             if clsid in effective_labels:
                 layout_boxes.append(LayoutBox(clsid, box, confidence))
-        page_detecion_outputs = PageDetectionResult(layout_boxes, image_path)
+            if clsid == E.SCANNED_DOCUMENT.value:
+                is_scanned_document = True
+                image = remove_watermark(image)
+
+        if is_scanned_document:
+            # 扫描件需要去水印后重新进行版面分析来识别出标题，因为训练的图片是去水印之后的
+            _page_detecion_outputs = pipeline(image)
+            for o in _page_detecion_outputs:
+                clsid, box, confidence = o
+                if clsid == E.TABLE_CAPTION.value:
+                    layout_boxes.append(LayoutBox(clsid, box, confidence))
+
+        page_detecion_outputs = PageDetectionResult(layout_boxes, image)
 
         scores = []
         poses = []
@@ -49,20 +67,38 @@ def layout_analysis(image_paths) -> List[PageDetectionResult]:
             _boxes.append(page_detecion_outputs.boxes[i])
         page_detecion_outputs.boxes = _boxes
 
-        for i in range(len(page_detecion_outputs.boxes) - 1, -1, -1):
-            box = page_detecion_outputs.boxes[i]
-            if box.clsid in (0, 5):
+        if not is_scanned_document:
+            for i in range(len(page_detecion_outputs.boxes) - 1, -1, -1):
                 # 移除Table box和Figure box中的Table caption box和Text box (有些扫描件会被识别为Figure)
-                for _box in page_detecion_outputs.boxes:
-                    if _box.clsid != 2 and _box.clsid != 4:
-                        continue
-                    if box.pos[0] > _box.pos[0] and box.pos[1] > _box.pos[1] and box.pos[2] < _box.pos[2] and box.pos[3] < _box.pos[3]:
-                        page_detecion_outputs.boxes.remove(box)
+                box = page_detecion_outputs.boxes[i]
+                if box.clsid in (E.TEXT.value, E.TABLE_CAPTION.value):
+                    for _box in page_detecion_outputs.boxes:
+                        if _box.clsid != E.FIGURE.value and _box.clsid != E.TABLE.value:
+                            continue
+                        if box.pos[0] > _box.pos[0] and box.pos[1] > _box.pos[1] and box.pos[2] < _box.pos[2] and box.pos[3] < _box.pos[3]:
+                            page_detecion_outputs.boxes.remove(box)
 
         # 将text和title合并起来，便于转成markdown格式
-        page_detecion_outputs.boxes = merge_text_and_title_boxes(page_detecion_outputs.boxes, (0, 1))
+        merged_labels = [E.TEXT.value, E.TITLE.value]
+        other_labels = list(set(effective_labels) - set(merged_labels))
+        page_detecion_outputs.boxes = merge_text_and_title_boxes(page_detecion_outputs.boxes, merged_labels, other_labels, E.TEXT.value)
+        
         # 对box进行排序
         page_detecion_outputs.boxes.sort(key=lambda x: (x.pos[1], x.pos[0]))
+
+        # box外扩，便于后续的ocr
+        h, w = image.shape[:2]
+        for layout in page_detecion_outputs.boxes:
+            layout.pos[0] -= expand_pixel
+            layout.pos[1] -= expand_pixel
+            layout.pos[2] += expand_pixel
+            layout.pos[3] += expand_pixel
+
+            layout.pos[0] = max(0, layout.pos[0])
+            layout.pos[1] = max(0, layout.pos[1])
+            layout.pos[2] = min(w, layout.pos[2])
+            layout.pos[3] = min(h, layout.pos[3])
+
         layout_analysis_results.append(page_detecion_outputs)
 
     return layout_analysis_results

helper/page_detection/pdf_detection.py

@@ -650,7 +650,7 @@ class PredictConfig():
             print(
                 'The RCNN export model is used for ONNX and it only supports batch_size = 1'
             )
-        self.print_config()
+        # self.print_config()
 
     def check_model(self, yml_conf):
         """
@@ -890,13 +890,11 @@ class Pipeline(object):
             output_dir=FLAGS.output_dir,
             use_fd_format=FLAGS.use_fd_format)
 
-    def __call__(self, image_path):
-        if FLAGS.image_dir is None and image_path is not None:
-            assert FLAGS.batch_size == 1, "batch_size should be 1, when image_file is not None"
-        if isinstance(image_path, str):
-            image_path = [image_path]
+    def __call__(self, image):
+        if isinstance(image, np.ndarray):
+            image = [image]
         results = self.detector.predict_image(
-            image_path,
+            image,
             visual=FLAGS.save_images)
         return results
             
@@ -904,7 +902,7 @@ class Pipeline(object):
 paddle.enable_static()
 parser = argsparser()
 FLAGS = parser.parse_args()
-print_arguments(FLAGS)
+# print_arguments(FLAGS)
 FLAGS.device = 'GPU'
 FLAGS.save_images = False
 FLAGS.device = FLAGS.device.upper()

helper/page_detection/test.py

@@ -0,0 +1,57 @@
+from typing import List
+import cv2
+from pdf_detection import Pipeline
+import pickle
+
+
+class LayoutBox(object):
+    def __init__(self, clsid: int, pos: List[float], confidence: float):
+        self.clsid = clsid
+        self.pos = pos
+        self.confidence = confidence
+
+
+class PageDetectionResult(object):
+    def __init__(self, boxes: List[LayoutBox], image_path: str):
+        self.boxes = boxes
+        self.image_path = image_path
+
+pipeline = Pipeline('/mnt/pdf2markdown/models/PaddleDetection/inference_model/picodet_lcnet_x1_0_fgd_layout_cdla_infer')
+
+
+def page_detection_visual(page_detection_result: PageDetectionResult):
+    img = cv2.imread(page_detection_result.image_path)
+    for box in page_detection_result.boxes:
+        pos = box.pos
+        clsid = box.clsid
+        confidence = box.confidence
+        if clsid == 0:
+            color = (0, 0, 0)
+            text = 'text'
+        elif clsid == 1:
+            color = (255, 0, 0)
+            text = 'title'
+        elif clsid == 2:
+            color = (0, 255, 0)
+            text = 'figure'
+        elif clsid == 4:
+            color = (0, 0, 255)
+            text = 'table'
+        if clsid == 5:
+            color = (255, 0, 255)
+            text = 'table caption'
+        text = f'{text} {confidence}'
+        img = cv2.rectangle(img, (int(pos[0]), int(pos[1])), (int(pos[2]), int(pos[3])), color, 2)
+        cv2.putText(img, text, (int(pos[0]), int(pos[1])), cv2.FONT_HERSHEY_TRIPLEX, 1, color, 2)
+    return img
+
+img_path = '/mnt/research/PaddleOCR/PaddleDetection/datasets/train_output/JPEGImages/0090.jpg'
+page_detecion_outputs = pipeline(img_path)
+boxes = []
+for output in page_detecion_outputs:
+    boxes.append(LayoutBox(output[0], output[1], output[2]))
+res = PageDetectionResult(boxes, img_path)
+with open('/mnt/pdf2markdown/a.pkl', 'wb') as f:
+    pickle.dump(res, f)
+# img = page_detection_visual(res)
+# cv2.imwrite('/mnt/pdf2markdown/0122.jpg', img)

helper/page_detection/utils.py

@@ -28,9 +28,9 @@ class LayoutBox(object):
 
 
 class PageDetectionResult(object):
-    def __init__(self, boxes: List[LayoutBox], image_path: str):
+    def __init__(self, boxes: List[LayoutBox], image: np.ndarray):
         self.boxes = boxes
-        self.image_path = image_path
+        self.image = image
 
 
 def argsparser():
@@ -590,9 +590,9 @@ def merge_boxes(boxes: List[List[float]]) -> List[float]:
     return [x1, y1, x2, y2]
 
 
-def merge_text_and_title_boxes(data: List[LayoutBox], merged_labels: Tuple[int]) -> List[LayoutBox]:
+def merge_text_and_title_boxes(data: List[LayoutBox], merged_labels: Tuple[int], other_labels: Tuple[int], merged_box_label: int) -> List[LayoutBox]:
     text_title_boxes = [(i, box) for i, box in enumerate(data) if box.clsid in merged_labels]
-    other_boxes = [box.pos for box in data if box.clsid in (2, 4, 5)]
+    other_boxes = [box.pos for box in data if box.clsid in other_labels]
 
     text_title_boxes.sort(key=lambda x: x[1].pos[1])  # sort by y1
 
@@ -618,7 +618,7 @@ def merge_text_and_title_boxes(data: List[LayoutBox], merged_labels: Tuple[int])
                 skip_indices.add(j)
                 j += 1
         if len(current_group) > 1:
-            merged_box = LayoutBox(0, merge_boxes(current_group), max(group_confidences))
+            merged_box = LayoutBox(merged_box_label, merge_boxes(current_group), max(group_confidences))
             merged.append(merged_box)
         else:
             idx = text_title_boxes[i][0]

models/PaddleDetection/inference_model/picodet_lcnet_x1_0_fgd_layout_cdla_infer_v2/infer_cfg.yml

@@ -2,7 +2,7 @@ mode: paddle
 draw_threshold: 0.5
 metric: COCO
 use_dynamic_shape: false
-arch: PicoDet
+arch: GFL
 min_subgraph_size: 3
 Preprocess:
 - interp: 2
@@ -25,11 +25,18 @@ Preprocess:
 - stride: 32
   type: PadStride
 label_list:
-- text
-- title
-- list
-- table
-- figure
+- _background_
+- Text
+- Title
+- Figure
+- Figure caption
+- Table
+- Table caption
+- Header
+- Footer
+- Reference
+- Equation
+- Scanning
 NMS:
   keep_top_k: 100
   name: MultiClassNMS

models/PaddleDetection/inference_model/picodet_lcnet_x1_0_fgd_layout_table_infer/infer_cfg.yml

@@ -1,39 +0,0 @@
-mode: paddle
-draw_threshold: 0.5
-metric: COCO
-use_dynamic_shape: false
-arch: PicoDet
-min_subgraph_size: 3
-Preprocess:
-- interp: 2
-  keep_ratio: false
-  target_size:
-  - 800
-  - 608
-  type: Resize
-- is_scale: true
-  mean:
-  - 0.485
-  - 0.456
-  - 0.406
-  std:
-  - 0.229
-  - 0.224
-  - 0.225
-  type: NormalizeImage
-- type: Permute
-- stride: 32
-  type: PadStride
-label_list:
-- table
-NMS:
-  keep_top_k: 100
-  name: MultiClassNMS
-  nms_threshold: 0.5
-  nms_top_k: 1000
-  score_threshold: 0.3
-fpn_stride:
-- 8
-- 16
-- 32
-- 64

models/PaddleDetection/inference_model/picodet_lcnet_x1_0_layout_infer/infer_cfg.yml

@@ -1,43 +0,0 @@
-mode: paddle
-draw_threshold: 0.5
-metric: COCO
-use_dynamic_shape: false
-arch: PicoDet
-min_subgraph_size: 3
-Preprocess:
-- interp: 2
-  keep_ratio: false
-  target_size:
-  - 800
-  - 608
-  type: Resize
-- is_scale: true
-  mean:
-  - 0.485
-  - 0.456
-  - 0.406
-  std:
-  - 0.229
-  - 0.224
-  - 0.225
-  type: NormalizeImage
-- type: Permute
-- stride: 32
-  type: PadStride
-label_list:
-- text
-- title
-- list
-- table
-- figure
-NMS:
-  keep_top_k: 100
-  name: MultiClassNMS
-  nms_threshold: 0.5
-  nms_top_k: 1000
-  score_threshold: 0.3
-fpn_stride:
-- 8
-- 16
-- 32
-- 64

pipeline.py

@@ -1,7 +1,9 @@
 from dotenv import load_dotenv
 import os
+from loguru import logger
 
 env = os.environ.get('env', 'dev')
+logger.info(f'Configure using this environment: {env}')
 load_dotenv(dotenv_path='.env.dev' if env == 'dev' else '.env', override=True)
 
 import time
@@ -11,84 +13,87 @@ from helper.image_helper import pdf2image, image_orient_cls, page_detection_visu
 from helper.page_detection.main import layout_analysis
 from helper.content_recognition.main import rec
 from helper.db_helper import insert_pdf2md_table
-import tempfile
-from loguru import logger
 import datetime
-import shutil
+from tqdm import tqdm
+from helper.constants import PDFAnalysisStatus
 
 
-def _pdf2markdown_pipeline(pdf_path, tmp_dir):
+def _pdf2markdown_pipeline(pdf_path, visual):
     start_time = time.time()
     # 1. pdf -> images
     t1 = time.time()
-    pdf2image(pdf_path, tmp_dir)
+    images = pdf2image(pdf_path)
     t2 = time.time()
 
     # 2. 图片方向分类
     t3 = time.time()
-    orient_cls_results = image_orient_cls(tmp_dir)
+    angles = image_orient_cls(images)
     t4 = time.time()
-    for r in orient_cls_results:
-        clsid = r[0]['class_ids'][0]
-        filename = r[0]['filename']
-        if clsid == 1 or clsid == 3:
-            img = cv2.imread(filename)
+    for i in range(len(angles)):
+        angle = angles[i]
+        img = images[i]
+        if angle == 0:
+            continue
+        if angle == 90:
+            img = cv2.rotate(img, cv2.ROTATE_90_COUNTERCLOCKWISE)
+        elif angle == 180:
+            img = cv2.rotate(img, cv2.ROTATE_180)
+        elif angle == 270:
             img = cv2.rotate(img, cv2.ROTATE_90_CLOCKWISE)
-            cv2.imwrite(filename, img)
-    
-    filepaths = os.listdir(tmp_dir)
-    filepaths.sort(key=lambda x: int(x.split('/')[-1].split('.')[0]))
-    filepaths = [f'{tmp_dir}/{_}' for _ in filepaths]
+        images[i] = img
 
-    # filepaths = filepaths[250:251]
+    # images = images[90: 123]
 
     # 3. 版面分析
     t5 = time.time()
-    layout_detection_results = layout_analysis(filepaths)
+    layout_detection_results = layout_analysis(images)
     t6 = time.time()
 
-    # 3.1 visual
-    if int(os.environ['VISUAL']):
+    # 3.1 版面分析可视化
+    if visual:
         visual_dir = './visual_images'
         for f in os.listdir(visual_dir):
             if f.endswith('.jpg'):
                 os.remove(f'{visual_dir}/{f}')
-        for i in range(len(layout_detection_results)):
+        for i in tqdm(range(len(layout_detection_results)), '版面分析可视化结果'):
             vis_img = page_detection_visual(layout_detection_results[i])
             cv2.imwrite(f'{visual_dir}/{i + 1}.jpg', vis_img)
 
+    # exit(0)
+    
     # 4. 内容识别
     t7 = time.time()
-    layout_recognition_results = rec(layout_detection_results, tmp_dir)
+    layout_recognition_results = rec(layout_detection_results)
     t8 = time.time()
 
     end_time = time.time()
-    logger.info(f'{pdf_path} analysis completed in {round(end_time - start_time, 3)} seconds, including {round(t2 - t1, 3)} for pdf to image, {round(t4 - t3, 3)} second for image orient classification, {round(t6 - t5, 3)} seconds for page detection, and {round(t8 - t7, 3)} seconds for layout recognition, page number: {len(filepaths)}')
+    logger.info(f'{pdf_path} analysis completed in {round(end_time - start_time, 3)} seconds, including {round(t2 - t1, 3)} for pdf to image, {round(t4 - t3, 3)} second for image orient classification, {round(t6 - t5, 3)} seconds for page detection, and {round(t8 - t7, 3)} seconds for layout recognition, page number: {len(images)}')
 
     return layout_recognition_results
 
 
-def pdf2markdown_pipeline(pdf_path: str):
+def pdf2markdown_pipeline(pdf_path: str, visual=False, insert_db=True):
     pdf_name = pdf_path.split('/')[-1]
     start_time = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')
     process_status = 0
-    tmp_dir = tempfile.mkdtemp()
     try:
-        results = _pdf2markdown_pipeline(pdf_path, tmp_dir)
+        results = _pdf2markdown_pipeline(pdf_path, visual)
     except Exception:
         logger.error(f'analysis pdf error! \n{traceback.format_exc()}')
-        process_status = 3
+        process_status = PDFAnalysisStatus.FAIL.value
         end_time = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')
-        insert_pdf2md_table(pdf_path, pdf_name, process_status, start_time, end_time, None)
+        if insert_db:
+            insert_pdf2md_table(pdf_path, pdf_name, process_status, start_time, end_time, None)
         pdf_id = None
     else:
-        process_status = 2
+        process_status = PDFAnalysisStatus.SUCCESS.value
         end_time = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')
-        pdf_id = insert_pdf2md_table(pdf_path, pdf_name, process_status, start_time, end_time, results)
-    finally:
-        shutil.rmtree(tmp_dir)
+        if insert_db:
+            pdf_id = insert_pdf2md_table(pdf_path, pdf_name, process_status, start_time, end_time, results)
     return process_status, pdf_id
 
 
 if __name__ == '__main__':
-    pdf2markdown_pipeline('/mnt/pdf2markdown/龙源电力：2023年年度审计报告.PDF')
+    pdf2markdown_pipeline('/mnt/pdf2markdown/龙源电力：2021年年度报告.PDF', visual=True, insert_db=True)
+    # pdf2markdown_pipeline('/mnt/pdf2markdown/龙源电力：2022年年度报告.PDF', visual=True, insert_db=True)
+    # pdf2markdown_pipeline('/mnt/pdf2markdown/龙源电力：2023年年度报告.PDF', visual=True, insert_db=True)

requirements.txt

@@ -16,6 +16,7 @@ cachetools==5.5.2
 certifi==2025.4.26
 cffi==1.17.1
 cfgv==3.4.0
+chardet==5.2.0
 charset-normalizer==3.4.1
 click==8.1.8
 cobble==0.1.4
@@ -33,6 +34,7 @@ distro==1.9.0
 doclayout_yolo==0.0.2b1
 easydict==1.13
 EbookLib==0.18
+einops==0.8.1
 et_xmlfile==2.0.0
 faiss-cpu==1.8.0.post1
 fast-langdetect==0.2.5
@@ -50,6 +52,7 @@ future==1.0.0
 gast==0.3.3
 google-auth==2.39.0
 google-genai==1.13.0
+GPUtil==1.4.0
 h11==0.16.0
 httpcore==1.0.9
 httpx==0.28.1
@@ -69,11 +72,11 @@ lazy_loader==0.4
 lmdb==1.6.2
 loguru==0.7.3
 lxml==5.4.0
-magic-pdf==1.3.10
+-e git+http://192.168.10.28:3000/Yaxin/pdf2markdown.git@f030719b330c56e9909196a8d4e00d3e9ec003dc#egg=magic_pdf&subdirectory=third_party/MinerU
 mammoth==1.9.0
 markdown2==2.5.3
 markdownify==0.13.1
--e marker
+-e git+http://192.168.10.28:3000/Yaxin/pdf2markdown.git@f030719b330c56e9909196a8d4e00d3e9ec003dc#egg=marker_pdf&subdirectory=third_party/marker
 MarkupSafe==3.0.2
 matplotlib==3.10.1
 modelscope==1.25.0
@@ -81,18 +84,29 @@ mpmath==1.3.0
 networkx==3.4.2
 nodeenv==1.9.1
 numpy==1.24.4
+nvidia-cublas-cu11==11.11.3.6
 nvidia-cublas-cu12==12.4.5.8
+nvidia-cuda-cupti-cu11==11.8.87
 nvidia-cuda-cupti-cu12==12.4.127
+nvidia-cuda-nvrtc-cu11==11.8.89
 nvidia-cuda-nvrtc-cu12==12.4.127
+nvidia-cuda-runtime-cu11==11.8.89
 nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cudnn-cu11==8.9.6.50
 nvidia-cudnn-cu12==9.1.0.70
+nvidia-cufft-cu11==10.9.0.58
 nvidia-cufft-cu12==11.2.1.3
+nvidia-curand-cu11==10.3.0.86
 nvidia-curand-cu12==10.3.5.147
+nvidia-cusolver-cu11==11.4.1.48
 nvidia-cusolver-cu12==11.6.1.9
+nvidia-cusparse-cu11==11.7.5.86
 nvidia-cusparse-cu12==12.3.1.170
 nvidia-cusparselt-cu12==0.6.2
+nvidia-nccl-cu11==2.19.3
 nvidia-nccl-cu12==2.21.5
 nvidia-nvjitlink-cu12==12.4.127
+nvidia-nvtx-cu11==11.8.86
 nvidia-nvtx-cu12==12.4.127
 omegaconf==2.3.0
 onnxruntime==1.21.1
@@ -105,10 +119,11 @@ opt-einsum==3.3.0
 packaging==25.0
 paddleclas==2.5.2
 paddleocr==2.10.0
-paddlepaddle-gpu==2.6.2
-pandas==2.2.3
+paddlepaddle-gpu==3.0.0
+paddlex==3.0.0rc1
+pandas==1.5.3
 pdf2image==1.17.0
-pdfminer.six==20250324
+pdfminer.six==20250506
 pdftext==0.6.2
 pillow==10.4.0
 platformdirs==4.3.7
@@ -144,6 +159,8 @@ regex==2024.11.6
 requests==2.32.3
 robust-downloader==0.0.2
 rsa==4.9.1
+ruamel.yaml==0.18.10
+ruamel.yaml.clib==0.2.12
 s3transfer==0.12.0
 safetensors==0.5.3
 scikit-image==0.25.2
@@ -156,7 +173,7 @@ six==1.17.0
 sniffio==1.3.1
 soupsieve==2.7
 stringzilla==3.12.5
-surya-ocr==0.13.1
+-e git+http://192.168.10.28:3000/Yaxin/pdf2markdown.git@f030719b330c56e9909196a8d4e00d3e9ec003dc#egg=surya_ocr&subdirectory=third_party/surya
 sympy==1.13.1
 termcolor==3.1.0
 thop==0.1.1.post2209072238

server.py

@@ -1,4 +1,4 @@
-from flask import Flask, request
+from flask import Flask, request, jsonify
 import requests
 from pipeline import pdf2markdown_pipeline
 
@@ -14,6 +14,7 @@ def pdf2markdown():
     for pdf_path in pdf_paths:
         process_status, pdf_id = pdf2markdown_pipeline(pdf_path)
         requests.post(callback_url, json={'pdfId': pdf_id, 'processStatus': process_status})
+    return jsonify({})
 
 
 if __name__ == '__main__':

third_party/MinerU/.gitattributes

@@ -0,0 +1,5 @@
+*.js linguist-vendored
+*.mjs linguist-vendored
+*.html linguist-documentation
+*.css linguist-vendored
+*.scss linguist-vendored

third_party/MinerU/.github/ISSUE_TEMPLATE/bug_report.yml

@@ -0,0 +1,135 @@
+name: 🐛 Bug Report
+description: Create a bug report for MinerU | MinerU 的 Bug 反馈
+labels: bug
+
+# We omit `title: "..."` so that the field defaults to blank. If we set it to
+# empty string, Github seems to reject this .yml file.
+
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thank you for submitting a MinerU 🐛 Bug Report! | 感谢您提交 MinerU 🐛 Bug 反馈！
+
+  - type: checkboxes
+    attributes:
+      label: 🔎 Search before asking | 提交之前请先搜索
+      description: >
+        Please search the MinerU [Readme](https://github.com/opendatalab/MinerU), [Issues](https://github.com/opendatalab/MinerU/issues) and [Discussions](https://github.com/opendatalab/MinerU/discussions) to see if a similar bug report already exists.
+      options:
+        - label: I have searched the MinerU [Readme](https://github.com/opendatalab/MinerU) and found no similar bug report.
+          required: true
+        - label: I have searched the MinerU [Issues](https://github.com/opendatalab/MinerU/issues) and found no similar bug report.
+          required: true
+        - label: I have searched the MinerU [Discussions](https://github.com/opendatalab/MinerU/discussions) and found no similar bug report.
+          required: true
+
+  - type: textarea
+    id: description
+    attributes:
+      label: Description of the bug | 错误描述
+      description: |
+        Provide console output with error messages and/or screenshots of the bug. | 请提供详细报错信息或者截图
+      placeholder: |
+        💡 ProTip! Include as much information as possible (screenshots, logs, tracebacks etc.) to receive the most helpful response.
+    validations:
+      required: true
+  
+  - type: textarea
+    id: reproduce
+    attributes:
+      label: How to reproduce the bug | 如何复现
+      
+      # Should not word-wrap this description here.
+      description: |
+        If you have questions about the parsing results or encounter errors during execution: | 如对解析结果有疑问或在运行中出现报错等异常:
+        * Provide a minimal reproducible example. | 请提供一个最小可复现的demo。 
+        * The demo should include the complete steps, code, and the PDF file to be parsed. | demo需要包含完整的操作步骤，代码，以及需要解析的PDF文件。
+        * When reporting parsing result anomalies and runtime errors, reproducible PDF files are essential. If the document is too large or confidential, you can print the problematic page(s) via the browser and submit the corresponding example file.
+        * 在反馈解析结果异常和运行时报错时，可复现的PDF文件是必不可少的，如文档过大或涉密，您可通过浏览器打印出出现问题的某一页或某几页再提交相应的示例文件。
+        
+        
+        For problems when building or installing MinerU: | 在构建或安装 MinerU 时遇到的问题:
+        * Give the **exact** build/install commands that were run. | 提供**确切**的构建/安装命令。
+        * Give the **complete** output from these commands. | 提供这些命令的**完整**输出。
+  
+    validations:
+      required: true
+
+#  - type: markdown
+#    attributes:
+#      value: |
+#        # The information below is required.
+
+
+  - type: dropdown
+    id: os_mode
+    attributes:
+      label: Operating System Mode | 操作系统类型
+      #multiple: true
+      options:
+        -
+        - Windows
+        - Linux
+        - MacOS
+    validations:
+      required: true
+
+  - type: textarea
+    id: os_name_version
+    attributes:
+      label: Operating System Version| 操作系统版本
+      #multiple: true
+      description: |
+        * 如果您使用的是Linux系统，请提供Linux系统的**发行版名称**和**版本号**来帮助开发人员排查问题。 
+        * If you are using a Linux system, please provide the Linux distribution and version number to help developers troubleshoot the issue.
+        * 如果您使用的是Windows或MacOS系统，请提供操作系统的**版本号**来帮助开发人员排查问题。
+        * If you are using a Windows or MacOS system, please provide the version number of the operating system to help developers troubleshoot the issue.
+        * 例如：Ubuntu 22.04, CentOS 7.9, MacOS 15.1, Windows 11
+        * For example: Ubuntu 22.04, CentOS 7.9, MacOS 15.1, Windows 11.
+
+    validations:
+      required: true
+
+  - type: dropdown
+    id: python_version
+    attributes:
+      label: Python version | Python 版本
+      #multiple: true
+      # Need quotes around `3.10` otherwise it is treated as a number and shows as `3.1`.
+      options:
+        -
+        - "3.13"
+        - "3.12"
+        - "3.11"
+        - "3.10"
+    validations:
+      required: true
+
+  - type: dropdown
+    id: software_version
+    attributes:
+      label: Software version | 软件版本 (magic-pdf --version)
+      #multiple: false
+      options:
+        -
+        - "1.0.x"
+        - "1.1.x"
+        - "1.2.x"
+        - "1.3.x"
+    validations:
+      required: true
+
+  - type: dropdown
+    id: device_mode
+    attributes:
+      label: Device mode | 设备模式
+      #multiple: true
+      options:
+        -
+        - cpu
+        - cuda
+        - mps
+        - npu
+    validations:
+      required: true

third_party/MinerU/.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,11 @@
+blank_issues_enabled: false
+contact_links:
+  - name: 🙏 Q&A
+    url: https://github.com/opendatalab/MinerU/discussions/categories/q-a
+    about: Ask the community for help
+  - name: 💡 Feature requests and ideas
+    url: https://github.com/opendatalab/MinerU/discussions/categories/ideas
+    about: Share ideas for new features
+  - name: 🙌 Show and tell
+    url: https://github.com/opendatalab/MinerU/discussions/categories/show-and-tell
+    about: Show off something you've made

third_party/MinerU/.github/pull_request_template.md

@@ -0,0 +1,32 @@
+Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.
+
+## Motivation
+
+Please describe the motivation of this PR and the goal you want to achieve through this PR.
+
+## Modification
+
+Please briefly describe what modification is made in this PR.
+
+## BC-breaking (Optional)
+
+Does the modification introduce changes that break the backward compatibility of the downstream repositories?
+If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.
+
+## Use cases (Optional)
+
+If this PR introduces a new feature, it is better to list some use cases here and update the documentation.
+
+## Checklist
+
+**Before PR**:
+
+- [ ] Pre-commit or other linting tools are used to fix the potential lint issues.
+- [ ] Bug fixes are fully covered by unit tests, the case that causes the bug should be added in the unit tests.
+- [ ] The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
+- [ ] The documentation has been modified accordingly, like docstring or example tutorials.
+
+**After PR**:
+
+- [ ] If the modification has potential influence on downstream or other related projects, this PR should be tested with those projects.
+- [ ] CLA has been signed and all committers have signed the CLA in this PR.

third_party/MinerU/.github/workflows/cla.yml

@@ -0,0 +1,43 @@
+name: "MinerU CLA Assistant"
+on:
+  issue_comment:
+    types: [created]
+  pull_request_target:
+    types: [opened,closed,synchronize]
+
+# explicitly configure permissions, in case your GITHUB_TOKEN workflow permissions are set to read-only in repository settings
+permissions:
+  actions: write
+  contents: write # this can be 'read' if the signatures are in remote repository
+  pull-requests: write
+  statuses: write
+
+jobs:
+  CLAAssistant:
+    runs-on: ubuntu-latest
+    steps:
+      - name: "CLA Assistant"
+        if: (github.event.comment.body == 'recheck' || github.event.comment.body == 'I have read the CLA Document and I hereby sign the CLA') || github.event_name == 'pull_request_target'
+        uses: contributor-assistant/github-action@v2.5.0
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          # the below token should have repo scope and must be manually added by you in the repository's secret
+          # This token is required only if you have configured to store the signatures in a remote repository/organization
+          PERSONAL_ACCESS_TOKEN: ${{ secrets.RELEASE_TOKEN }}
+        with:
+          path-to-signatures: 'signatures/version1/cla.json'
+          path-to-document: 'https://github.com/opendatalab/MinerU/blob/master/MinerU_CLA.md' # e.g. a CLA or a DCO document
+          # branch should not be protected
+          branch: 'master'
+          allowlist: myhloli,dt-yy,Focusshang,renpengli01,icecraft,drunkpig,wangbinDL,qiangqiang199,GDDGCZ518,papayalove,conghui,quyuan,LollipopsAndWine
+
+         # the followings are the optional inputs - If the optional inputs are not given, then default values will be taken
+          #remote-organization-name: enter the remote organization name where the signatures should be stored (Default is storing the signatures in the same repository)
+          #remote-repository-name: enter the  remote repository name where the signatures should be stored (Default is storing the signatures in the same repository)
+          #create-file-commit-message: 'For example: Creating file for storing CLA Signatures'
+          #signed-commit-message: 'For example: $contributorName has signed the CLA in $owner/$repo#$pullRequestNo'
+          #custom-notsigned-prcomment: 'pull request comment with Introductory message to ask new contributors to sign'
+          #custom-pr-sign-comment: 'The signature to be committed in order to sign the CLA'
+          #custom-allsigned-prcomment: 'pull request comment when all contributors has signed, defaults to **CLA Assistant Lite bot** All Contributors have signed the CLA.'
+          #lock-pullrequest-aftermerge: false - if you don't want this bot to automatically lock the pull request after merging (default - true)
+          #use-dco-flag: true - If you are using DCO instead of CLA

third_party/MinerU/.github/workflows/cli.yml

@@ -0,0 +1,46 @@
+# This workflow will install Python dependencies, run tests and lint with a variety of Python versions
+# For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python
+
+name: mineru
+on:
+  pull_request:
+    branches:
+      - "master"
+      - "dev"
+    paths-ignore:
+      - "cmds/**"
+      - "**.md"
+  workflow_dispatch:
+jobs:
+  cli-test:
+    if: github.repository == 'opendatalab/MinerU'
+    runs-on: pdf
+    timeout-minutes: 240
+    strategy:
+      fail-fast: true
+
+    steps:
+    - name: PDF cli
+      uses: actions/checkout@v3
+      with:
+        fetch-depth: 2
+
+    - name: install&test
+      run: |
+        source activate mineru
+        conda env list
+        pip show coverage
+        cd $GITHUB_WORKSPACE && sh tests/retry_env.sh
+        cd $GITHUB_WORKSPACE && python tests/clean_coverage.py      
+        cd $GITHUB_WORKSPACE && coverage run -m pytest tests/unittest/ --cov=magic_pdf/  --cov-report html --cov-report term-missing
+        cd $GITHUB_WORKSPACE && python tests/get_coverage.py
+        cd $GITHUB_WORKSPACE && pytest -m P0 -s -v tests/test_cli/test_cli_sdk.py
+
+  notify_to_feishu:
+    if: ${{ always() && !cancelled() && contains(needs.*.result, 'failure')}}
+    needs: cli-test
+    runs-on: pdf
+    steps:
+    - name: notify
+      run: |
+        curl -X POST -H "Content-Type: application/json" -d '{"msg_type":"post","content":{"post":{"zh_cn":{"title":"'${{ github.repository }}' GitHubAction Failed","content":[[{"tag":"text","text":""},{"tag":"a","text":"Please click here for details ","href":"https://github.com/'${{ github.repository }}'/actions/runs/'${GITHUB_RUN_ID}'"},{"tag":"at","user_id":"'$USER_ID'"}]]}}}}'  $WEBHOOK_URL

third_party/MinerU/.github/workflows/huigui.yml

@@ -0,0 +1,61 @@
+# This workflow will install Python dependencies, run tests and lint with a variety of Python versions
+# For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python
+
+name: mineru
+on:
+  push:
+    branches:
+      - "master"
+      - "dev"
+    paths-ignore:
+      - "cmds/**"
+      - "**.md"
+jobs:
+  cli-test:
+    if: github.repository == 'opendatalab/MinerU'
+    runs-on: pdf
+    timeout-minutes: 240
+    strategy:
+      fail-fast: true
+
+    steps:
+    - name: PDF cli
+      uses: actions/checkout@v3
+      with:
+        fetch-depth: 2
+
+    - name: install&test
+      run: |
+        source activate mineru
+        conda env list
+        pip show coverage
+        cd $GITHUB_WORKSPACE && sh tests/retry_env.sh
+        cd $GITHUB_WORKSPACE && python tests/clean_coverage.py      
+        cd $GITHUB_WORKSPACE && coverage run -m pytest tests/unittest/ --cov=magic_pdf/  --cov-report html --cov-report term-missing
+        cd $GITHUB_WORKSPACE && python tests/get_coverage.py
+        cd $GITHUB_WORKSPACE && pytest -s -v tests/test_cli/test_cli_sdk.py
+
+  notify_to_feishu:
+    if: ${{ always() && !cancelled() && contains(needs.*.result, 'failure')}}
+    needs: cli-test
+    runs-on: pdf
+    steps:
+    - name: get_actor
+      run: |
+          metion_list="dt-yy"
+          echo $GITHUB_ACTOR
+          if [[ $GITHUB_ACTOR == "drunkpig" ]]; then
+            metion_list="xuchao"
+          elif [[ $GITHUB_ACTOR == "myhloli" ]]; then
+            metion_list="zhaoxiaomeng"
+          elif [[ $GITHUB_ACTOR == "icecraft" ]]; then
+            metion_list="xurui1"
+          fi
+          echo $metion_list
+          echo "METIONS=$metion_list" >> "$GITHUB_ENV"
+          echo ${{ env.METIONS }}
+
+    - name: notify
+      run: |
+        #echo ${{ secrets.USER_ID }}
+        curl -X POST -H "Content-Type: application/json" -d '{"msg_type":"post","content":{"post":{"zh_cn":{"title":"'${{ github.repository }}' GitHubAction Failed","content":[[{"tag":"text","text":""},{"tag":"a","text":"Please click here for details ","href":"https://github.com/'${{ github.repository }}'/actions/runs/'${GITHUB_RUN_ID}'"},{"tag":"at","user_id":"'$USER_ID'"}]]}}}}'  $WEBHOOK_URL

third_party/MinerU/.github/workflows/python-package.yml

@@ -0,0 +1,143 @@
+# This workflow will install Python dependencies, run tests and lint with a variety of Python versions
+# For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python
+
+name: Python package
+
+on:
+  push:
+    tags:
+      - '*released'
+  workflow_dispatch:
+
+
+jobs:
+
+  update-version:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          ref: master
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.10"
+
+      - name: Update version.py
+        run: |
+          python update_version.py
+
+      - name: Verify version.py
+        run: |
+          ls -l magic_pdf/libs/version.py
+          cat magic_pdf/libs/version.py
+
+      - name: Commit changes
+        run: |
+          git config --local user.email "moe@myhloli.com"
+          git config --local user.name "myhloli"
+          git add magic_pdf/libs/version.py
+          if git diff-index --quiet HEAD; then
+            echo "No changes to commit"
+          else
+            git commit -m "Update version.py with new version"
+          fi
+        id: commit_changes
+
+      - name: Push changes
+        if: steps.commit_changes.outcome == 'success'
+        env:
+          GITHUB_TOKEN: ${{ secrets.RELEASE_TOKEN }}
+        run: |
+          git push origin HEAD:master
+
+  check-install:
+    needs: [ update-version ]
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.10", "3.11", "3.12", "3.13"]
+
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+      with:
+        ref: master
+        fetch-depth: 0
+
+    - name: Verify version.py
+      run: |
+        ls -l magic_pdf/libs/version.py
+        cat magic_pdf/libs/version.py
+
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v5
+      with:
+        python-version: ${{ matrix.python-version }}
+
+    - name: Install magic-pdf
+      run: |
+        python -m pip install --upgrade pip
+        pip install -e .[full]
+
+  build:
+    needs: [ check-install ]
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: [ "3.10"]
+
+    steps:
+
+    - name: Checkout code
+      uses: actions/checkout@v4
+      with:
+        ref: master
+        fetch-depth: 0
+
+    - name: Install wheel
+      run: |
+        python -m pip install wheel
+
+    - name: Build wheel
+      run: |
+        python setup.py bdist_wheel
+
+    - name: Upload artifact
+      uses: actions/upload-artifact@v4
+      with:
+        name: wheel-file
+        path: dist/*.whl
+        retention-days: 30
+
+  release:
+    needs: [ build ]
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Download artifact
+        uses: actions/download-artifact@v4
+        with:
+          name: wheel-file
+          path: dist
+
+      - name: Create and Upload Release
+        id: create_release
+        uses: softprops/action-gh-release@4634c16e79c963813287e889244c50009e7f0981
+        with:
+          files: './dist/*.whl'
+        env:
+          GITHUB_TOKEN: ${{ secrets.RELEASE_TOKEN }}
+
+      - name: Publish distribution to PyPI
+        run: |
+          pip install -U twine id keyring packaging readme-renderer requests requests-toolbelt rfc3986 rich urllib3
+          twine check dist/*
+          twine upload dist/* -u __token__ -p ${{ secrets.PYPI_TOKEN }}

third_party/MinerU/.github/workflows/rerun.yml

@@ -0,0 +1,23 @@
+name: check-status
+
+on:
+  workflow_run:
+    workflows: [ci]
+    types: [completed]
+
+jobs:
+  on-failure:
+    runs-on: pdf
+    permissions:
+      actions: write
+    if: ${{ (github.event.workflow_run.head_branch == 'master') && github.event.workflow_run.conclusion == 'failure' && github.event.workflow_run.run_attempt < 3 }}
+    steps:
+      - run: |
+          echo 'The triggering workflow failed'
+          sleep 600
+          curl -L \
+          -X POST \
+          -H "Accept: application/vnd.github+json" \
+          -H "Authorization: Bearer ${{ github.token }}" \
+          -H "X-GitHub-Api-Version: 2022-11-28" \
+          https://api.github.com/repos/${{ github.repository }}/actions/runs/${{ github.event.workflow_run.id }}/rerun-failed-jobs 

third_party/MinerU/.gitignore

@@ -0,0 +1,53 @@
+*.tar
+*.tar.gz
+*.zip
+venv*/
+envs/
+slurm_logs/
+
+sync1.sh
+data_preprocess_pj1
+data-preparation1
+__pycache__
+*.log
+*.pyc
+.vscode
+debug/
+*.ipynb
+.idea
+
+# vscode history
+.history
+
+.DS_Store
+.env
+
+bad_words/
+bak/
+
+app/tests/*
+temp/
+tmp/
+tmp
+.vscode
+.vscode/
+ocr_demo
+.coveragerc
+/app/common/__init__.py
+/magic_pdf/config/__init__.py
+source.dev.env
+
+tmp
+
+projects/web/node_modules
+projects/web/dist
+
+projects/web_demo/web_demo/static/
+cli_debug/
+debug_utils/
+
+# sphinx docs
+_build/
+
+
+output/

third_party/MinerU/.pre-commit-config.yaml

@@ -0,0 +1,47 @@
+repos:
+  - repo: https://github.com/PyCQA/flake8
+    rev: 5.0.4
+    hooks:
+      - id: flake8
+        args: ["--max-line-length=150", "--ignore=E131,E125,W503,W504,E203"]
+  - repo: https://github.com/PyCQA/isort
+    rev: 5.11.5
+    hooks:
+      - id: isort
+  - repo: https://github.com/pre-commit/mirrors-yapf
+    rev: v0.32.0
+    hooks:
+      - id: yapf
+        args: ["--style={based_on_style: google, column_limit: 150, indent_width: 4}"]
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.2.1
+    hooks:
+      - id: codespell
+        args: ['--skip', '*.json']
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.3.0
+    hooks:
+      - id: trailing-whitespace
+      - id: check-yaml
+      - id: end-of-file-fixer
+      - id: requirements-txt-fixer
+      - id: double-quote-string-fixer
+      - id: check-merge-conflict
+      - id: fix-encoding-pragma
+        args: ["--remove"]
+      - id: mixed-line-ending
+        args: ["--fix=lf"]
+  - repo: https://github.com/executablebooks/mdformat
+    rev: 0.7.9
+    hooks:
+      - id: mdformat
+        args: ["--number", "--table-width", "200"]
+        additional_dependencies:
+          - mdformat-openmmlab
+          - mdformat_frontmatter
+          - linkify-it-py
+  - repo: https://github.com/myint/docformatter
+    rev: v1.3.1
+    hooks:
+      - id: docformatter
+        args: ["--in-place", "--wrap-descriptions", "119"]

third_party/MinerU/.readthedocs.yaml

@@ -0,0 +1,16 @@
+version: 2
+
+build:
+  os: ubuntu-22.04
+  tools:
+    python: "3.10"
+
+formats:
+  - epub
+
+python:
+  install:
+    - requirements: next_docs/zh_cn/requirements.txt
+
+sphinx:
+  configuration: next_docs/zh_cn/conf.py

third_party/MinerU/LICENSE.md

@@ -0,0 +1,662 @@
+                    GNU AFFERO GENERAL PUBLIC LICENSE
+                       Version 3, 19 November 2007
+
+ Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+                            Preamble
+
+  The GNU Affero General Public License is a free, copyleft license for
+software and other kinds of works, specifically designed to ensure
+cooperation with the community in the case of network server software.
+
+  The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works.  By contrast,
+our General Public Licenses are intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+them if you wish), that you receive source code or can get it if you
+want it, that you can change the software or use pieces of it in new
+free programs, and that you know you can do these things.
+
+  Developers that use our General Public Licenses protect your rights
+with two steps: (1) assert copyright on the software, and (2) offer
+you this License which gives you legal permission to copy, distribute
+and/or modify the software.
+
+  A secondary benefit of defending all users' freedom is that
+improvements made in alternate versions of the program, if they
+receive widespread use, become available for other developers to
+incorporate.  Many developers of free software are heartened and
+encouraged by the resulting cooperation.  However, in the case of
+software used on network servers, this result may fail to come about.
+The GNU General Public License permits making a modified version and
+letting the public access it on a server without ever releasing its
+source code to the public.
+
+  The GNU Affero General Public License is designed specifically to
+ensure that, in such cases, the modified source code becomes available
+to the community.  It requires the operator of a network server to
+provide the source code of the modified version running there to the
+users of that server.  Therefore, public use of a modified version, on
+a publicly accessible server, gives the public access to the source
+code of the modified version.
+
+  An older license, called the Affero General Public License and
+published by Affero, was designed to accomplish similar goals.  This is
+a different license, not a version of the Affero GPL, but Affero has
+released a new version of the Affero GPL which permits relicensing under
+this license.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                       TERMS AND CONDITIONS
+
+  0. Definitions.
+
+  "This License" refers to version 3 of the GNU Affero General Public License.
+
+  "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
+
+  "The Program" refers to any copyrightable work licensed under this
+License.  Each licensee is addressed as "you".  "Licensees" and
+"recipients" may be individuals or organizations.
+
+  To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy.  The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
+
+  A "covered work" means either the unmodified Program or a work based
+on the Program.
+
+  To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy.  Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
+
+  To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies.  Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
+
+  An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License.  If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
+
+  1. Source Code.
+
+  The "source code" for a work means the preferred form of the work
+for making modifications to it.  "Object code" means any non-source
+form of a work.
+
+  A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
+
+  The "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form.  A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
+
+  The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities.  However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work.  For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
+
+  The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
+
+  The Corresponding Source for a work in source code form is that
+same work.
+
+  2. Basic Permissions.
+
+  All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met.  This License explicitly affirms your unlimited
+permission to run the unmodified Program.  The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work.  This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+  You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force.  You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright.  Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+  Conveying under any other circumstances is permitted solely under
+the conditions stated below.  Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+  3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+  No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+  When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+  4. Conveying Verbatim Copies.
+
+  You may convey verbatim copies of the Program's source code as you
+receive it, in any medium, provided that you conspicuously and
+appropriately publish on each copy an appropriate copyright notice;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+  You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+  5. Conveying Modified Source Versions.
+
+  You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+    a) The work must carry prominent notices stating that you modified
+    it, and giving a relevant date.
+
+    b) The work must carry prominent notices stating that it is
+    released under this License and any conditions added under section
+    7.  This requirement modifies the requirement in section 4 to
+    "keep intact all notices".
+
+    c) You must license the entire work, as a whole, under this
+    License to anyone who comes into possession of a copy.  This
+    License will therefore apply, along with any applicable section 7
+    additional terms, to the whole of the work, and all its parts,
+    regardless of how they are packaged.  This License gives no
+    permission to license the work in any other way, but it does not
+    invalidate such permission if you have separately received it.
+
+    d) If the work has interactive user interfaces, each must display
+    Appropriate Legal Notices; however, if the Program has interactive
+    interfaces that do not display Appropriate Legal Notices, your
+    work need not make them do so.
+
+  A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit.  Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+  6. Conveying Non-Source Forms.
+
+  You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+    a) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by the
+    Corresponding Source fixed on a durable physical medium
+    customarily used for software interchange.
+
+    b) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by a
+    written offer, valid for at least three years and valid for as
+    long as you offer spare parts or customer support for that product
+    model, to give anyone who possesses the object code either (1) a
+    copy of the Corresponding Source for all the software in the
+    product that is covered by this License, on a durable physical
+    medium customarily used for software interchange, for a price no
+    more than your reasonable cost of physically performing this
+    conveying of source, or (2) access to copy the
+    Corresponding Source from a network server at no charge.
+
+    c) Convey individual copies of the object code with a copy of the
+    written offer to provide the Corresponding Source.  This
+    alternative is allowed only occasionally and noncommercially, and
+    only if you received the object code with such an offer, in accord
+    with subsection 6b.
+
+    d) Convey the object code by offering access from a designated
+    place (gratis or for a charge), and offer equivalent access to the
+    Corresponding Source in the same way through the same place at no
+    further charge.  You need not require recipients to copy the
+    Corresponding Source along with the object code.  If the place to
+    copy the object code is a network server, the Corresponding Source
+    may be on a different server (operated by you or a third party)
+    that supports equivalent copying facilities, provided you maintain
+    clear directions next to the object code saying where to find the
+    Corresponding Source.  Regardless of what server hosts the
+    Corresponding Source, you remain obligated to ensure that it is
+    available for as long as needed to satisfy these requirements.
+
+    e) Convey the object code using peer-to-peer transmission, provided
+    you inform other peers where the object code and Corresponding
+    Source of the work are being offered to the general public at no
+    charge under subsection 6d.
+
+  A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+  A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling.  In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage.  For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product.  A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+  "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source.  The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+  If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information.  But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+  The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed.  Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+  Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+  7. Additional Terms.
+
+  "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law.  If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+  When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it.  (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.)  You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+  Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+    a) Disclaiming warranty or limiting liability differently from the
+    terms of sections 15 and 16 of this License; or
+
+    b) Requiring preservation of specified reasonable legal notices or
+    author attributions in that material or in the Appropriate Legal
+    Notices displayed by works containing it; or
+
+    c) Prohibiting misrepresentation of the origin of that material, or
+    requiring that modified versions of such material be marked in
+    reasonable ways as different from the original version; or
+
+    d) Limiting the use for publicity purposes of names of licensors or
+    authors of the material; or
+
+    e) Declining to grant rights under trademark law for use of some
+    trade names, trademarks, or service marks; or
+
+    f) Requiring indemnification of licensors and authors of that
+    material by anyone who conveys the material (or modified versions of
+    it) with contractual assumptions of liability to the recipient, for
+    any liability that these contractual assumptions directly impose on
+    those licensors and authors.
+
+  All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10.  If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term.  If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+  If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+  Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+  8. Termination.
+
+  You may not propagate or modify a covered work except as expressly
+provided under this License.  Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+  However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+  Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+  Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License.  If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+  9. Acceptance Not Required for Having Copies.
+
+  You are not required to accept this License in order to receive or
+run a copy of the Program.  Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance.  However,
+nothing other than this License grants you permission to propagate or
+modify any covered work.  These actions infringe copyright if you do
+not accept this License.  Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+  10. Automatic Licensing of Downstream Recipients.
+
+  Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License.  You are not responsible
+for enforcing compliance by third parties with this License.
+
+  An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations.  If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+  You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License.  For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+  11. Patents.
+
+  A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based.  The
+work thus licensed is called the contributor's "contributor version".
+
+  A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version.  For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
+this License.
+
+  Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+  In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement).  To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+  If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients.  "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+  If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+  A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License.  You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+  Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+  12. No Surrender of Others' Freedom.
+
+  If conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all.  For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying the Program.
+
+  13. Remote Network Interaction; Use with the GNU General Public License.
+
+  Notwithstanding any other provision of this License, if you modify the
+Program, your modified version must prominently offer all users
+interacting with it remotely through a computer network (if your version
+supports such interaction) an opportunity to receive the Corresponding
+Source of your version by providing access to the Corresponding Source
+from a network server at no charge, through some standard or customary
+means of facilitating copying of software.  This Corresponding Source
+shall include the Corresponding Source for any work covered by version 3
+of the GNU General Public License that is incorporated pursuant to the
+following paragraph.
+
+  Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU General Public License into a single
+combined work, and to convey the resulting work.  The terms of this
+License will continue to apply to the part which is the covered work,
+but the work with which it is combined will remain governed by version
+3 of the GNU General Public License.
+
+  14. Revised Versions of this License.
+
+  The Free Software Foundation may publish revised and/or new versions of
+the GNU Affero General Public License from time to time.  Such new versions
+will be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+  Each version is given a distinguishing version number.  If the
+Program specifies that a certain numbered version of the GNU Affero General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation.  If the Program does not specify a version number of the
+GNU Affero General Public License, you may choose any version ever published
+by the Free Software Foundation.
+
+  If the Program specifies that a proxy can decide which future
+versions of the GNU Affero General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
+
+  Later license versions may give you additional or different
+permissions.  However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
+
+  15. Disclaimer of Warranty.
+
+  THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
+APPLICABLE LAW.  EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
+HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
+OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
+THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
+IS WITH YOU.  SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
+ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+  16. Limitation of Liability.
+
+  IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
+GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
+USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
+DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
+PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
+EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGES.
+
+  17. Interpretation of Sections 15 and 16.
+
+  If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+state the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published
+    by the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+  If your software can interact with users remotely through a computer
+network, you should also make sure that it provides a way for users to
+get its source.  For example, if your program is a web application, its
+interface could display a "Source" link that leads users to an archive
+of the code.  There are many ways you could offer source, and different
+solutions will be better for different programs; see section 13 for the
+specific requirements.
+
+  You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU AGPL, see
+<https://www.gnu.org/licenses/>.
+

third_party/MinerU/MinerU_CLA.md

@@ -0,0 +1,14 @@
+# MinerU Contributor License Agreement
+In order to clarify the intellectual property license granted with Contributions from any person or entity, the open source project MinerU ("MinerU") must have a Contributor License Agreement (CLA) on file that has been signed by each Contributor, indicating agreement to the license terms below. This license is for your protection as a Contributor as well as the protection of MinerU and its users; it does not change your rights to use your own Contributions for any other purpose.
+
+You accept and agree to the following terms and conditions for Your present and future Contributions submitted to MinerU. Except for the license granted herein to MinerU and recipients of software distributed by MinerU, You reserve all right, title, and interest in and to Your Contributions.
+
+1. Definitions. "You" (or "Your") shall mean the copyright owner or legal entity authorized by the copyright owner that is making this Agreement with MinerU. For legal entities, the entity making a Contribution and all other entities that control, are controlled by, or are under common control with that entity are considered to be a single Contributor. For the purposes of this definition, "control" means (i) the power, direct or indirect, to cause the direction or management of such entity, whether by contract or otherwise, or (ii) ownership of fifty percent (50%) or more of the outstanding shares, or (iii) beneficial ownership of such entity. "Contribution" shall mean the code, documentation or any original work of authorship, including any modifications or additions to an existing work, that is intentionally submitted by You to MinerU for inclusion in, or documentation of, any of the products owned or managed by MinerU (the "Work"). For the purposes of this definition, "submitted" means any form of electronic, verbal, or written communication sent to MinerU or its representatives, including but not limited to communication on electronic mailing lists, source code control systems, and issue tracking systems that are managed by, or on behalf of, MinerU for the purpose of discussing and improving the Work, but excluding communication that is conspicuously marked or otherwise designated in writing by You as "Not a Contribution."
+2. Grant of Copyright License. Subject to the terms and conditions of this Agreement, You hereby grant to MinerU and to recipients of software distributed by MinerU a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable copyright license to reproduce, prepare derivative works of, publicly display, publicly perform, sublicense, and distribute Your Contributions and such derivative works.
+3. Grant of Patent License. Subject to the terms and conditions of this Agreement, You hereby grant to MinerU and to recipients of software distributed by MinerU a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Work, where such license applies only to those patent claims licensable by You that are necessarily infringed by Your Contribution(s) alone or by combination of Your Contribution(s) with the Work to which such Contribution(s) was submitted. If any entity institutes patent litigation against You or any other entity (including a cross-claim or counterclaim in a lawsuit) alleging that Your Contribution, or the Work to which You have contributed, constitutes direct or contributory patent infringement, then any patent licenses granted to that entity under this Agreement for that Contribution or Work shall terminate as of the date such litigation is filed.
+4. You represent that You are legally entitled to grant the above license. If You are an entity, You represent further that each of Your employee designated by You is authorized to submit Contributions on behalf of You. If You are an individual and Your employer(s) has rights to intellectual property that You create that includes Your Contributions, You represent further that You have received permission to make Contributions on behalf of that employer, that Your employer has waived such rights for Your Contributions to MinerU, or that Your employer has executed a separate CLA with MinerU.
+5. If you do post content or submit material on MinerU and unless we indicate otherwise, you grant MinerU a nonexclusive, royalty-free, perpetual, irrevocable, and fully sublicensable right to use, reproduce, modify, adapt, publish, perform, translate, create derivative works from, distribute, and display such content throughout the world in any media. You grant MinerU and sublicensees the right to use your GitHub Public Profile, including but not limited to name, that you submit in connection with such content. You represent and warrant that you own or otherwise control all of the rights to the content that you post; that the content is accurate; that use of the content you supply does not violate this policy and will not cause injury to any person or entity; and that you will indemnify MinerU for all claims resulting from content you supply. MinerU has the right but not the obligation to monitor and edit or remove any activity or content. MinerU takes no responsibility and assumes no liability for any content posted by you or any third party.
+6. You represent that each of Your Contributions is Your original creation. Should You wish to submit work that is not Your original creation, You may submit it to MinerU separately from any Contribution, identifying the complete details of its source and of any license or other restriction (including, but not limited to, related patents, trademarks, and license agreements) of which You are personally aware, and conspicuously marking the work as "Submitted on behalf of a third-party: [named here]".
+7. You are not expected to provide support for Your Contributions, except to the extent You desire to provide support. You may provide support for free, for a fee, or not at all. Unless required by applicable law or agreed to in writing, You provide Your Contributions on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied, including, without limitation, any warranties or conditions of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A PARTICULAR PURPOSE.
+8. You agree to notify MinerU of any facts or circumstances of which You become aware that would make these representations inaccurate in any respect.
+9. MinerU reserves the right to update or change this Agreement at any time, by posting the most current version of the Agreement on MinerU, with a new Effective Date shown on Jul. 24th, 2024. All such changes in the Agreement are effective from the Effective Date. Your continued use of MinerU after we post any such changes signifies your agreement to those changes. If you do not agree to the then-current Agreement, you must immediately discontinue using MinerU.

third_party/MinerU/README.md

@@ -0,0 +1,604 @@
+<div align="center" xmlns="http://www.w3.org/1999/html">
+<!-- logo -->
+<p align="center">
+  <img src="docs/images/MinerU-logo.png" width="300px" style="vertical-align:middle;">
+</p>
+
+<!-- icon -->
+
+[![stars](https://img.shields.io/github/stars/opendatalab/MinerU.svg)](https://github.com/opendatalab/MinerU)
+[![forks](https://img.shields.io/github/forks/opendatalab/MinerU.svg)](https://github.com/opendatalab/MinerU)
+[![open issues](https://img.shields.io/github/issues-raw/opendatalab/MinerU)](https://github.com/opendatalab/MinerU/issues)
+[![issue resolution](https://img.shields.io/github/issues-closed-raw/opendatalab/MinerU)](https://github.com/opendatalab/MinerU/issues)
+[![PyPI version](https://img.shields.io/pypi/v/magic-pdf)](https://pypi.org/project/magic-pdf/)
+[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/magic-pdf)](https://pypi.org/project/magic-pdf/)
+[![Downloads](https://static.pepy.tech/badge/magic-pdf)](https://pepy.tech/project/magic-pdf)
+[![Downloads](https://static.pepy.tech/badge/magic-pdf/month)](https://pepy.tech/project/magic-pdf)
+
+[![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://mineru.net/OpenSourceTools/Extractor?source=github)
+[![HuggingFace](https://img.shields.io/badge/Demo_on_HuggingFace-yellow.svg?logo=data:image/png;base64,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&labelColor=white)](https://huggingface.co/spaces/opendatalab/MinerU)
+[![ModelScope](https://img.shields.io/badge/Demo_on_ModelScope-purple?logo=data:image/svg+xml;base64,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&labelColor=white)](https://www.modelscope.cn/studios/OpenDataLab/MinerU)
+[![Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/gist/myhloli/3b3a00a4a0a61577b6c30f989092d20d/mineru_demo.ipynb)
+[![Paper](https://img.shields.io/badge/Paper-arXiv-green)](https://arxiv.org/abs/2409.18839)
+
+
+<a href="https://trendshift.io/repositories/11174" target="_blank"><img src="https://trendshift.io/api/badge/repositories/11174" alt="opendatalab%2FMinerU | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+
+<!-- language -->
+
+[English](README.md) | [简体中文](README_zh-CN.md)
+
+<!-- hot link -->
+
+<p align="center">
+<a href="https://github.com/opendatalab/PDF-Extract-Kit">PDF-Extract-Kit: High-Quality PDF Extraction Toolkit</a>🔥🔥🔥
+<br>
+<br>
+<a href="https://mineru.net/client?source=github">
+Easier to use: Just grab MinerU Desktop. No coding, no login, just a simple interface and smooth interactions. Enjoy it without any fuss!</a>🚀🚀🚀
+
+</p>
+
+<!-- join us -->
+
+<p align="center">
+    👋 join us on <a href="https://discord.gg/Tdedn9GTXq" target="_blank">Discord</a> and <a href="http://mineru.space/s/V85Yl" target="_blank">WeChat</a>
+</p>
+
+</div>
+
+# Changelog
+- 2025/04/29 1.3.10 Released
+  - Support for custom formula delimiters can be achieved by modifying the `latex-delimiter-config` item in the `magic-pdf.json` file under the user directory.
+- 2025/04/27 1.3.9 Released  
+  - Optimized the formula parsing function to improve the success rate of formula rendering
+- 2025/04/23 1.3.8 Released
+  - The default `ocr` model (`ch`) has been updated to `PP-OCRv4_server_rec_doc` (model update required)
+    - `PP-OCRv4_server_rec_doc` is trained on a mix of more Chinese document data and PP-OCR training data, enhancing recognition capabilities for some traditional Chinese characters, Japanese, and special characters. It supports over 15,000 recognizable characters, improving text recognition in documents while also boosting general text recognition.
+    - [Performance comparison between PP-OCRv4_server_rec_doc, PP-OCRv4_server_rec, and PP-OCRv4_mobile_rec](https://paddlepaddle.github.io/PaddleX/latest/en/module_usage/tutorials/ocr_modules/text_recognition.html#ii-supported-model-list)
+    - Verified results show that the `PP-OCRv4_server_rec_doc` model significantly improves accuracy in both single-language (`Chinese`, `English`, `Japanese`, `Traditional Chinese`) and mixed-language scenarios, with speed comparable to `PP-OCRv4_server_rec`, making it suitable for most use cases.
+    - In a small number of pure English scenarios, the `PP-OCRv4_server_rec_doc` model may encounter word concatenation issues, whereas `PP-OCRv4_server_rec` performs better in such cases. Therefore, we have retained the `PP-OCRv4_server_rec` model, which users can invoke by passing the parameter `lang='ch_server'`(python api) or `--lang ch_server`(cli).
+- 2025/04/22 1.3.7 Released
+  - Fixed the issue where the `lang` parameter was ineffective during table parsing model initialization.
+  - Fixed the significant slowdown in OCR and table parsing speed in `cpu` mode.
+- 2025/04/16 1.3.4 Released
+  - Slightly improved the speed of OCR detection by removing some unused blocks.
+  - Fixed page-level sorting errors caused by footnotes in certain cases.
+- 2025/04/12 1.3.2 released
+  - Fixed the issue of incompatible dependency package versions when installing in Python 3.13 environment on Windows systems.
+  - Optimized memory usage during batch inference.
+  - Improved the parsing effect of tables rotated by 90 degrees.
+  - Enhanced the parsing accuracy for large tables in financial report samples.
+  - Fixed the occasional word concatenation issue in English text areas when OCR language is not specified.(The model needs to be updated)
+- 2025/04/08 1.3.1 released, fixed some compatibility issues
+  - Supported Python 3.13
+  - Made the final adaptation for some outdated Linux systems (e.g., CentOS 7), and no further support will be guaranteed for subsequent versions. [Installation Instructions](https://github.com/opendatalab/MinerU/issues/1004)
+- 2025/04/03 1.3.0 released, in this version we made many optimizations and improvements:
+  - Installation and compatibility optimization
+    - By removing the use of `layoutlmv3` in layout, resolved compatibility issues caused by `detectron2`.
+    - Torch version compatibility extended to 2.2~2.6 (excluding 2.5).
+    - CUDA compatibility supports 11.8/12.4/12.6/12.8 (CUDA version determined by torch), resolving compatibility issues for some users with 50-series and H-series GPUs.
+    - Python compatible versions expanded to 3.10~3.12, solving the problem of automatic downgrade to 0.6.1 during installation in non-3.10 environments.
+    - Offline deployment process optimized; no internet connection required after successful deployment to download any model files.
+  - Performance optimization
+    - By supporting batch processing of multiple PDF files ([script example](demo/batch_demo.py)), improved parsing speed for small files in batches (compared to version 1.0.1, formula parsing speed increased by over 1400%, overall parsing speed increased by over 500%).
+    - Optimized loading and usage of the mfr model, reducing GPU memory usage and improving parsing speed (requires re-execution of the [model download process](docs/how_to_download_models_en.md) to obtain incremental updates of model files).
+    - Optimized GPU memory usage, requiring only a minimum of 6GB to run this project.
+    - Improved running speed on MPS devices.
+  - Parsing effect optimization
+    - Updated the mfr model to `unimernet(2503)`, solving the issue of lost line breaks in multi-line formulas.
+  - Usability Optimization
+    - By using `paddleocr2torch`, completely replaced the use of the `paddle` framework and `paddleocr` in the project, resolving conflicts between `paddle` and `torch`, as well as thread safety issues caused by the `paddle` framework.
+    - Added a real-time progress bar during the parsing process to accurately track progress, making the wait less painful.
+<details>
+<summary>2025/03/03 1.2.1 released</summary>
+<ul>
+  <li>Fixed the impact on punctuation marks during full-width to half-width conversion of letters and numbers</li>
+  <li>Fixed caption matching inaccuracies in certain scenarios</li>
+  <li>Fixed formula span loss issues in certain scenarios</li>
+</ul>
+</details>
+
+<details>
+<summary>2025/02/24 1.2.0 released</summary>
+<p>This version includes several fixes and improvements to enhance parsing efficiency and accuracy:</p>
+<ul>
+  <li><strong>Performance Optimization</strong>
+    <ul>
+      <li>Increased classification speed for PDF documents in auto mode.</li>
+    </ul>
+  </li>
+  <li><strong>Parsing Optimization</strong>
+    <ul>
+      <li>Improved parsing logic for documents containing watermarks, significantly enhancing the parsing results for such documents.</li>
+      <li>Enhanced the matching logic for multiple images/tables and captions within a single page, improving the accuracy of image-text matching in complex layouts.</li>
+    </ul>
+  </li>
+  <li><strong>Bug Fixes</strong>
+    <ul>
+      <li>Fixed an issue where image/table spans were incorrectly filled into text blocks under certain conditions.</li>
+      <li>Resolved an issue where title blocks were empty in some cases.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2025/01/22 1.1.0 released</summary>
+<p>In this version we have focused on improving parsing accuracy and efficiency:</p>
+<ul>
+  <li><strong>Model capability upgrade</strong> (requires re-executing the <a href="https://github.com/opendatalab/MinerU/blob/master/docs/how_to_download_models_en.md">model download process</a> to obtain incremental updates of model files)
+    <ul>
+      <li>The layout recognition model has been upgraded to the latest <code>doclayout_yolo(2501)</code> model, improving layout recognition accuracy.</li>
+      <li>The formula parsing model has been upgraded to the latest <code>unimernet(2501)</code> model, improving formula recognition accuracy.</li>
+    </ul>
+  </li>
+  <li><strong>Performance optimization</strong>
+    <ul>
+      <li>On devices that meet certain configuration requirements (16GB+ VRAM), by optimizing resource usage and restructuring the processing pipeline, overall parsing speed has been increased by more than 50%.</li>
+    </ul>
+  </li>
+  <li><strong>Parsing effect optimization</strong>
+    <ul>
+      <li>Added a new heading classification feature (testing version, enabled by default) to the online demo (<a href="https://mineru.net/OpenSourceTools/Extractor">mineru.net</a>/<a href="https://huggingface.co/spaces/opendatalab/MinerU">huggingface</a>/<a href="https://www.modelscope.cn/studios/OpenDataLab/MinerU">modelscope</a>), which supports hierarchical classification of headings, thereby enhancing document structuring.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2025/01/10 1.0.1 released</summary>
+<p>This is our first official release, where we have introduced a completely new API interface and enhanced compatibility through extensive refactoring, as well as a brand new automatic language identification feature:</p>
+<ul>
+  <li><strong>New API Interface</strong>
+    <ul>
+      <li>For the data-side API, we have introduced the Dataset class, designed to provide a robust and flexible data processing framework. This framework currently supports a variety of document formats, including images (.jpg and .png), PDFs, Word documents (.doc and .docx), and PowerPoint presentations (.ppt and .pptx). It ensures effective support for data processing tasks ranging from simple to complex.</li>
+      <li>For the user-side API, we have meticulously designed the MinerU processing workflow as a series of composable Stages. Each Stage represents a specific processing step, allowing users to define new Stages according to their needs and creatively combine these stages to customize their data processing workflows.</li>
+    </ul>
+  </li>
+  <li><strong>Enhanced Compatibility</strong>
+    <ul>
+      <li>By optimizing the dependency environment and configuration items, we ensure stable and efficient operation on ARM architecture Linux systems.</li>
+      <li>We have deeply integrated with Huawei Ascend NPU acceleration, providing autonomous and controllable high-performance computing capabilities. This supports the localization and development of AI application platforms in China. <a href="https://github.com/opendatalab/MinerU/blob/master/docs/README_Ascend_NPU_Acceleration_zh_CN.md">Ascend NPU Acceleration</a></li>
+    </ul>
+  </li>
+  <li><strong>Automatic Language Identification</strong>
+    <ul>
+      <li>By introducing a new language recognition model, setting the <code>lang</code> configuration to <code>auto</code> during document parsing will automatically select the appropriate OCR language model, improving the accuracy of scanned document parsing.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2024/11/22 0.10.0 released</summary>
+<p>Introducing hybrid OCR text extraction capabilities:</p>
+<ul>
+  <li>Significantly improved parsing performance in complex text distribution scenarios such as dense formulas, irregular span regions, and text represented by images.</li>
+  <li>Combines the dual advantages of accurate content extraction and faster speed in text mode, and more precise span/line region recognition in OCR mode.</li>
+</ul>
+</details>
+
+<details>
+<summary>2024/11/15 0.9.3 released</summary>
+<p>Integrated <a href="https://github.com/RapidAI/RapidTable">RapidTable</a> for table recognition, improving single-table parsing speed by more than 10 times, with higher accuracy and lower GPU memory usage.</p>
+</details>
+
+<details>
+<summary>2024/11/06 0.9.2 released</summary>
+<p>Integrated the <a href="https://huggingface.co/U4R/StructTable-InternVL2-1B">StructTable-InternVL2-1B</a> model for table recognition functionality.</p>
+</details>
+
+<details>
+<summary>2024/10/31 0.9.0 released</summary>
+<p>This is a major new version with extensive code refactoring, addressing numerous issues, improving performance, reducing hardware requirements, and enhancing usability:</p>
+<ul>
+  <li>Refactored the sorting module code to use <a href="https://github.com/ppaanngggg/layoutreader">layoutreader</a> for reading order sorting, ensuring high accuracy in various layouts.</li>
+  <li>Refactored the paragraph concatenation module to achieve good results in cross-column, cross-page, cross-figure, and cross-table scenarios.</li>
+  <li>Refactored the list and table of contents recognition functions, significantly improving the accuracy of list blocks and table of contents blocks, as well as the parsing of corresponding text paragraphs.</li>
+  <li>Refactored the matching logic for figures, tables, and descriptive text, greatly enhancing the accuracy of matching captions and footnotes to figures and tables, and reducing the loss rate of descriptive text to near zero.</li>
+  <li>Added multi-language support for OCR, supporting detection and recognition of 84 languages. For the list of supported languages, see <a href="https://paddlepaddle.github.io/PaddleOCR/latest/en/ppocr/blog/multi_languages.html#5-support-languages-and-abbreviations">OCR Language Support List</a>.</li>
+  <li>Added memory recycling logic and other memory optimization measures, significantly reducing memory usage. The memory requirement for enabling all acceleration features except table acceleration (layout/formula/OCR) has been reduced from 16GB to 8GB, and the memory requirement for enabling all acceleration features has been reduced from 24GB to 10GB.</li>
+  <li>Optimized configuration file feature switches, adding an independent formula detection switch to significantly improve speed and parsing results when formula detection is not needed.</li>
+  <li>Integrated <a href="https://github.com/opendatalab/PDF-Extract-Kit">PDF-Extract-Kit 1.0</a>:
+    <ul>
+      <li>Added the self-developed <code>doclayout_yolo</code> model, which speeds up processing by more than 10 times compared to the original solution while maintaining similar parsing effects, and can be freely switched with <code>layoutlmv3</code> via the configuration file.</li>
+      <li>Upgraded formula parsing to <code>unimernet 0.2.1</code>, improving formula parsing accuracy while significantly reducing memory usage.</li>
+      <li>Due to the repository change for <code>PDF-Extract-Kit 1.0</code>, you need to re-download the model. Please refer to <a href="https://github.com/opendatalab/MinerU/blob/master/docs/how_to_download_models_en.md">How to Download Models</a> for detailed steps.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2024/09/27 Version 0.8.1 released</summary>
+<p>Fixed some bugs, and providing a <a href="https://github.com/opendatalab/MinerU/blob/master/projects/web_demo/README.md">localized deployment version</a> of the <a href="https://opendatalab.com/OpenSourceTools/Extractor/PDF/">online demo</a> and the <a href="https://github.com/opendatalab/MinerU/blob/master/projects/web/README.md">front-end interface</a>.</p>
+</details>
+
+<details>
+<summary>2024/09/09 Version 0.8.0 released</summary>
+<p>Supporting fast deployment with Dockerfile, and launching demos on Huggingface and Modelscope.</p>
+</details>
+
+<details>
+<summary>2024/08/30 Version 0.7.1 released</summary>
+<p>Add paddle tablemaster table recognition option</p>
+</details>
+
+<details>
+<summary>2024/08/09 Version 0.7.0b1 released</summary>
+<p>Simplified installation process, added table recognition functionality</p>
+</details>
+
+<details>
+<summary>2024/08/01 Version 0.6.2b1 released</summary>
+<p>Optimized dependency conflict issues and installation documentation</p>
+</details>
+
+<details>
+<summary>2024/07/05 Initial open-source release</summary>
+</details>
+
+<!-- TABLE OF CONTENT -->
+
+<details open="open">
+  <summary><h2 style="display: inline-block">Table of Contents</h2></summary>
+  <ol>
+    <li>
+      <a href="#mineru">MinerU</a>
+      <ul>
+        <li><a href="#project-introduction">Project Introduction</a></li>
+        <li><a href="#key-features">Key Features</a></li>
+        <li><a href="#quick-start">Quick Start</a>
+            <ul>
+            <li><a href="#online-demo">Online Demo</a></li>
+            <li><a href="#quick-cpu-demo">Quick CPU Demo</a></li>
+            <li><a href="#using-gpu">Using GPU</a></li>
+            <li><a href="#using-npu">Using NPU</a></li>
+            </ul>
+        </li>
+        <li><a href="#usage">Usage</a>
+            <ul>
+            <li><a href="#command-line">Command Line</a></li>
+            <li><a href="#api">API</a></li>
+            <li><a href="#deploy-derived-projects">Deploy Derived Projects</a></li>
+            <li><a href="#development-guide">Development Guide</a></li>
+            </ul>
+        </li>
+      </ul>
+    </li>
+    <li><a href="#todo">TODO</a></li>
+    <li><a href="#known-issues">Known Issues</a></li>
+    <li><a href="#faq">FAQ</a></li>
+    <li><a href="#all-thanks-to-our-contributors">All Thanks To Our Contributors</a></li>
+    <li><a href="#license-information">License Information</a></li>
+    <li><a href="#acknowledgments">Acknowledgments</a></li>
+    <li><a href="#citation">Citation</a></li>
+    <li><a href="#star-history">Star History</a></li>
+    <li><a href="#magic-doc">Magic-doc</a></li>
+    <li><a href="#magic-html">Magic-html</a></li>
+    <li><a href="#links">Links</a></li>
+  </ol>
+</details>
+
+# MinerU
+
+## Project Introduction
+
+MinerU is a tool that converts PDFs into machine-readable formats (e.g., markdown, JSON), allowing for easy extraction into any format.
+MinerU was born during the pre-training process of [InternLM](https://github.com/InternLM/InternLM). We focus on solving symbol conversion issues in scientific literature and hope to contribute to technological development in the era of large models.
+Compared to well-known commercial products, MinerU is still young. If you encounter any issues or if the results are not as expected, please submit an issue on [issue](https://github.com/opendatalab/MinerU/issues) and **attach the relevant PDF**.
+
+https://github.com/user-attachments/assets/4bea02c9-6d54-4cd6-97ed-dff14340982c
+
+## Key Features
+
+- Remove headers, footers, footnotes, page numbers, etc., to ensure semantic coherence.
+- Output text in human-readable order, suitable for single-column, multi-column, and complex layouts.
+- Preserve the structure of the original document, including headings, paragraphs, lists, etc.
+- Extract images, image descriptions, tables, table titles, and footnotes.
+- Automatically recognize and convert formulas in the document to LaTeX format.
+- Automatically recognize and convert tables in the document to HTML format.
+- Automatically detect scanned PDFs and garbled PDFs and enable OCR functionality.
+- OCR supports detection and recognition of 84 languages.
+- Supports multiple output formats, such as multimodal and NLP Markdown, JSON sorted by reading order, and rich intermediate formats.
+- Supports various visualization results, including layout visualization and span visualization, for efficient confirmation of output quality.
+- Supports running in a pure CPU environment, and also supports GPU(CUDA)/NPU(CANN)/MPS acceleration
+- Compatible with Windows, Linux, and Mac platforms.
+
+## Quick Start
+
+If you encounter any installation issues, please first consult the <a href="#faq">FAQ</a>. </br>
+If the parsing results are not as expected, refer to the <a href="#known-issues">Known Issues</a>. </br>
+There are three different ways to experience MinerU:
+
+- [Online Demo (No Installation Required)](#online-demo)
+- [Quick CPU Demo (Windows, Linux, Mac)](#quick-cpu-demo)
+- Accelerate inference by using CUDA/CANN/MPS
+  - [Linux/Windows + CUDA](#Using-GPU)
+  - [Linux + CANN](#using-npu)
+  - [MacOS + MPS](#using-mps)
+
+> [!WARNING]
+> **Pre-installation Notice—Hardware and Software Environment Support**
+>
+> To ensure the stability and reliability of the project, we only optimize and test for specific hardware and software environments during development. This ensures that users deploying and running the project on recommended system configurations will get the best performance with the fewest compatibility issues.
+>
+> By focusing resources on the mainline environment, our team can more efficiently resolve potential bugs and develop new features.
+>
+> In non-mainline environments, due to the diversity of hardware and software configurations, as well as third-party dependency compatibility issues, we cannot guarantee 100% project availability. Therefore, for users who wish to use this project in non-recommended environments, we suggest carefully reading the documentation and FAQ first. Most issues already have corresponding solutions in the FAQ. We also encourage community feedback to help us gradually expand support.
+
+<table>
+    <tr>
+        <td colspan="3" rowspan="2">Operating System</td>
+    </tr>
+    <tr>
+        <td>Linux after 2019</td>
+        <td>Windows 10 / 11</td>
+        <td>macOS 11+</td>
+    </tr>
+    <tr>
+        <td colspan="3">CPU</td>
+        <td>x86_64 / arm64</td>
+        <td>x86_64(unsupported ARM Windows)</td>
+        <td>x86_64 / arm64</td>
+    </tr>
+    <tr>
+        <td colspan="3">Memory Requirements</td>
+        <td colspan="3">16GB or more, recommended 32GB+</td>
+    </tr>
+    <tr>
+        <td colspan="3">Storage Requirements</td>
+        <td colspan="3">20GB or more, with a preference for SSD</td>
+    </tr>
+    <tr>
+        <td colspan="3">Python Version</td>
+        <td colspan="3">3.10~3.13</td>
+    </tr>
+    <tr>
+        <td colspan="3">Nvidia Driver Version</td>
+        <td>latest (Proprietary Driver)</td>
+        <td>latest</td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td colspan="3">CUDA Environment</td>
+        <td colspan="2"><a href="https://pytorch.org/get-started/locally/">Refer to the PyTorch official website</a></td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td colspan="3">CANN Environment(NPU support)</td>
+        <td>8.0+(Ascend 910b)</td>
+        <td>None</td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td rowspan="2">GPU/MPS Hardware Support List</td>
+        <td colspan="2">GPU VRAM 6GB or more</td>
+        <td colspan="2">All GPUs with Tensor Cores produced from Volta(2017) onwards.<br>
+        More than 6GB VRAM </td>
+        <td rowspan="2">Apple silicon</td>
+    </tr>
+</table>
+
+### Online Demo
+
+Synced with dev branch updates:
+
+[![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://mineru.net/OpenSourceTools/Extractor?source=github)
+[![HuggingFace](https://img.shields.io/badge/Demo_on_HuggingFace-yellow.svg?logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAF8AAABYCAMAAACkl9t/AAAAk1BMVEVHcEz/nQv/nQv/nQr/nQv/nQr/nQv/nQv/nQr/wRf/txT/pg7/yRr/rBD/zRz/ngv/oAz/zhz/nwv/txT/ngv/0B3+zBz/nQv/0h7/wxn/vRb/thXkuiT/rxH/pxD/ogzcqyf/nQvTlSz/czCxky7/SjifdjT/Mj3+Mj3wMj15aTnDNz+DSD9RTUBsP0FRO0Q6O0WyIxEIAAAAGHRSTlMADB8zSWF3krDDw8TJ1NbX5efv8ff9/fxKDJ9uAAAGKklEQVR42u2Z63qjOAyGC4RwCOfB2JAGqrSb2WnTw/1f3UaWcSGYNKTdf/P+mOkTrE+yJBulvfvLT2A5ruenaVHyIks33npl/6C4s/ZLAM45SOi/1FtZPyFur1OYofBX3w7d54Bxm+E8db+nDr12ttmESZ4zludJEG5S7TO72YPlKZFyE+YCYUJTBZsMiNS5Sd7NlDmKM2Eg2JQg8awbglfqgbhArjxkS7dgp2RH6hc9AMLdZYUtZN5DJr4molC8BfKrEkPKEnEVjLbgW1fLy77ZVOJagoIcLIl+IxaQZGjiX597HopF5CkaXVMDO9Pyix3AFV3kw4lQLCbHuMovz8FallbcQIJ5Ta0vks9RnolbCK84BtjKRS5uA43hYoZcOBGIG2Epbv6CvFVQ8m8loh66WNySsnN7htL58LNp+NXT8/PhXiBXPMjLSxtwp8W9f/1AngRierBkA+kk/IpUSOeKByzn8y3kAAAfh//0oXgV4roHm/kz4E2z//zRc3/lgwBzbM2mJxQEa5pqgX7d1L0htrhx7LKxOZlKbwcAWyEOWqYSI8YPtgDQVjpB5nvaHaSnBaQSD6hweDi8PosxD6/PT09YY3xQA7LTCTKfYX+QHpA0GCcqmEHvr/cyfKQTEuwgbs2kPxJEB0iNjfJcCTPyocx+A0griHSmADiC91oNGVwJ69RudYe65vJmoqfpul0lrqXadW0jFKH5BKwAeCq+Den7s+3zfRJzA61/Uj/9H/VzLKTx9jFPPdXeeP+L7WEvDLAKAIoF8bPTKT0+TM7W8ePj3Rz/Yn3kOAp2f1Kf0Weony7pn/cPydvhQYV+eFOfmOu7VB/ViPe34/EN3RFHY/yRuT8ddCtMPH/McBAT5s+vRde/gf2c/sPsjLK+m5IBQF5tO+h2tTlBGnP6693JdsvofjOPnnEHkh2TnV/X1fBl9S5zrwuwF8NFrAVJVwCAPTe8gaJlomqlp0pv4Pjn98tJ/t/fL++6unpR1YGC2n/KCoa0tTLoKiEeUPDl94nj+5/Tv3/eT5vBQ60X1S0oZr+IWRR8Ldhu7AlLjPISlJcO9vrFotky9SpzDequlwEir5beYAc0R7D9KS1DXva0jhYRDXoExPdc6yw5GShkZXe9QdO/uOvHofxjrV/TNS6iMJS+4TcSTgk9n5agJdBQbB//IfF/HpvPt3Tbi7b6I6K0R72p6ajryEJrENW2bbeVUGjfgoals4L443c7BEE4mJO2SpbRngxQrAKRudRzGQ8jVOL2qDVjjI8K1gc3TIJ5KiFZ1q+gdsARPB4NQS4AjwVSt72DSoXNyOWUrU5mQ9nRYyjp89Xo7oRI6Bga9QNT1mQ/ptaJq5T/7WcgAZywR/XlPGAUDdet3LE+qS0TI+g+aJU8MIqjo0Kx8Ly+maxLjJmjQ18rA0YCkxLQbUZP1WqdmyQGJLUm7VnQFqodmXSqmRrdVpqdzk5LvmvgtEcW8PMGdaS23EOWyDVbACZzUJPaqMbjDxpA3Qrgl0AikimGDbqmyT8P8NOYiqrldF8rX+YN7TopX4UoHuSCYY7cgX4gHwclQKl1zhx0THf+tCAUValzjI7Wg9EhptrkIcfIJjA94evOn8B2eHaVzvBrnl2ig0So6hvPaz0IGcOvTHvUIlE2+prqAxLSQxZlU2stql1NqCCLdIiIN/i1DBEHUoElM9dBravbiAnKqgpi4IBkw+utSPIoBijDXJipSVV7MpOEJUAc5Qmm3BnUN+w3hteEieYKfRZSIUcXKMVf0u5wD4EwsUNVvZOtUT7A2GkffHjByWpHqvRBYrTV72a6j8zZ6W0DTE86Hn04bmyWX3Ri9WH7ZU6Q7h+ZHo0nHUAcsQvVhXRDZHChwiyi/hnPuOsSEF6Exk3o6Y9DT1eZ+6cASXk2Y9k+6EOQMDGm6WBK10wOQJCBwren86cPPWUcRAnTVjGcU1LBgs9FURiX/e6479yZcLwCBmTxiawEwrOcleuu12t3tbLv/N4RLYIBhYexm7Fcn4OJcn0+zc+s8/VfPeddZHAGN6TT8eGczHdR/Gts1/MzDkThr23zqrVfAMFT33Nx1RJsx1k5zuWILLnG/vsH+Fv5D4NTVcp1Gzo8AAAAAElFTkSuQmCC&labelColor=white)](https://huggingface.co/spaces/opendatalab/MinerU)
+[![ModelScope](https://img.shields.io/badge/Demo_on_ModelScope-purple?logo=data:image/svg+xml;base64,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&labelColor=white)](https://www.modelscope.cn/studios/OpenDataLab/MinerU)
+
+### Quick CPU Demo
+
+#### 1. Install magic-pdf
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+pip install -U "magic-pdf[full]"
+```
+
+#### 2. Download model weight files
+
+Refer to [How to Download Model Files](docs/how_to_download_models_en.md) for detailed instructions.
+
+#### 3. Modify the Configuration File for Additional Configuration
+
+After completing the [2. Download model weight files](#2-download-model-weight-files) step, the script will automatically generate a `magic-pdf.json` file in the user directory and configure the default model path.
+You can find the `magic-pdf.json` file in your 【user directory】.
+
+> [!TIP]
+> The user directory for Windows is "C:\\Users\\username", for Linux it is "/home/username", and for macOS it is "/Users/username".
+
+You can modify certain configurations in this file to enable or disable features, such as table recognition:
+
+
+> [!NOTE]
+> If the following items are not present in the JSON, please manually add the required items and remove the comment content (standard JSON does not support comments).
+
+```json
+{
+    // other config
+    "layout-config": {
+        "model": "doclayout_yolo" 
+    },
+    "formula-config": {
+        "mfd_model": "yolo_v8_mfd",
+        "mfr_model": "unimernet_small",
+        "enable": true  // The formula recognition feature is enabled by default. If you need to disable it, please change the value here to "false".
+    },
+    "table-config": {
+        "model": "rapid_table", 
+        "sub_model": "slanet_plus",
+        "enable": true, // The table recognition feature is enabled by default. If you need to disable it, please change the value here to "false".
+        "max_time": 400
+    }
+}
+```
+
+### Using GPU
+
+If your device supports CUDA and meets the GPU requirements of the mainline environment, you can use GPU acceleration. Please select the appropriate guide based on your system:
+
+- [Ubuntu 22.04 LTS + GPU](docs/README_Ubuntu_CUDA_Acceleration_en_US.md)
+- [Windows 10/11 + GPU](docs/README_Windows_CUDA_Acceleration_en_US.md)
+- Quick Deployment with Docker
+> [!IMPORTANT]
+> Docker requires a GPU with at least 6GB of VRAM, and all acceleration features are enabled by default.
+>
+> Before running this Docker, you can use the following command to check if your device supports CUDA acceleration on Docker.
+> 
+> ```bash
+> docker run --rm --gpus=all nvidia/cuda:12.1.0-base-ubuntu22.04 nvidia-smi
+> ```
+  ```bash
+  wget https://github.com/opendatalab/MinerU/raw/master/docker/global/Dockerfile -O Dockerfile
+  docker build -t mineru:latest .
+  docker run -it --name mineru --gpus=all mineru:latest /bin/bash -c "echo 'source /opt/mineru_venv/bin/activate' >> ~/.bashrc && exec bash"
+  magic-pdf --help
+  ```
+
+### Using NPU
+
+If your device has NPU acceleration hardware, you can follow the tutorial below to use NPU acceleration:
+
+[Ascend NPU Acceleration](docs/README_Ascend_NPU_Acceleration_zh_CN.md)
+
+### Using MPS
+
+If your device uses Apple silicon chips, you can enable MPS acceleration for your tasks.
+
+You can enable MPS acceleration by setting the `device-mode` parameter to `mps` in the `magic-pdf.json` configuration file.
+
+```json
+{
+    // other config
+    "device-mode": "mps"
+}
+```
+
+
+## Usage
+
+### Command Line
+
+[Using MinerU via Command Line](https://mineru.readthedocs.io/en/latest/user_guide/usage/command_line.html)
+
+> [!TIP]
+> For more information about the output files, please refer to the [Output File Description](docs/output_file_en_us.md).
+
+### API
+
+[Using MinerU via Python API](https://mineru.readthedocs.io/en/latest/user_guide/usage/api.html)
+
+
+### Deploy Derived Projects
+
+Derived projects include secondary development projects based on MinerU by project developers and community developers,  
+such as application interfaces based on Gradio, RAG based on llama, web demos similar to the official website, lightweight multi-GPU load balancing client/server ends, etc.
+These projects may offer more features and a better user experience.  
+For specific deployment methods, please refer to the [Derived Project README](projects/README.md)
+
+
+### Development Guide
+
+TODO
+
+# TODO
+
+- [x] Reading order based on the model  
+- [x] Recognition of `index` and `list` in the main text  
+- [x] Table recognition
+- [x] Heading Classification
+- [ ] Code block recognition in the main text
+- [ ] [Chemical formula recognition](docs/chemical_knowledge_introduction/introduction.pdf)
+- [ ] Geometric shape recognition
+
+# Known Issues
+
+- Reading order is determined by the model based on the spatial distribution of readable content, and may be out of order in some areas under extremely complex layouts.
+- Vertical text is not supported.
+- Tables of contents and lists are recognized through rules, and some uncommon list formats may not be recognized.
+- Code blocks are not yet supported in the layout model.
+- Comic books, art albums, primary school textbooks, and exercises cannot be parsed well.
+- Table recognition may result in row/column recognition errors in complex tables.
+- OCR recognition may produce inaccurate characters in PDFs of lesser-known languages (e.g., diacritical marks in Latin script, easily confused characters in Arabic script).
+- Some formulas may not render correctly in Markdown.
+
+# FAQ
+
+[FAQ in Chinese](docs/FAQ_zh_cn.md)
+
+[FAQ in English](docs/FAQ_en_us.md)
+
+# All Thanks To Our Contributors
+
+<a href="https://github.com/opendatalab/MinerU/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=opendatalab/MinerU" />
+</a>
+
+# License Information
+
+[LICENSE.md](LICENSE.md)
+
+This project currently uses PyMuPDF to achieve advanced functionality. However, since it adheres to the AGPL license, it may impose restrictions on certain usage scenarios. In future iterations, we plan to explore and replace it with a more permissive PDF processing library to enhance user-friendliness and flexibility.
+
+# Acknowledgments
+
+- [PDF-Extract-Kit](https://github.com/opendatalab/PDF-Extract-Kit)
+- [DocLayout-YOLO](https://github.com/opendatalab/DocLayout-YOLO)
+- [StructEqTable](https://github.com/UniModal4Reasoning/StructEqTable-Deploy)
+- [RapidTable](https://github.com/RapidAI/RapidTable)
+- [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)
+- [RapidOCR](https://github.com/RapidAI/RapidOCR)
+- [PaddleOCR2Pytorch](https://github.com/frotms/PaddleOCR2Pytorch)
+- [PyMuPDF](https://github.com/pymupdf/PyMuPDF)
+- [layoutreader](https://github.com/ppaanngggg/layoutreader)
+- [fast-langdetect](https://github.com/LlmKira/fast-langdetect)
+- [pdfminer.six](https://github.com/pdfminer/pdfminer.six)
+
+# Citation
+
+```bibtex
+@misc{wang2024mineruopensourcesolutionprecise,
+      title={MinerU: An Open-Source Solution for Precise Document Content Extraction}, 
+      author={Bin Wang and Chao Xu and Xiaomeng Zhao and Linke Ouyang and Fan Wu and Zhiyuan Zhao and Rui Xu and Kaiwen Liu and Yuan Qu and Fukai Shang and Bo Zhang and Liqun Wei and Zhihao Sui and Wei Li and Botian Shi and Yu Qiao and Dahua Lin and Conghui He},
+      year={2024},
+      eprint={2409.18839},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV},
+      url={https://arxiv.org/abs/2409.18839}, 
+}
+
+@article{he2024opendatalab,
+  title={Opendatalab: Empowering general artificial intelligence with open datasets},
+  author={He, Conghui and Li, Wei and Jin, Zhenjiang and Xu, Chao and Wang, Bin and Lin, Dahua},
+  journal={arXiv preprint arXiv:2407.13773},
+  year={2024}
+}
+```
+
+# Star History
+
+<a>
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date" />
+   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date" />
+ </picture>
+</a>
+
+# Magic-doc
+
+[Magic-Doc](https://github.com/InternLM/magic-doc) Fast speed ppt/pptx/doc/docx/pdf extraction tool
+
+# Magic-html
+
+[Magic-HTML](https://github.com/opendatalab/magic-html) Mixed web page extraction tool
+
+# Links
+
+- [LabelU (A Lightweight Multi-modal Data Annotation Tool)](https://github.com/opendatalab/labelU)
+- [LabelLLM (An Open-source LLM Dialogue Annotation Platform)](https://github.com/opendatalab/LabelLLM)
+- [PDF-Extract-Kit (A Comprehensive Toolkit for High-Quality PDF Content Extraction)](https://github.com/opendatalab/PDF-Extract-Kit)

third_party/MinerU/README_zh-CN.md

@@ -0,0 +1,596 @@
+<div align="center" xmlns="http://www.w3.org/1999/html">
+<!-- logo -->
+<p align="center">
+  <img src="docs/images/MinerU-logo.png" width="300px" style="vertical-align:middle;">
+</p>
+
+<!-- icon -->
+
+[![stars](https://img.shields.io/github/stars/opendatalab/MinerU.svg)](https://github.com/opendatalab/MinerU)
+[![forks](https://img.shields.io/github/forks/opendatalab/MinerU.svg)](https://github.com/opendatalab/MinerU)
+[![open issues](https://img.shields.io/github/issues-raw/opendatalab/MinerU)](https://github.com/opendatalab/MinerU/issues)
+[![issue resolution](https://img.shields.io/github/issues-closed-raw/opendatalab/MinerU)](https://github.com/opendatalab/MinerU/issues)
+[![PyPI version](https://img.shields.io/pypi/v/magic-pdf)](https://pypi.org/project/magic-pdf/)
+[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/magic-pdf)](https://pypi.org/project/magic-pdf/)
+[![Downloads](https://static.pepy.tech/badge/magic-pdf)](https://pepy.tech/project/magic-pdf)
+[![Downloads](https://static.pepy.tech/badge/magic-pdf/month)](https://pepy.tech/project/magic-pdf)
+
+[![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://mineru.net/OpenSourceTools/Extractor?source=github)
+[![ModelScope](https://img.shields.io/badge/Demo_on_ModelScope-purple?logo=data:image/svg+xml;base64,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&labelColor=white)](https://www.modelscope.cn/studios/OpenDataLab/MinerU)
+[![HuggingFace](https://img.shields.io/badge/Demo_on_HuggingFace-yellow.svg?logo=data:image/png;base64,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&labelColor=white)](https://huggingface.co/spaces/opendatalab/MinerU)
+[![Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/gist/myhloli/3b3a00a4a0a61577b6c30f989092d20d/mineru_demo.ipynb)
+[![Paper](https://img.shields.io/badge/Paper-arXiv-green)](https://arxiv.org/abs/2409.18839)
+
+
+<a href="https://trendshift.io/repositories/11174" target="_blank"><img src="https://trendshift.io/api/badge/repositories/11174" alt="opendatalab%2FMinerU | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+
+<!-- language -->
+
+[English](README.md) | [简体中文](README_zh-CN.md)
+
+<!-- hot link -->
+
+<p align="center">
+<a href="https://github.com/opendatalab/PDF-Extract-Kit">PDF-Extract-Kit: 高质量PDF解析工具箱</a>🔥🔥🔥
+<br>
+<br>
+<a href="https://mineru.net/client?source=github">更便捷的使用方式：MinerU桌面端。无需编程，无需登录，图形界面，简单交互，畅用无忧。</a>🚀🚀🚀
+
+</p>
+
+<!-- join us -->
+
+<p align="center">
+    👋 join us on <a href="https://discord.gg/Tdedn9GTXq" target="_blank">Discord</a> and <a href="http://mineru.space/s/V85Yl" target="_blank">WeChat</a>
+</p>
+
+</div>
+
+# 更新记录
+- 2025/04/29 1.3.10 发布
+  - 支持使用自定义公式标识符，可通过修改用户目录下的`magic-pdf.json`文件中的`latex-delimiter-config`项实现。
+- 2025/04/27 1.3.9 发布
+  - 优化公式解析功能，提升公式渲染的成功率
+- 2025/04/23 1.3.8 发布
+  - `ocr`默认模型(`ch`)更新为`PP-OCRv4_server_rec_doc`（需更新模型）
+    - `PP-OCRv4_server_rec_doc`是在`PP-OCRv4_server_rec`的基础上，在更多中文文档数据和PP-OCR训练数据的混合数据训练而成，增加了部分繁体字、日文、特殊字符的识别能力，可支持识别的字符为1.5万+，除文档相关的文字识别能力提升外，也同时提升了通用文字的识别能力。
+    - [PP-OCRv4_server_rec_doc/PP-OCRv4_server_rec/PP-OCRv4_mobile_rec 性能对比](https://paddlepaddle.github.io/PaddleX/latest/module_usage/tutorials/ocr_modules/text_recognition.html#_3)
+    - 经验证，`PP-OCRv4_server_rec_doc`模型在`中英日繁`单种语言或多种语言混合场景均有明显精度提升，且速度与`PP-OCRv4_server_rec`相当，适合绝大部分场景使用。
+    - `PP-OCRv4_server_rec_doc`在小部分纯英文场景可能会发生单词粘连问题，`PP-OCRv4_server_rec`则在此场景下表现更好，因此我们保留了`PP-OCRv4_server_rec`模型，用户可通过增加参数`lang='ch_server'`(python api)或`--lang ch_server`(命令行)调用。
+- 2025/04/22 1.3.7 发布
+  - 修复表格解析模型初始化时lang参数失效的问题
+  - 修复在`cpu`模式下ocr和表格解析速度大幅下降的问题
+- 2025/04/16 1.3.4 发布
+  - 通过移除一些无用的块，小幅提升了ocr-det的速度
+  - 修复部分情况下由footnote导致的页面内排序错误
+- 2025/04/12 1.3.2 发布
+  - 修复了windows系统下，在python3.13环境安装时一些依赖包版本不兼容的问题
+  - 优化批量推理时的内存占用
+  - 优化旋转90度表格的解析效果
+  - 优化财报样本中超大表格的解析效果
+  - 修复了在未指定OCR语言时，英文文本区域偶尔出现的单词黏连问题（需要更新模型）
+- 2025/04/08 1.3.1 发布，修复了一些兼容问题
+  - 支持python 3.13
+  - 为部分过时的linux系统（如centos7）做出最后适配，并不再保证后续版本的继续支持，[安装说明](https://github.com/opendatalab/MinerU/issues/1004)
+- 2025/04/03 1.3.0 发布，在这个版本我们做出了许多优化和改进：
+  - 安装与兼容性优化
+    - 通过移除layout中`layoutlmv3`的使用，解决了由`detectron2`导致的兼容问题
+    - torch版本兼容扩展到2.2~2.6(2.5除外)
+    - cuda兼容支持11.8/12.4/12.6/12.8（cuda版本由torch决定），解决部分用户50系显卡与H系显卡的兼容问题
+    - python兼容版本扩展到3.10~3.12，解决了在非3.10环境下安装时自动降级到0.6.1的问题
+    - 优化离线部署流程，部署成功后不需要联网下载任何模型文件
+  - 性能优化
+    - 通过支持多个pdf文件的batch处理（[脚本样例](demo/batch_demo.py)），提升了批量小文件的解析速度 (与1.0.1版本相比，公式解析速度最高提升超过1400%，整体解析速度最高提升超过500%)
+    - 通过优化mfr模型的加载和使用，降低了显存占用并提升了解析速度(需重新执行[模型下载流程](docs/how_to_download_models_zh_cn.md)以获得模型文件的增量更新)
+    - 优化显存占用，最低仅需6GB即可运行本项目
+    - 优化了在mps设备上的运行速度
+  - 解析效果优化
+    - mfr模型更新到`unimernet(2503)`，解决多行公式中换行丢失的问题
+  - 易用性优化
+    - 通过使用`paddleocr2torch`，完全替代`paddle`框架以及`paddleocr`在项目中的使用，解决了`paddle`和`torch`的冲突问题，和由于`paddle`框架导致的线程不安全问题
+    - 解析过程增加实时进度条显示，精准把握解析进度，让等待不再痛苦
+<details>
+<summary>2025/03/03 1.2.1 发布，修复了一些问题</summary>
+<ul>
+    <li>修复在字母与数字的全角转半角操作时对标点符号的影响</li>
+    <li>修复在某些情况下caption的匹配不准确问题</li>
+    <li>修复在某些情况下的公式span丢失问题</li>
+</ul>
+</details>
+
+<details>
+<summary>2025/02/24 1.2.0 发布，这个版本我们修复了一些问题，提升了解析的效率与精度：</summary>
+<ul>
+    <li>性能优化
+        <ul>
+            <li>auto模式下pdf文档的分类速度提升</li>
+        </ul>
+    </li>
+    <li>解析优化
+        <ul>
+            <li>优化对包含水印文档的解析逻辑，显著提升包含水印文档的解析效果</li>
+            <li>改进了单页内多个图像/表格与caption的匹配逻辑，提升了复杂布局下图文匹配的准确性</li>
+        </ul>
+    </li>
+    <li>问题修复
+        <ul>
+            <li>修复在某些情况下图片/表格span被填充进textblock导致的异常</li>
+            <li>修复在某些情况下标题block为空的问题</li>
+        </ul>
+    </li>
+</ul>
+</details>
+
+<details>
+<summary>2025/01/22 1.1.0 发布，在这个版本我们重点提升了解析的精度与效率：</summary>
+<ul>
+    <li>模型能力升级（需重新执行 <a href="https://github.com/opendatalab/MinerU/docs/how_to_download_models_zh_cn.md">模型下载流程</a> 以获得模型文件的增量更新）
+        <ul>
+            <li>布局识别模型升级到最新的 `doclayout_yolo(2501)` 模型，提升了layout识别精度</li>
+            <li>公式解析模型升级到最新的 `unimernet(2501)` 模型，提升了公式识别精度</li>
+        </ul>
+    </li>
+    <li>性能优化
+        <ul>
+            <li>在配置满足一定条件（显存16GB+）的设备上，通过优化资源占用和重构处理流水线，整体解析速度提升50%以上</li>
+        </ul>
+    </li>
+    <li>解析效果优化
+        <ul>
+            <li>在线demo（<a href="https://mineru.net/OpenSourceTools/Extractor">mineru.net</a> / <a href="https://huggingface.co/spaces/opendatalab/MinerU">huggingface</a> / <a href="https://www.modelscope.cn/studios/OpenDataLab/MinerU">modelscope</a>）上新增标题分级功能（测试版本，默认开启），支持对标题进行分级，提升文档结构化程度</li>
+        </ul>
+    </li>
+</ul>
+</details>
+
+<details>
+<summary>2025/01/10 1.0.1 发布，这是我们的第一个正式版本，在这个版本中，我们通过大量重构带来了全新的API接口和更广泛的兼容性，以及全新的自动语言识别功能：</summary>
+<ul>
+    <li>全新API接口
+        <ul>
+            <li>对于数据侧API，我们引入了Dataset类，旨在提供一个强大而灵活的数据处理框架。该框架当前支持包括图像（.jpg及.png）、PDF、Word（.doc及.docx）、以及PowerPoint（.ppt及.pptx）在内的多种文档格式，确保了从简单到复杂的数据处理任务都能得到有效的支持。</li>
+            <li>针对用户侧API，我们将MinerU的处理流程精心设计为一系列可组合的Stage阶段。每个Stage代表了一个特定的处理步骤，用户可以根据自身需求自由地定义新的Stage，并通过创造性地组合这些阶段来定制专属的数据处理流程。</li>
+        </ul>
+    </li>
+    <li>更广泛的兼容性适配
+        <ul>
+            <li>通过优化依赖环境和配置项，确保在ARM架构的Linux系统上能够稳定高效运行。</li>
+            <li>深度适配华为昇腾NPU加速，积极响应信创要求，提供自主可控的高性能计算能力，助力人工智能应用平台的国产化应用与发展。 <a href="https://github.com/opendatalab/MinerU/docs/README_Ascend_NPU_Acceleration_zh_CN.md">NPU加速教程</a></li>
+        </ul>
+    </li>
+    <li>自动语言识别
+        <ul>
+            <li>通过引入全新的语言识别模型， 在文档解析中将 `lang` 配置为 `auto`，即可自动选择合适的OCR语言模型，提升扫描类文档解析的准确性。</li>
+        </ul>
+    </li>
+</ul>
+</details>
+
+<details>
+<summary>2024/11/22 0.10.0发布，通过引入混合OCR文本提取能力，</summary>
+<ul>
+    <li>在公式密集、span区域不规范、部分文本使用图像表现等复杂文本分布场景下获得解析效果的显著提升</li>
+    <li>同时具备文本模式内容提取准确、速度更快与OCR模式span/line区域识别更准的双重优势</li>
+</ul>
+</details>
+
+<details>
+<summary>2024/11/15 0.9.3发布，为表格识别功能接入了<a href="https://github.com/RapidAI/RapidTable">RapidTable</a>,单表解析速度提升10倍以上，准确率更高，显存占用更低</summary>
+</details>
+
+<details>
+<summary>2024/11/06 0.9.2发布，为表格识别功能接入了<a href="https://huggingface.co/U4R/StructTable-InternVL2-1B">StructTable-InternVL2-1B</a>模型</summary>
+</details>
+
+<details>
+<summary>2024/10/31 0.9.0发布，这是我们进行了大量代码重构的全新版本，解决了众多问题，提升了性能，降低了硬件需求，并提供了更丰富的易用性：</summary>
+<ul>
+    <li>重构排序模块代码，使用 <a href="https://github.com/ppaanngggg/layoutreader">layoutreader</a> 进行阅读顺序排序，确保在各种排版下都能实现极高准确率</li>
+    <li>重构段落拼接模块，在跨栏、跨页、跨图、跨表情况下均能实现良好的段落拼接效果</li>
+    <li>重构列表和目录识别功能，极大提升列表块和目录块识别的准确率及对应文本段落的解析效果</li>
+    <li>重构图、表与描述性文本的匹配逻辑，大幅提升 caption 和 footnote 与图表的匹配准确率，并将描述性文本的丢失率降至接近0</li>
+    <li>增加 OCR 的多语言支持，支持 84 种语言的检测与识别，语言支持列表详见 <a href="https://paddlepaddle.github.io/PaddleOCR/latest/ppocr/blog/multi_languages.html#5">OCR 语言支持列表</a></li>
+    <li>增加显存回收逻辑及其他显存优化措施，大幅降低显存使用需求。开启除表格加速外的全部加速功能(layout/公式/OCR)的显存需求从16GB降至8GB，开启全部加速功能的显存需求从24GB降至10GB</li>
+    <li>优化配置文件的功能开关，增加独立的公式检测开关，无需公式检测时可大幅提升速度和解析效果</li>
+    <li>集成 <a href="https://github.com/opendatalab/PDF-Extract-Kit">PDF-Extract-Kit 1.0</a>
+        <ul>
+            <li>加入自研的 `doclayout_yolo` 模型，在相近解析效果情况下比原方案提速10倍以上，可通过配置文件与 `layoutlmv3` 自由切换</li>
+            <li>公式解析升级至 `unimernet 0.2.1`，在提升公式解析准确率的同时，大幅降低显存需求</li>
+            <li>因 `PDF-Extract-Kit 1.0` 更换仓库，需要重新下载模型，步骤详见 <a href="https://github.com/opendatalab/MinerU/docs/how_to_download_models_zh_cn.md">如何下载模型</a></li>
+        </ul>
+    </li>
+</ul>
+</details>
+
+<details>
+<summary>2024/09/27 0.8.1发布，修复了一些bug，同时提供了<a href="https://opendatalab.com/OpenSourceTools/Extractor/PDF/">在线demo</a>的<a href="https://github.com/opendatalab/MinerU/projects/web_demo/README_zh-CN.md">本地化部署版本</a>和<a href="https://github.com/opendatalab/MinerU/projects/web/README_zh-CN.md">前端界面</a></summary>
+</details>
+
+<details>
+<summary>2024/09/09 0.8.0发布，支持Dockerfile快速部署，同时上线了huggingface、modelscope demo</summary>
+</details>
+
+<details>
+<summary>2024/08/30 0.7.1发布，集成了paddle tablemaster表格识别功能</summary>
+</details>
+
+<details>
+<summary>2024/08/09 0.7.0b1发布，简化安装步骤提升易用性，加入表格识别功能</summary>
+</details>
+
+<details>
+<summary>2024/08/01 0.6.2b1发布，优化了依赖冲突问题和安装文档</summary>
+</details>
+
+<details>
+<summary>2024/07/05 首次开源</summary>
+</details>
+
+
+<!-- TABLE OF CONTENT -->
+
+<details open="open">
+  <summary><h2 style="display: inline-block">文档目录</h2></summary>
+  <ol>
+    <li>
+      <a href="#mineru">MinerU</a>
+      <ul>
+        <li><a href="#项目简介">项目简介</a></li>
+        <li><a href="#主要功能">主要功能</a></li>
+        <li><a href="#快速开始">快速开始</a>
+            <ul>
+            <li><a href="#在线体验">在线体验</a></li>
+            <li><a href="#使用CPU快速体验">使用CPU快速体验</a></li>
+            <li><a href="#使用GPU">使用GPU</a></li>
+            <li><a href="#使用NPU">使用NPU</a></li>
+            </ul>
+        </li>
+        <li><a href="#使用">使用方式</a>
+            <ul>
+            <li><a href="#命令行">命令行</a></li>
+            <li><a href="#api">API</a></li>
+            <li><a href="#部署衍生项目">部署衍生项目</a></li>
+            <li><a href="#二次开发">二次开发</a></li>
+            </ul>
+        </li>
+      </ul>
+    </li>
+    <li><a href="#todo">TODO</a></li>
+    <li><a href="#known-issues">Known Issues</a></li>
+    <li><a href="#faq">FAQ</a></li>
+    <li><a href="#all-thanks-to-our-contributors">Contributors</a></li>
+    <li><a href="#license-information">License Information</a></li>
+    <li><a href="#acknowledgments">Acknowledgements</a></li>
+    <li><a href="#citation">Citation</a></li>
+    <li><a href="#star-history">Star History</a></li>
+    <li><a href="#magic-doc">magic-doc快速提取PPT/DOC/PDF</a></li>
+    <li><a href="#magic-html">magic-html提取混合网页内容</a></li>
+    <li><a href="#links">Links</a></li>
+  </ol>
+</details>
+
+# MinerU
+
+## 项目简介
+
+MinerU是一款将PDF转化为机器可读格式的工具（如markdown、json），可以很方便地抽取为任意格式。
+MinerU诞生于[书生-浦语](https://github.com/InternLM/InternLM)的预训练过程中，我们将会集中精力解决科技文献中的符号转化问题，希望在大模型时代为科技发展做出贡献。
+相比国内外知名商用产品MinerU还很年轻，如果遇到问题或者结果不及预期请到[issue](https://github.com/opendatalab/MinerU/issues)提交问题，同时**附上相关PDF**。
+
+https://github.com/user-attachments/assets/4bea02c9-6d54-4cd6-97ed-dff14340982c
+
+## 主要功能
+
+- 删除页眉、页脚、脚注、页码等元素，确保语义连贯
+- 输出符合人类阅读顺序的文本，适用于单栏、多栏及复杂排版
+- 保留原文档的结构，包括标题、段落、列表等
+- 提取图像、图片描述、表格、表格标题及脚注
+- 自动识别并转换文档中的公式为LaTeX格式
+- 自动识别并转换文档中的表格为HTML格式
+- 自动检测扫描版PDF和乱码PDF，并启用OCR功能
+- OCR支持84种语言的检测与识别
+- 支持多种输出格式，如多模态与NLP的Markdown、按阅读顺序排序的JSON、含有丰富信息的中间格式等
+- 支持多种可视化结果，包括layout可视化、span可视化等，便于高效确认输出效果与质检
+- 支持纯CPU环境运行，并支持 GPU(CUDA)/NPU(CANN)/MPS 加速
+- 兼容Windows、Linux和Mac平台
+
+## 快速开始
+
+如果遇到任何安装问题，请先查询 <a href="#faq">FAQ</a> </br>
+如果遇到解析效果不及预期，参考 <a href="#known-issues">Known Issues</a></br>
+有3种不同方式可以体验MinerU的效果：
+
+- [在线体验(无需任何安装)](#在线体验)
+- [使用CPU快速体验（Windows，Linux，Mac）](#使用cpu快速体验)
+- 使用 CUDA/CANN/MPS 加速推理 
+  - [Linux/Windows + CUDA](#使用gpu)
+  - [Linux + CANN](#使用npu)
+  - [MacOS + MPS](#使用mps)
+
+
+> [!WARNING]
+> **安装前必看——软硬件环境支持说明**
+> 
+> 为了确保项目的稳定性和可靠性，我们在开发过程中仅对特定的软硬件环境进行优化和测试。这样当用户在推荐的系统配置上部署和运行项目时，能够获得最佳的性能表现和最少的兼容性问题。
+>
+> 通过集中资源和精力于主线环境，我们团队能够更高效地解决潜在的BUG，及时开发新功能。
+>
+> 在非主线环境中，由于硬件、软件配置的多样性，以及第三方依赖项的兼容性问题，我们无法100%保证项目的完全可用性。因此，对于希望在非推荐环境中使用本项目的用户，我们建议先仔细阅读文档以及FAQ，大多数问题已经在FAQ中有对应的解决方案，除此之外我们鼓励社区反馈问题，以便我们能够逐步扩大支持范围。
+
+<table>
+    <tr>
+        <td colspan="3" rowspan="2">操作系统</td>
+    </tr>
+    <tr>
+        <td>Linux after 2019</td>
+        <td>Windows 10 / 11</td>
+        <td>macOS 11+</td>
+    </tr>
+    <tr>
+        <td colspan="3">CPU</td>
+        <td>x86_64 / arm64</td>
+        <td>x86_64(暂不支持ARM Windows)</td>
+        <td>x86_64 / arm64</td>
+    </tr>
+    <tr>
+        <td colspan="3">内存</td>
+        <td colspan="3">大于等于16GB，推荐32G以上</td>
+    </tr>
+    <tr>
+        <td colspan="3">存储空间</td>
+        <td colspan="3">大于等于20GB，推荐使用SSD以获得最佳性能</td>
+    </tr>
+    <tr>
+        <td colspan="3">python版本</td>
+        <td colspan="3">3.10~3.13</td>
+    </tr>
+    <tr>
+        <td colspan="3">Nvidia Driver 版本</td>
+        <td>latest(专有驱动)</td>
+        <td>latest</td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td colspan="3">CUDA环境</td>
+        <td colspan="2"><a href="https://pytorch.org/get-started/locally/">Refer to the PyTorch official website</a></td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td colspan="3">CANN环境(NPU支持)</td>
+        <td>8.0+(Ascend 910b)</td>
+        <td>None</td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td rowspan="2">GPU/MPS 硬件支持列表</td>
+        <td colspan="2">显存6G以上</td>
+        <td colspan="2">
+        Volta(2017)及之后生产的全部带Tensor Core的GPU <br>
+        6G显存及以上</td>
+        <td rowspan="2">Apple silicon</td>
+    </tr>
+</table>
+
+### 在线体验
+
+同步dev分支更新：
+
+[![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://mineru.net/OpenSourceTools/Extractor?source=github)
+[![ModelScope](https://img.shields.io/badge/Demo_on_ModelScope-purple?logo=data:image/svg+xml;base64,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&labelColor=white)](https://www.modelscope.cn/studios/OpenDataLab/MinerU)
+[![HuggingFace](https://img.shields.io/badge/Demo_on_HuggingFace-yellow.svg?logo=data:image/png;base64,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&labelColor=white)](https://huggingface.co/spaces/opendatalab/MinerU)
+
+### 使用CPU快速体验
+
+#### 1. 安装magic-pdf
+
+> [!NOTE]
+> 最新版本国内镜像源同步可能会有延迟，请耐心等待
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+pip install -U "magic-pdf[full]" -i https://mirrors.aliyun.com/pypi/simple
+```
+
+#### 2. 下载模型权重文件
+
+详细参考 [如何下载模型文件](docs/how_to_download_models_zh_cn.md)
+
+#### 3. 修改配置文件以进行额外配置
+
+完成[2. 下载模型权重文件](#2-下载模型权重文件)步骤后，脚本会自动生成用户目录下的magic-pdf.json文件，并自动配置默认模型路径。
+您可在【用户目录】下找到magic-pdf.json文件。
+
+> [!TIP]
+> windows的用户目录为 "C:\\Users\\用户名", linux用户目录为 "/home/用户名", macOS用户目录为 "/Users/用户名"
+
+您可修改该文件中的部分配置实现功能的开关，如表格识别功能：
+
+> [!NOTE]
+>如json内没有如下项目，请手动添加需要的项目，并删除注释内容（标准json不支持注释）
+
+```json
+{
+    // other config
+    "layout-config": {
+        "model": "doclayout_yolo" 
+    },
+    "formula-config": {
+        "mfd_model": "yolo_v8_mfd",
+        "mfr_model": "unimernet_small",
+        "enable": true  // 公式识别功能默认是开启的，如果需要关闭请修改此处的值为"false"
+    },
+    "table-config": {
+        "model": "rapid_table",
+        "sub_model": "slanet_plus",
+        "enable": true, // 表格识别功能默认是开启的，如果需要关闭请修改此处的值为"false"
+        "max_time": 400
+    }
+}
+```
+
+### 使用GPU
+
+如果您的设备支持CUDA，且满足主线环境中的显卡要求，则可以使用GPU加速，请根据自己的系统选择适合的教程：
+
+- [Ubuntu22.04LTS + GPU](docs/README_Ubuntu_CUDA_Acceleration_zh_CN.md)
+- [Windows10/11 + GPU](docs/README_Windows_CUDA_Acceleration_zh_CN.md)
+- 使用Docker快速部署
+> [!IMPORTANT]
+> Docker 需设备gpu显存大于等于6GB，默认开启所有加速功能
+> 
+> 运行本docker前可以通过以下命令检测自己的设备是否支持在docker上使用CUDA加速
+> 
+> ```bash
+> docker run --rm --gpus=all nvidia/cuda:12.1.0-base-ubuntu22.04 nvidia-smi
+> ```
+  ```bash
+  wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/docker/china/Dockerfile -O Dockerfile
+  docker build -t mineru:latest .
+  docker run -it --name mineru --gpus=all mineru:latest /bin/bash -c "echo 'source /opt/mineru_venv/bin/activate' >> ~/.bashrc && exec bash"
+  magic-pdf --help
+  ```
+### 使用NPU
+
+如果您的设备存在NPU加速硬件，则可以通过以下教程使用NPU加速：
+
+[NPU加速教程](docs/README_Ascend_NPU_Acceleration_zh_CN.md)
+
+### 使用MPS
+如果您的设备使用Apple silicon 芯片，您可以开启mps加速：
+
+您可以通过在 `magic-pdf.json` 配置文件中将 `device-mode` 参数设置为 `mps` 来启用 MPS 加速。
+
+```json
+{
+    // other config
+    "device-mode": "mps"
+}
+```
+
+
+
+## 使用
+
+### 命令行
+
+[通过命令行使用MinerU](https://mineru.readthedocs.io/en/latest/user_guide/usage/command_line.html)
+
+> [!TIP]
+> 更多有关输出文件的信息，请参考[输出文件说明](docs/output_file_zh_cn.md)
+
+### API
+
+[通过Python代码调用MinerU](https://mineru.readthedocs.io/en/latest/user_guide/usage/api.html)
+
+
+### 部署衍生项目
+
+衍生项目包含项目开发者和社群开发者们基于MinerU的二次开发项目，
+例如基于Gradio的应用界面、基于llama的RAG、官网同款web demo、轻量级的多卡负载均衡c/s端等，
+这些项目可能会提供更多的功能和更好的用户体验。
+具体部署方式请参考 [衍生项目readme](projects/README_zh-CN.md)
+
+
+### 二次开发
+
+TODO
+
+# TODO
+
+- [x] 基于模型的阅读顺序  
+- [x] 正文中目录、列表识别  
+- [x] 表格识别
+- [x] 标题分级
+- [ ] 正文中代码块识别
+- [ ] [化学式识别](docs/chemical_knowledge_introduction/introduction.pdf)
+- [ ] 几何图形识别
+
+# Known Issues
+
+- 阅读顺序基于模型对可阅读内容在空间中的分布进行排序，在极端复杂的排版下可能会部分区域乱序
+- 不支持竖排文字
+- 目录和列表通过规则进行识别，少部分不常见的列表形式可能无法识别
+- 代码块在layout模型里还没有支持
+- 漫画书、艺术图册、小学教材、习题尚不能很好解析
+- 表格识别在复杂表格上可能会出现行/列识别错误
+- 在小语种PDF上，OCR识别可能会出现字符不准确的情况（如拉丁文的重音符号、阿拉伯文易混淆字符等）
+- 部分公式可能会无法在markdown中渲染
+
+# FAQ
+
+[常见问题](docs/FAQ_zh_cn.md)
+
+
+[FAQ](docs/FAQ_en_us.md)
+
+# All Thanks To Our Contributors
+
+<a href="https://github.com/opendatalab/MinerU/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=opendatalab/MinerU" />
+</a>
+
+# License Information
+
+[LICENSE.md](LICENSE.md)
+
+本项目目前采用PyMuPDF以实现高级功能，但因其遵循AGPL协议，可能对某些使用场景构成限制。未来版本迭代中，我们计划探索并替换为许可条款更为宽松的PDF处理库，以提升用户友好度及灵活性。
+
+# Acknowledgments
+
+- [PDF-Extract-Kit](https://github.com/opendatalab/PDF-Extract-Kit)
+- [DocLayout-YOLO](https://github.com/opendatalab/DocLayout-YOLO)
+- [StructEqTable](https://github.com/UniModal4Reasoning/StructEqTable-Deploy)
+- [RapidTable](https://github.com/RapidAI/RapidTable)
+- [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)
+- [RapidOCR](https://github.com/RapidAI/RapidOCR)
+- [PaddleOCR2Pytorch](https://github.com/frotms/PaddleOCR2Pytorch)
+- [PyMuPDF](https://github.com/pymupdf/PyMuPDF)
+- [layoutreader](https://github.com/ppaanngggg/layoutreader)
+- [fast-langdetect](https://github.com/LlmKira/fast-langdetect)
+- [pdfminer.six](https://github.com/pdfminer/pdfminer.six)
+
+# Citation
+
+```bibtex
+@misc{wang2024mineruopensourcesolutionprecise,
+      title={MinerU: An Open-Source Solution for Precise Document Content Extraction}, 
+      author={Bin Wang and Chao Xu and Xiaomeng Zhao and Linke Ouyang and Fan Wu and Zhiyuan Zhao and Rui Xu and Kaiwen Liu and Yuan Qu and Fukai Shang and Bo Zhang and Liqun Wei and Zhihao Sui and Wei Li and Botian Shi and Yu Qiao and Dahua Lin and Conghui He},
+      year={2024},
+      eprint={2409.18839},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV},
+      url={https://arxiv.org/abs/2409.18839}, 
+}
+
+@article{he2024opendatalab,
+  title={Opendatalab: Empowering general artificial intelligence with open datasets},
+  author={He, Conghui and Li, Wei and Jin, Zhenjiang and Xu, Chao and Wang, Bin and Lin, Dahua},
+  journal={arXiv preprint arXiv:2407.13773},
+  year={2024}
+}
+```
+
+# Star History
+
+<a>
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date" />
+   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date" />
+ </picture>
+</a>
+
+# Magic-doc
+
+[Magic-Doc](https://github.com/InternLM/magic-doc) Fast speed ppt/pptx/doc/docx/pdf extraction tool
+
+# Magic-html
+
+[Magic-HTML](https://github.com/opendatalab/magic-html) Mixed web page extraction tool
+
+# Links
+
+- [LabelU (A Lightweight Multi-modal Data Annotation Tool)](https://github.com/opendatalab/labelU)
+- [LabelLLM (An Open-source LLM Dialogue Annotation Platform)](https://github.com/opendatalab/LabelLLM)
+- [PDF-Extract-Kit (A Comprehensive Toolkit for High-Quality PDF Content Extraction)](https://github.com/opendatalab/PDF-Extract-Kit)

third_party/MinerU/demo/batch_demo.py

@@ -0,0 +1,23 @@
+import os
+from pathlib import Path
+from magic_pdf.data.batch_build_dataset import batch_build_dataset
+from magic_pdf.tools.common import batch_do_parse
+
+
+def batch(pdf_dir, output_dir, method, lang):
+    os.makedirs(output_dir, exist_ok=True)
+    doc_paths = []
+    for doc_path in Path(pdf_dir).glob('*'):
+        if doc_path.suffix == '.pdf':
+            doc_paths.append(doc_path)
+
+    # build dataset with 2 workers
+    datasets = batch_build_dataset(doc_paths, 4, lang)
+
+    # os.environ["MINERU_MIN_BATCH_INFERENCE_SIZE"] = "200"  # every 200 pages will be parsed in one batch
+    batch_do_parse(output_dir, [str(doc_path.stem) for doc_path in doc_paths], datasets, method)
+
+
+if __name__ == '__main__':
+    batch("pdfs", "output", "auto", "")
+

third_party/MinerU/demo/demo.py

@@ -0,0 +1,68 @@
+# Copyright (c) Opendatalab. All rights reserved.
+import os
+
+from magic_pdf.data.data_reader_writer import FileBasedDataWriter, FileBasedDataReader
+from magic_pdf.data.dataset import PymuDocDataset
+from magic_pdf.model.doc_analyze_by_custom_model import doc_analyze
+from magic_pdf.config.enums import SupportedPdfParseMethod
+
+# args
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+pdf_file_name = os.path.join(__dir__, "pdfs", "demo1.pdf")  # replace with the real pdf path
+name_without_extension = os.path.basename(pdf_file_name).split('.')[0]
+
+# prepare env
+local_image_dir = os.path.join(__dir__, "output", name_without_extension, "images")
+local_md_dir = os.path.join(__dir__, "output", name_without_extension)
+image_dir = str(os.path.basename(local_image_dir))
+os.makedirs(local_image_dir, exist_ok=True)
+
+image_writer, md_writer = FileBasedDataWriter(local_image_dir), FileBasedDataWriter(local_md_dir)
+
+# read bytes
+reader1 = FileBasedDataReader("")
+pdf_bytes = reader1.read(pdf_file_name)  # read the pdf content
+
+# proc
+## Create Dataset Instance
+ds = PymuDocDataset(pdf_bytes)
+
+## inference
+if ds.classify() == SupportedPdfParseMethod.OCR:
+    infer_result = ds.apply(doc_analyze, ocr=True)
+
+    ## pipeline
+    pipe_result = infer_result.pipe_ocr_mode(image_writer)
+
+else:
+    infer_result = ds.apply(doc_analyze, ocr=False)
+
+    ## pipeline
+    pipe_result = infer_result.pipe_txt_mode(image_writer)
+
+### get model inference result
+model_inference_result = infer_result.get_infer_res()
+
+### draw layout result on each page
+pipe_result.draw_layout(os.path.join(local_md_dir, f"{name_without_extension}_layout.pdf"))
+
+### draw spans result on each page
+pipe_result.draw_span(os.path.join(local_md_dir, f"{name_without_extension}_spans.pdf"))
+
+### get markdown content
+md_content = pipe_result.get_markdown(image_dir)
+
+### dump markdown
+pipe_result.dump_md(md_writer, f"{name_without_extension}.md", image_dir)
+
+### get content list content
+content_list_content = pipe_result.get_content_list(image_dir)
+
+### dump content list
+pipe_result.dump_content_list(md_writer, f"{name_without_extension}_content_list.json", image_dir)
+
+### get middle json
+middle_json_content = pipe_result.get_middle_json()
+
+### dump middle json
+pipe_result.dump_middle_json(md_writer, f'{name_without_extension}_middle.json')

third_party/MinerU/docker/ascend_npu/Dockerfile

@@ -0,0 +1,51 @@
+# Use the official Ubuntu base image
+FROM swr.cn-central-221.ovaijisuan.com/mindformers/mindformers1.2_mindspore2.3:20240722
+
+USER root
+
+# Set environment variables to non-interactive to avoid prompts during installation
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Update the package list and install necessary packages
+RUN apt-get update && \
+    apt-get install -y \
+        software-properties-common && \
+    add-apt-repository -y ppa:deadsnakes/ppa && \
+    apt-get update && \
+    apt-get install -y \
+        python3.10 \
+        python3.10-venv \
+        python3.10-distutils \
+        python3.10-dev \
+        python3-pip \
+        wget \
+        git \
+        libgl1 \
+        libglib2.0-0 \
+        && rm -rf /var/lib/apt/lists/*
+
+# Set Python 3.10 as the default python3
+RUN update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.10 1
+
+# Create a virtual environment for MinerU
+RUN python3 -m venv /opt/mineru_venv
+
+# Copy the configuration file template and install magic-pdf latest
+RUN /bin/bash -c "wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/magic-pdf.template.json && \
+    cp magic-pdf.template.json /root/magic-pdf.json && \
+    source /opt/mineru_venv/bin/activate && \
+    pip3 install --upgrade pip -i https://mirrors.aliyun.com/pypi/simple && \
+    pip3 install torch==2.3.1 torchvision==0.18.1 -i https://mirrors.aliyun.com/pypi/simple && \
+    pip3 install -U magic-pdf[full] 'numpy<2' decorator attrs absl-py cloudpickle ml-dtypes tornado einops -i https://mirrors.aliyun.com/pypi/simple && \
+    wget https://gitee.com/ascend/pytorch/releases/download/v6.0.rc2-pytorch2.3.1/torch_npu-2.3.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl && \
+    pip3 install torch_npu-2.3.1-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl"
+
+# Download models and update the configuration file
+RUN /bin/bash -c "source /opt/mineru_venv/bin/activate && \
+    pip3 install modelscope -i https://mirrors.aliyun.com/pypi/simple && \
+    wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/scripts/download_models.py -O download_models.py && \
+    python3 download_models.py && \
+    sed -i 's|cpu|npu|g' /root/magic-pdf.json"
+
+# Set the entry point to activate the virtual environment and run the command line tool
+ENTRYPOINT ["/bin/bash", "-c", "source /opt/mineru_venv/bin/activate && exec \"$@\"", "--"]

third_party/MinerU/docker/china/Dockerfile

@@ -0,0 +1,54 @@
+# Use the official Ubuntu base image
+FROM ubuntu:22.04
+
+# Set environment variables to non-interactive to avoid prompts during installation
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Update the package list and install necessary packages
+RUN apt-get update && \
+    apt-get install -y \
+        software-properties-common && \
+    add-apt-repository ppa:deadsnakes/ppa && \
+    apt-get update && \
+    apt-get install -y \
+        python3.10 \
+        python3.10-venv \
+        python3.10-distutils \
+        python3-pip \
+        wget \
+        git \
+        libgl1 \
+        libreoffice \
+        fonts-noto-cjk \
+        fonts-wqy-zenhei \
+        fonts-wqy-microhei \
+        ttf-mscorefonts-installer \
+        fontconfig \
+        libglib2.0-0 \
+        libxrender1 \
+        libsm6 \
+        libxext6 \
+        poppler-utils \
+        && rm -rf /var/lib/apt/lists/*
+
+# Set Python 3.10 as the default python3
+RUN update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.10 1
+
+# Create a virtual environment for MinerU
+RUN python3 -m venv /opt/mineru_venv
+
+# Copy the configuration file template and install magic-pdf latest
+RUN /bin/bash -c "wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/magic-pdf.template.json && \
+    cp magic-pdf.template.json /root/magic-pdf.json && \
+    source /opt/mineru_venv/bin/activate && \
+    pip3 install --upgrade pip -i https://mirrors.aliyun.com/pypi/simple && \
+    pip3 install -U magic-pdf[full] -i https://mirrors.aliyun.com/pypi/simple"
+
+# Download models and update the configuration file
+RUN /bin/bash -c "pip3 install modelscope -i https://mirrors.aliyun.com/pypi/simple && \
+    wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/scripts/download_models.py -O download_models.py && \
+    python3 download_models.py && \
+    sed -i 's|cpu|cuda|g' /root/magic-pdf.json"
+
+# Set the entry point to activate the virtual environment and run the command line tool
+ENTRYPOINT ["/bin/bash", "-c", "source /opt/mineru_venv/bin/activate && exec \"$@\"", "--"]

third_party/MinerU/docker/global/Dockerfile

@@ -0,0 +1,54 @@
+# Use the official Ubuntu base image
+FROM ubuntu:22.04
+
+# Set environment variables to non-interactive to avoid prompts during installation
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Update the package list and install necessary packages
+RUN apt-get update && \
+    apt-get install -y \
+        software-properties-common && \
+    add-apt-repository ppa:deadsnakes/ppa && \
+    apt-get update && \
+    apt-get install -y \
+        python3.10 \
+        python3.10-venv \
+        python3.10-distutils \
+        python3-pip \
+        wget \
+        git \
+        libgl1 \
+        libreoffice \
+        fonts-noto-cjk \
+        fonts-wqy-zenhei \
+        fonts-wqy-microhei \
+        ttf-mscorefonts-installer \
+        fontconfig \
+        libglib2.0-0 \
+        libxrender1 \
+        libsm6 \
+        libxext6 \
+        poppler-utils \
+        && rm -rf /var/lib/apt/lists/*
+
+# Set Python 3.10 as the default python3
+RUN update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.10 1
+
+# Create a virtual environment for MinerU
+RUN python3 -m venv /opt/mineru_venv
+
+# Copy the configuration file template and install magic-pdf latest
+RUN /bin/bash -c "wget https://github.com/opendatalab/MinerU/raw/master/magic-pdf.template.json && \
+    cp magic-pdf.template.json /root/magic-pdf.json && \
+    source /opt/mineru_venv/bin/activate && \
+    pip3 install --upgrade pip && \
+    pip3 install -U magic-pdf[full]"
+
+# Download models and update the configuration file
+RUN /bin/bash -c "pip3 install huggingface_hub && \
+    wget https://github.com/opendatalab/MinerU/raw/master/scripts/download_models_hf.py -O download_models.py && \
+    python3 download_models.py && \
+    sed -i 's|cpu|cuda|g' /root/magic-pdf.json"
+
+# Set the entry point to activate the virtual environment and run the command line tool
+ENTRYPOINT ["/bin/bash", "-c", "source /opt/mineru_venv/bin/activate && exec \"$@\"", "--"]

third_party/MinerU/docs/FAQ_en_us.md

@@ -0,0 +1,95 @@
+# Frequently Asked Questions
+
+### 1. When using the command `pip install magic-pdf[full]` on newer versions of macOS, the error `zsh: no matches found: magic-pdf[full]` occurs.
+
+On macOS, the default shell has switched from Bash to Z shell, which has special handling logic for certain types of string matching. This can lead to the "no matches found" error. You can try disabling the globbing feature in the command line and then run the installation command again.
+
+```bash
+setopt no_nomatch
+pip install magic-pdf[full]
+```
+
+### 2. Encountering the error `pickle.UnpicklingError: invalid load key, 'v'.` during use
+
+This might be due to an incomplete download of the model file. You can try re-downloading the model file and then try again.
+Reference: https://github.com/opendatalab/MinerU/issues/143
+
+### 3. Where should the model files be downloaded and how should the `/models-dir` configuration be set?
+
+The path for the model files is configured in "magic-pdf.json". just like:
+
+```json
+{
+  "models-dir": "/tmp/models"
+}
+```
+
+This path is an absolute path, not a relative path. You can obtain the absolute path in the models directory using the "pwd" command.
+Reference: https://github.com/opendatalab/MinerU/issues/155#issuecomment-2230216874
+
+### 4. Encountered the error `ImportError: libGL.so.1: cannot open shared object file: No such file or directory` in Ubuntu 22.04 on WSL2
+
+The `libgl` library is missing in Ubuntu 22.04 on WSL2. You can install the `libgl` library with the following command to resolve the issue:
+
+```bash
+sudo apt-get install libgl1-mesa-glx
+```
+
+Reference: https://github.com/opendatalab/MinerU/issues/388
+
+### 5. Encountered error `ModuleNotFoundError: No module named 'fairscale'`
+
+You need to uninstall the module and reinstall it:
+
+```bash
+pip uninstall fairscale
+pip install fairscale
+```
+
+Reference: https://github.com/opendatalab/MinerU/issues/411
+
+### 6. On some newer devices like the H100, the text parsed during OCR using CUDA acceleration is garbled.
+
+The compatibility of cuda11 with new graphics cards is poor, and the CUDA version used by Paddle needs to be upgraded.
+
+```bash
+pip install paddlepaddle-gpu==3.0.0b1 -i https://www.paddlepaddle.org.cn/packages/stable/cu123/
+```
+
+Reference: https://github.com/opendatalab/MinerU/issues/558
+
+### 7. On some Linux servers, the program immediately reports an error `Illegal instruction (core dumped)`
+
+This might be because the server's CPU does not support the AVX/AVX2 instruction set, or the CPU itself supports it but has been disabled by the system administrator. You can try contacting the system administrator to remove the restriction or change to a different server.
+
+References: https://github.com/opendatalab/MinerU/issues/591 , https://github.com/opendatalab/MinerU/issues/736
+
+
+### 8. Error when installing MinerU on CentOS 7 or Ubuntu 18: `ERROR: Failed building wheel for simsimd`
+
+The new version of albumentations (1.4.21) introduces a dependency on simsimd. Since the pre-built package of simsimd for Linux requires a glibc version greater than or equal to 2.28, this causes installation issues on some Linux distributions released before 2019. You can resolve this issue by using the following command:
+```
+pip install -U magic-pdf[full,old_linux] --extra-index-url https://wheels.myhloli.com
+```
+
+Reference: https://github.com/opendatalab/MinerU/issues/1004
+
+### 9. Old Graphics Cards Such as M40 Encounter "RuntimeError: CUDA error: CUBLAS_STATUS_NOT_SUPPORTED"
+
+An error occurs during operation (cuda):
+```
+RuntimeError: CUDA error: CUBLAS_STATUS_NOT_SUPPORTED when calling cublasGemmStridedBatchedEx(handle, opa, opb, (int)m, (int)n, (int)k, (void*)&falpha, a, CUDA_R_16BF, (int)lda, stridea, b, CUDA_R_16BF, (int)ldb, strideb, (void*)&fbeta, c, CUDA_R_16BF, (int)ldc, stridec, (int)num_batches, compute_type, CUBLAS_GEMM_DEFAULT_TENSOR_OP)
+```
+Because BF16 precision is not supported on graphics cards before the Turing architecture and some graphics cards are not recognized by torch, it is necessary to manually disable BF16 precision.
+Modify the code in lines 287-290 of the "pdf_parse_union_core_v2.py" file (note that the location may vary in different versions):
+```
+if torch.cuda.is_bf16_supported():
+    supports_bfloat16 = True
+else:
+    supports_bfloat16 = False
+```
+Change it to:
+```
+supports_bfloat16 = False
+```
+Reference: https://github.com/opendatalab/MinerU/issues/1508

third_party/MinerU/docs/FAQ_zh_cn.md

@@ -0,0 +1,97 @@
+# 常见问题解答
+
+### 1.在较新版本的mac上使用命令安装pip install magic-pdf\[full\] zsh: no matches found: magic-pdf\[full\]
+
+在 macOS 上，默认的 shell 从 Bash 切换到了 Z shell，而 Z shell 对于某些类型的字符串匹配有特殊的处理逻辑，这可能导致no matches found错误。
+可以通过在命令行禁用globbing特性，再尝试运行安装命令
+
+```bash
+setopt no_nomatch
+pip install magic-pdf[full]
+```
+
+### 2.使用过程中遇到_pickle.UnpicklingError: invalid load key, 'v'.错误
+
+可能是由于模型文件未下载完整导致，可尝试重新下载模型文件后再试
+参考：https://github.com/opendatalab/MinerU/issues/143
+
+### 3.模型文件应该下载到哪里/models-dir的配置应该怎么填
+
+模型文件的路径输入是在"magic-pdf.json"中通过
+
+```json
+{
+  "models-dir": "/tmp/models"
+}
+```
+
+进行配置的。
+这个路径是绝对路径而不是相对路径，绝对路径的获取可在models目录中通过命令 "pwd" 获取。
+参考：https://github.com/opendatalab/MinerU/issues/155#issuecomment-2230216874
+
+### 4.在WSL2的Ubuntu22.04中遇到报错`ImportError: libGL.so.1: cannot open shared object file: No such file or directory`
+
+WSL2的Ubuntu22.04中缺少`libgl`库，可通过以下命令安装`libgl`库解决：
+
+```bash
+sudo apt-get install libgl1-mesa-glx
+```
+
+参考：https://github.com/opendatalab/MinerU/issues/388
+
+### 5.遇到报错 `ModuleNotFoundError : Nomodulenamed 'fairscale'`
+
+需要卸载该模块并重新安装
+
+```bash
+pip uninstall fairscale
+pip install fairscale
+```
+
+参考：https://github.com/opendatalab/MinerU/issues/411
+
+### 6.在部分较新的设备如H100上，使用CUDA加速OCR时解析出的文字乱码。
+
+cuda11对新显卡的兼容性不好，需要升级paddle使用的cuda版本
+
+```bash
+pip install paddlepaddle-gpu==3.0.0b1 -i https://www.paddlepaddle.org.cn/packages/stable/cu123/
+```
+参考：https://github.com/opendatalab/MinerU/issues/558
+
+### 7.在部分Linux服务器上，程序一运行就报错 `非法指令 (核心已转储)` 或 `Illegal instruction (core dumped)`
+
+可能是因为服务器CPU不支持AVX/AVX2指令集，或cpu本身支持但被运维禁用了，可以尝试联系运维解除限制或更换服务器。
+
+参考：https://github.com/opendatalab/MinerU/issues/591 ， https://github.com/opendatalab/MinerU/issues/736
+
+### 8.在 CentOS 7 或 Ubuntu 18 系统安装MinerU时报错`ERROR: Failed building wheel for simsimd`
+
+新版本albumentations(1.4.21)引入了依赖simsimd,由于simsimd在linux的预编译包要求glibc的版本大于等于2.28，导致部分2019年之前发布的Linux发行版无法正常安装，可通过如下命令安装:
+```
+pip install -U magic-pdf[full,old_linux] --extra-index-url https://wheels.myhloli.com
+```
+
+参考：https://github.com/opendatalab/MinerU/issues/1004
+
+### 9. 旧显卡如M40出现 "RuntimeError: CUDA error: CUBLAS_STATUS_NOT_SUPPORTED"
+
+在运行过程中（使用CUDA）出现以下错误：
+```
+RuntimeError: CUDA error: CUBLAS_STATUS_NOT_SUPPORTED when calling cublasGemmStridedBatchedEx(handle, opa, opb, (int)m, (int)n, (int)k, (void*)&falpha, a, CUDA_R_16BF, (int)lda, stridea, b, CUDA_R_16BF, (int)ldb, strideb, (void*)&fbeta, c, CUDA_R_16BF, (int)ldc, stridec, (int)num_batches, compute_type, CUBLAS_GEMM_DEFAULT_TENSOR_OP)
+```
+由于Turing架构之前的显卡不支持BF16精度，并且部分显卡未能被PyTorch正确识别，因此需要手动关闭BF16精度。
+
+请找到并修改`pdf_parse_union_core_v2.py`文件中的第287至290行代码（注意：不同版本中位置可能有所不同），原代码如下：
+```python
+if torch.cuda.is_bf16_supported():
+    supports_bfloat16 = True
+else:
+    supports_bfloat16 = False
+```
+将其修改为：
+```python
+supports_bfloat16 = False
+```
+
+参考：https://github.com/opendatalab/MinerU/issues/1508

third_party/MinerU/docs/README_Ascend_NPU_Acceleration_zh_CN.md

@@ -0,0 +1,51 @@
+# Ascend NPU 加速
+
+## 简介
+
+本文档介绍如何在 Ascend NPU 上使用 MinerU。本文档内容已在`华为 Atlas 800T A2`服务器上测试通过。
+```
+CPU：鲲鹏 920 aarch64 2.6GHz
+NPU：Ascend 910B 64GB
+OS：openEuler 22.03 (LTS-SP3)/ Ubuntu 22.04.5 LTS
+CANN：8.0.RC2
+驱动版本：24.1.rc2.1
+```
+由于适配 Ascend NPU 的环境较为复杂，建议使用 Docker 容器运行 MinerU。
+
+通过docker运行MinerU前需确保物理机已安装支持CANN 8.0.RC2的驱动和固件。
+
+
+## 构建镜像
+请保持网络状况良好，并执行以下代码构建镜像。    
+```bash
+wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/docker/ascend_npu/Dockerfile -O Dockerfile
+docker build -t mineru_npu:latest .
+```
+如果构建过程中未发生报错则说明镜像构建成功。
+
+
+## 运行容器
+
+```bash
+docker run -it -u root --name mineru-npu --privileged=true \
+    --ipc=host \
+    --network=host \
+    --device=/dev/davinci0 \
+    --device=/dev/davinci1 \
+    --device=/dev/davinci2 \
+    --device=/dev/davinci3 \
+    --device=/dev/davinci4 \
+    --device=/dev/davinci5 \
+    --device=/dev/davinci6 \
+    --device=/dev/davinci7 \
+    --device=/dev/davinci_manager \
+    --device=/dev/devmm_svm \
+    --device=/dev/hisi_hdc \
+    -v /var/log/npu/:/usr/slog \
+    -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
+    -v /usr/local/Ascend/driver:/usr/local/Ascend/driver \
+    mineru_npu:latest \
+    /bin/bash -c "echo 'source /opt/mineru_venv/bin/activate' >> ~/.bashrc && exec bash"
+
+magic-pdf --help
+```

third_party/MinerU/docs/README_Ubuntu_CUDA_Acceleration_en_US.md

@@ -0,0 +1,111 @@
+# Ubuntu 22.04 LTS
+
+### 1. Check if NVIDIA Drivers Are Installed
+
+```sh
+nvidia-smi
+```
+
+If you see information similar to the following, it means that the NVIDIA drivers are already installed, and you can skip Step 2.
+
+> [!NOTE]
+> Notice:`CUDA Version` should be >= 12.4, If the displayed version number is less than 12.4, please upgrade the driver.
+
+```plaintext
++---------------------------------------------------------------------------------------+
+| NVIDIA-SMI 570.133.07             Driver Version: 572.83         CUDA Version: 12.8   |
+|-----------------------------------------+----------------------+----------------------+
+| GPU  Name                     TCC/WDDM  | Bus-Id        Disp.A | Volatile Uncorr. ECC |
+| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
+|                                         |                      |               MIG M. |
+|=========================================+======================+======================|
+|   0  NVIDIA GeForce RTX 3060 Ti   WDDM  | 00000000:01:00.0  On |                  N/A |
+|  0%   51C    P8              12W / 200W |   1489MiB /  8192MiB |      5%      Default |
+|                                         |                      |                  N/A |
++-----------------------------------------+----------------------+----------------------+
+```
+
+### 2. Install the Driver
+
+If no driver is installed, use the following command:
+
+```sh
+sudo apt-get update
+sudo apt-get install nvidia-driver-570-server
+```
+
+Install the proprietary driver and restart your computer after installation.
+
+```sh
+reboot
+```
+
+### 3. Install Anaconda
+
+If Anaconda is already installed, skip this step.
+
+```sh
+wget https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Linux-x86_64.sh
+bash Anaconda3-2024.06-1-Linux-x86_64.sh
+```
+
+In the final step, enter `yes`, close the terminal, and reopen it.
+
+### 4. Create an Environment Using Conda
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+```
+
+### 5. Install Applications
+
+```sh
+pip install -U magic-pdf[full]
+```
+> [!TIP]
+> After installation, you can check the version of `magic-pdf` using the following command:
+>
+> ```sh
+> magic-pdf --version
+> ```
+
+
+### 6. Download Models
+
+
+Refer to detailed instructions on [how to download model files](how_to_download_models_en.md).
+
+
+## 7. Understand the Location of the Configuration File
+
+After completing the [6. Download Models](#6-download-models) step, the script will automatically generate a `magic-pdf.json` file in the user directory and configure the default model path.
+You can find the `magic-pdf.json` file in your user directory.
+
+> [!TIP]
+> The user directory for Linux is "/home/username".
+
+
+### 8. First Run
+
+Download a sample file from the repository and test it.
+
+```sh
+wget https://github.com/opendatalab/MinerU/raw/master/demo/pdfs/small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
+```
+
+### 9. Test CUDA Acceleration
+
+If your graphics card has at least **6GB** of VRAM, follow these steps to test CUDA acceleration:
+
+1. Modify the value of `"device-mode"` in the `magic-pdf.json` configuration file located in your home directory.
+   ```json
+   {
+     "device-mode": "cuda"
+   }
+   ```
+2. Test CUDA acceleration with the following command:
+   ```sh
+   magic-pdf -p small_ocr.pdf -o ./output
+   ```

third_party/MinerU/docs/README_Ubuntu_CUDA_Acceleration_zh_CN.md

@@ -0,0 +1,115 @@
+# Ubuntu 22.04 LTS
+
+## 1. 检测是否已安装nvidia驱动
+
+```bash
+nvidia-smi
+```
+
+如果看到类似如下的信息，说明已经安装了nvidia驱动，可以跳过步骤2
+
+> [!NOTE]
+> `CUDA Version` 显示的版本号应 >= 12.4，如显示的版本号小于12.4，请升级驱动
+
+```plaintext
++---------------------------------------------------------------------------------------+
+| NVIDIA-SMI 570.133.07             Driver Version: 572.83         CUDA Version: 12.8   |
+|-----------------------------------------+----------------------+----------------------+
+| GPU  Name                     TCC/WDDM  | Bus-Id        Disp.A | Volatile Uncorr. ECC |
+| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
+|                                         |                      |               MIG M. |
+|=========================================+======================+======================|
+|   0  NVIDIA GeForce RTX 3060 Ti   WDDM  | 00000000:01:00.0  On |                  N/A |
+|  0%   51C    P8              12W / 200W |   1489MiB /  8192MiB |      5%      Default |
+|                                         |                      |                  N/A |
++-----------------------------------------+----------------------+----------------------+
+```
+
+## 2. 安装驱动
+
+如没有驱动，则通过如下命令
+
+```bash
+sudo apt-get update
+sudo apt-get install nvidia-driver-570-server
+```
+
+安装专有驱动，安装完成后，重启电脑
+
+```bash
+reboot
+```
+
+## 3. 安装anacoda
+
+如果已安装conda，可以跳过本步骤
+
+```bash
+wget -U NoSuchBrowser/1.0 https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/Anaconda3-2024.06-1-Linux-x86_64.sh
+bash Anaconda3-2024.06-1-Linux-x86_64.sh
+```
+
+最后一步输入yes，关闭终端重新打开
+
+## 4. 使用conda 创建环境
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+```
+
+## 5. 安装应用
+
+```bash
+pip install -U magic-pdf[full] -i https://mirrors.aliyun.com/pypi/simple
+```
+
+> [!TIP]
+> 下载完成后，您可以通过以下命令检查`magic-pdf`的版本：
+>
+> ```bash
+> magic-pdf --version
+> ```
+
+
+## 6. 下载模型
+
+
+详细参考 [如何下载模型文件](how_to_download_models_zh_cn.md)
+
+## 7. 了解配置文件存放的位置
+
+完成[6.下载模型](#6-下载模型)步骤后，脚本会自动生成用户目录下的magic-pdf.json文件，并自动配置默认模型路径。
+您可在【用户目录】下找到magic-pdf.json文件。
+
+> [!TIP]
+> linux用户目录为 "/home/用户名"
+
+## 8. 第一次运行
+
+从仓库中下载样本文件，并测试
+
+```bash
+wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/demo/pdfs/small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
+```
+
+## 9. 测试CUDA加速
+
+如果您的显卡显存大于等于 **6GB** ，可以进行以下流程，测试CUDA解析加速效果
+
+**1.修改【用户目录】中配置文件magic-pdf.json中"device-mode"的值**
+
+```json
+{
+  "device-mode":"cuda"
+}
+```
+
+**2.运行以下命令测试cuda加速效果**
+
+```bash
+magic-pdf -p small_ocr.pdf -o ./output
+```
+> [!TIP]
+> CUDA加速是否生效可以根据log中输出的各个阶段cost耗时来简单判断，通常情况下，使用cuda加速会比cpu更快。

third_party/MinerU/docs/README_Windows_CUDA_Acceleration_en_US.md

@@ -0,0 +1,83 @@
+# Windows 10/11
+
+### 1. Install CUDA and cuDNN
+
+You need to install a CUDA version that is compatible with torch's requirements. For details, please refer to the [official PyTorch website](https://pytorch.org/get-started/locally/).
+
+- CUDA 11.8 https://developer.nvidia.com/cuda-11-8-0-download-archive
+- CUDA 12.4 https://developer.nvidia.com/cuda-12-4-0-download-archive
+- CUDA 12.6 https://developer.nvidia.com/cuda-12-6-0-download-archive
+- CUDA 12.8 https://developer.nvidia.com/cuda-12-8-0-download-archive
+
+### 2. Install Anaconda
+
+If Anaconda is already installed, you can skip this step.
+
+Download link: https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Windows-x86_64.exe
+
+### 3. Create an Environment Using Conda
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+```
+
+### 4. Install Applications
+
+```
+pip install -U magic-pdf[full]
+```
+
+> [!IMPORTANT]
+> After installation, you can check the version of `magic-pdf` using the following command:
+>
+> ```bash
+> magic-pdf --version
+> ```
+
+
+### 5. Download Models
+
+Refer to detailed instructions on [how to download model files](how_to_download_models_en.md).
+
+### 6. Understand the Location of the Configuration File
+
+After completing the [5. Download Models](#5-download-models) step, the script will automatically generate a `magic-pdf.json` file in the user directory and configure the default model path.
+You can find the `magic-pdf.json` file in your 【user directory】 .
+
+> [!TIP]
+> The user directory for Windows is "C:/Users/username".
+
+### 7. First Run
+
+Download a sample file from the repository and test it.
+
+```powershell
+  wget https://github.com/opendatalab/MinerU/raw/master/demo/pdfs/small_ocr.pdf -O small_ocr.pdf
+  magic-pdf -p small_ocr.pdf -o ./output
+```
+
+### 8. Test CUDA Acceleration
+
+If your graphics card has at least 6GB of VRAM, follow these steps to test CUDA-accelerated parsing performance.
+
+1. **Overwrite the installation of torch and torchvision** supporting CUDA.(Please select the appropriate index-url based on your CUDA version. For more details, refer to the [PyTorch official website](https://pytorch.org/get-started/locally/).)
+
+   ```
+   pip install --force-reinstall torch torchvision --index-url https://download.pytorch.org/whl/cu124
+   ```
+
+2. **Modify the value of `"device-mode"`** in the `magic-pdf.json` configuration file located in your user directory.
+
+   ```json
+   {
+     "device-mode": "cuda"
+   }
+   ```
+
+
+3. **Run the following command to test CUDA acceleration**:
+
+   ```
+   magic-pdf -p small_ocr.pdf -o ./output
+   ```

third_party/MinerU/docs/README_Windows_CUDA_Acceleration_zh_CN.md

@@ -0,0 +1,86 @@
+# Windows10/11
+
+## 1. 安装cuda环境
+
+需要安装符合torch要求的cuda版本，具体可参考[torch官网](https://pytorch.org/get-started/locally/)
+
+- CUDA 11.8 https://developer.nvidia.com/cuda-11-8-0-download-archive
+- CUDA 12.4 https://developer.nvidia.com/cuda-12-4-0-download-archive
+- CUDA 12.6 https://developer.nvidia.com/cuda-12-6-0-download-archive
+- CUDA 12.8 https://developer.nvidia.com/cuda-12-8-0-download-archive
+
+## 2. 安装anaconda
+
+如果已安装conda，可以跳过本步骤
+
+下载链接：
+https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/Anaconda3-2024.06-1-Windows-x86_64.exe
+
+## 3. 使用conda 创建环境
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+```
+
+## 4. 安装应用
+
+```bash
+pip install -U magic-pdf[full] -i https://mirrors.aliyun.com/pypi/simple
+```
+
+> [!IMPORTANT]
+> 下载完成后，您可以通过以下命令检查magic-pdf的版本
+>
+> ```bash
+> magic-pdf --version
+> ```
+
+
+## 5. 下载模型
+
+详细参考 [如何下载模型文件](how_to_download_models_zh_cn.md)
+
+## 6. 了解配置文件存放的位置
+
+完成[5.下载模型](#5-下载模型)步骤后，脚本会自动生成用户目录下的magic-pdf.json文件，并自动配置默认模型路径。
+您可在【用户目录】下找到magic-pdf.json文件。
+
+> [!TIP]
+> windows用户目录为 "C:/Users/用户名"
+
+## 7. 第一次运行
+
+从仓库中下载样本文件，并测试
+
+```powershell
+ wget https://github.com/opendatalab/MinerU/raw/master/demo/pdfs/small_ocr.pdf -O small_ocr.pdf
+ magic-pdf -p small_ocr.pdf -o ./output
+```
+
+## 8. 测试CUDA加速
+
+如果您的显卡显存大于等于 **6GB** ，可以进行以下流程，测试CUDA解析加速效果
+
+**1.覆盖安装支持cuda的torch和torchvision**(请根据cuda版本选择合适的index-url，具体可参考[torch官网](https://pytorch.org/get-started/locally/))
+
+```bash
+pip install --force-reinstall torch torchvision --index-url https://download.pytorch.org/whl/cu124
+```
+
+**2.修改【用户目录】中配置文件magic-pdf.json中"device-mode"的值**
+
+```json
+{
+  "device-mode":"cuda"
+}
+```
+
+**3.运行以下命令测试cuda加速效果**
+
+```bash
+magic-pdf -p small_ocr.pdf -o ./output
+```
+
+> [!TIP]
+> CUDA加速是否生效可以根据log中输出的各个阶段的耗时来简单判断，通常情况下，cuda加速后运行速度比cpu更快。

third_party/MinerU/docs/how_to_download_models_en.md

@@ -0,0 +1,23 @@
+Model downloads are divided into initial downloads and updates to the model directory. Please refer to the corresponding documentation for instructions on how to proceed.
+
+
+# Initial download of model files
+
+### Download the Model from Hugging Face
+
+Use a Python Script to Download Model Files from Hugging Face
+```bash
+pip install huggingface_hub
+wget https://github.com/opendatalab/MinerU/raw/master/scripts/download_models_hf.py -O download_models_hf.py
+python download_models_hf.py
+```
+The Python script will automatically download the model files and configure the model directory in the configuration file.
+
+The configuration file can be found in the user directory, with the filename `magic-pdf.json`.
+
+
+# How to update models previously downloaded
+
+## 1. Models downloaded via Hugging Face or Model Scope
+
+If you previously downloaded models via Hugging Face or Model Scope, you can rerun the Python script used for the initial download. This will automatically update the model directory to the latest version.

third_party/MinerU/docs/how_to_download_models_zh_cn.md

@@ -0,0 +1,37 @@
+模型下载分为首次下载和更新模型目录，请参考对应的文档内容进行操作
+
+# 首次下载模型文件
+
+模型文件可以从 Hugging Face 或 Model Scope 下载，由于网络原因，国内用户访问HF可能会失败，请使用 ModelScope。
+
+<details>
+  <summary>方法一：从 Hugging Face 下载模型</summary>
+  <p>使用python脚本 从Hugging Face下载模型文件</p>
+  <pre><code>pip install huggingface_hub
+wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/scripts/download_models_hf.py -O download_models_hf.py
+python download_models_hf.py</code></pre>
+  <p>python脚本会自动下载模型文件并配置好配置文件中的模型目录</p>
+</details>
+
+## 方法二：从 ModelScope 下载模型
+
+### 使用python脚本 从ModelScope下载模型文件
+
+```bash
+pip install modelscope
+wget https://gcore.jsdelivr.net/gh/opendatalab/MinerU@master/scripts/download_models.py -O download_models.py
+python download_models.py
+```
+python脚本会自动下载模型文件并配置好配置文件中的模型目录
+
+配置文件可以在用户目录中找到，文件名为`magic-pdf.json`
+
+> [!TIP]
+> windows的用户目录为 "C:\\Users\\用户名", linux用户目录为 "/home/用户名", macOS用户目录为 "/Users/用户名"
+
+
+# 此前下载过模型，如何更新
+
+## 1. 通过 Hugging Face 或 Model Scope 下载过模型
+
+如此前通过 HuggingFace 或 Model Scope 下载过模型，可以重复执行此前的模型下载python脚本，将会自动将模型目录更新到最新版本。

third_party/MinerU/docs/output_file_en_us.md

@@ -0,0 +1,330 @@
+## Overview
+
+After executing the `magic-pdf` command, in addition to outputting files related to markdown, several other files unrelated to markdown will also be generated. These files will be introduced one by one.
+
+### some_pdf_layout.pdf
+
+Each page layout consists of one or more boxes. The number at the top left of each box indicates its sequence number. Additionally, in `layout.pdf`, different content blocks are highlighted with different background colors.
+
+![layout example](images/layout_example.png)
+
+### some_pdf_spans.pdf
+
+All spans on the page are drawn with different colored line frames according to the span type. This file can be used for quality control, allowing for quick identification of issues such as missing text or unrecognized inline formulas.
+
+![spans example](images/spans_example.png)
+
+### some_pdf_model.json
+
+#### Structure Definition
+
+```python
+from pydantic import BaseModel, Field
+from enum import IntEnum
+
+class CategoryType(IntEnum):
+     title = 0               # Title
+     plain_text = 1          # Text
+     abandon = 2             # Includes headers, footers, page numbers, and page annotations
+     figure = 3              # Image
+     figure_caption = 4      # Image description
+     table = 5               # Table
+     table_caption = 6       # Table description
+     table_footnote = 7      # Table footnote
+     isolate_formula = 8     # Block formula
+     formula_caption = 9     # Formula label
+
+     embedding = 13          # Inline formula
+     isolated = 14           # Block formula
+     text = 15               # OCR recognition result
+
+
+class PageInfo(BaseModel):
+    page_no: int = Field(description="Page number, the first page is 0", ge=0)
+    height: int = Field(description="Page height", gt=0)
+    width: int = Field(description="Page width", ge=0)
+
+class ObjectInferenceResult(BaseModel):
+    category_id: CategoryType = Field(description="Category", ge=0)
+    poly: list[float] = Field(description="Quadrilateral coordinates, representing the coordinates of the top-left, top-right, bottom-right, and bottom-left points respectively")
+    score: float = Field(description="Confidence of the inference result")
+    latex: str | None = Field(description="LaTeX parsing result", default=None)
+    html: str | None = Field(description="HTML parsing result", default=None)
+
+class PageInferenceResults(BaseModel):
+     layout_dets: list[ObjectInferenceResult] = Field(description="Page recognition results", ge=0)
+     page_info: PageInfo = Field(description="Page metadata")
+
+
+# The inference results of all pages, ordered by page number, are stored in a list as the inference results of MinerU
+inference_result: list[PageInferenceResults] = []
+
+```
+
+The format of the poly coordinates is \[x0, y0, x1, y1, x2, y2, x3, y3\], representing the coordinates of the top-left, top-right, bottom-right, and bottom-left points respectively.
+![Poly Coordinate Diagram](images/poly.png)
+
+#### example
+
+```json
+[
+    {
+        "layout_dets": [
+            {
+                "category_id": 2,
+                "poly": [
+                    99.1906967163086,
+                    100.3119125366211,
+                    730.3707885742188,
+                    100.3119125366211,
+                    730.3707885742188,
+                    245.81326293945312,
+                    99.1906967163086,
+                    245.81326293945312
+                ],
+                "score": 0.9999997615814209
+            }
+        ],
+        "page_info": {
+            "page_no": 0,
+            "height": 2339,
+            "width": 1654
+        }
+    },
+    {
+        "layout_dets": [
+            {
+                "category_id": 5,
+                "poly": [
+                    99.13092803955078,
+                    2210.680419921875,
+                    497.3183898925781,
+                    2210.680419921875,
+                    497.3183898925781,
+                    2264.78076171875,
+                    99.13092803955078,
+                    2264.78076171875
+                ],
+                "score": 0.9999997019767761
+            }
+        ],
+        "page_info": {
+            "page_no": 1,
+            "height": 2339,
+            "width": 1654
+        }
+    }
+]
+```
+
+### some_pdf_middle.json
+
+| Field Name     | Description                                                                                                    |
+| :------------- | :------------------------------------------------------------------------------------------------------------- |
+| pdf_info       | list, each element is a dict representing the parsing result of each PDF page, see the table below for details |
+| \_parse_type   | ocr \| txt, used to indicate the mode used in this intermediate parsing state                                  |
+| \_version_name | string, indicates the version of magic-pdf used in this parsing                                                |
+
+<br>
+
+**pdf_info**
+
+Field structure description
+
+| Field Name          | Description                                                                                                        |
+| :------------------ | :----------------------------------------------------------------------------------------------------------------- |
+| preproc_blocks      | Intermediate result after PDF preprocessing, not yet segmented                                                     |
+| layout_bboxes       | Layout segmentation results, containing layout direction (vertical, horizontal), and bbox, sorted by reading order |
+| page_idx            | Page number, starting from 0                                                                                       |
+| page_size           | Page width and height                                                                                              |
+| \_layout_tree       | Layout tree structure                                                                                              |
+| images              | list, each element is a dict representing an img_block                                                             |
+| tables              | list, each element is a dict representing a table_block                                                            |
+| interline_equations | list, each element is a dict representing an interline_equation_block                                              |
+| discarded_blocks    | List, block information returned by the model that needs to be dropped                                             |
+| para_blocks         | Result after segmenting preproc_blocks                                                                             |
+
+In the above table, `para_blocks` is an array of dicts, each dict representing a block structure. A block can support up to one level of nesting.
+
+<br>
+
+**block**
+
+The outer block is referred to as a first-level block, and the fields in the first-level block include:
+
+| Field Name | Description                                                    |
+| :--------- | :------------------------------------------------------------- |
+| type       | Block type (table\|image)                                      |
+| bbox       | Block bounding box coordinates                                 |
+| blocks     | list, each element is a dict representing a second-level block |
+
+<br>
+There are only two types of first-level blocks: "table" and "image". All other blocks are second-level blocks.
+
+The fields in a second-level block include:
+
+| Field Name | Description                                                                                                 |
+| :--------- | :---------------------------------------------------------------------------------------------------------- |
+| type       | Block type                                                                                                  |
+| bbox       | Block bounding box coordinates                                                                              |
+| lines      | list, each element is a dict representing a line, used to describe the composition of a line of information |
+
+Detailed explanation of second-level block types
+
+| type               | Description            |
+| :----------------- | :--------------------- |
+| image_body         | Main body of the image |
+| image_caption      | Image description text |
+| image_footnote     | Image footnote         |
+| table_body         | Main body of the table |
+| table_caption      | Table description text |
+| table_footnote     | Table footnote         |
+| text               | Text block             |
+| title              | Title block            |
+| index              | Index block            |
+| list               | List block             |
+| interline_equation | Block formula          |
+
+<br>
+
+**line**
+
+The field format of a line is as follows:
+
+| Field Name | Description                                                                                             |
+| :--------- | :------------------------------------------------------------------------------------------------------ |
+| bbox       | Bounding box coordinates of the line                                                                    |
+| spans      | list, each element is a dict representing a span, used to describe the composition of the smallest unit |
+
+<br>
+
+**span**
+
+| Field Name          | Description                                                                                              |
+| :------------------ | :------------------------------------------------------------------------------------------------------- |
+| bbox                | Bounding box coordinates of the span                                                                     |
+| type                | Type of the span                                                                                         |
+| content \| img_path | Text spans use content, chart spans use img_path to store the actual text or screenshot path information |
+
+The types of spans are as follows:
+
+| type               | Description    |
+| :----------------- | :------------- |
+| image              | Image          |
+| table              | Table          |
+| text               | Text           |
+| inline_equation    | Inline formula |
+| interline_equation | Block formula  |
+
+**Summary**
+
+A span is the smallest storage unit for all elements.
+
+The elements stored within para_blocks are block information.
+
+The block structure is as follows:
+
+First-level block (if any) -> Second-level block -> Line -> Span
+
+#### example
+
+```json
+{
+    "pdf_info": [
+        {
+            "preproc_blocks": [
+                {
+                    "type": "text",
+                    "bbox": [
+                        52,
+                        61.956024169921875,
+                        294,
+                        82.99800872802734
+                    ],
+                    "lines": [
+                        {
+                            "bbox": [
+                                52,
+                                61.956024169921875,
+                                294,
+                                72.0000228881836
+                            ],
+                            "spans": [
+                                {
+                                    "bbox": [
+                                        54.0,
+                                        61.956024169921875,
+                                        296.2261657714844,
+                                        72.0000228881836
+                                    ],
+                                    "content": "dependent on the service headway and the reliability of the departure ",
+                                    "type": "text",
+                                    "score": 1.0
+                                }
+                            ]
+                        }
+                    ]
+                }
+            ],
+            "layout_bboxes": [
+                {
+                    "layout_bbox": [
+                        52,
+                        61,
+                        294,
+                        731
+                    ],
+                    "layout_label": "V",
+                    "sub_layout": []
+                }
+            ],
+            "page_idx": 0,
+            "page_size": [
+                612.0,
+                792.0
+            ],
+            "_layout_tree": [],
+            "images": [],
+            "tables": [],
+            "interline_equations": [],
+            "discarded_blocks": [],
+            "para_blocks": [
+                {
+                    "type": "text",
+                    "bbox": [
+                        52,
+                        61.956024169921875,
+                        294,
+                        82.99800872802734
+                    ],
+                    "lines": [
+                        {
+                            "bbox": [
+                                52,
+                                61.956024169921875,
+                                294,
+                                72.0000228881836
+                            ],
+                            "spans": [
+                                {
+                                    "bbox": [
+                                        54.0,
+                                        61.956024169921875,
+                                        296.2261657714844,
+                                        72.0000228881836
+                                    ],
+                                    "content": "dependent on the service headway and the reliability of the departure ",
+                                    "type": "text",
+                                    "score": 1.0
+                                }
+                            ]
+                        }
+                    ]
+                }
+            ]
+        }
+    ],
+    "_parse_type": "txt",
+    "_version_name": "0.6.1"
+}
+```

third_party/MinerU/docs/output_file_zh_cn.md

@@ -0,0 +1,329 @@
+## 概览
+
+`magic-pdf` 命令执行后除了输出和 markdown 有关的文件以外，还会生成若干个和 markdown 无关的文件。现在将一一介绍这些文件
+
+### some_pdf_layout.pdf
+
+每一页的 layout 均由一个或多个框组成。 每个框左上脚的数字表明它们的序号。此外 layout.pdf 框内用不同的背景色块圈定不同的内容块。
+
+![layout 页面示例](images/layout_example.png)
+
+### some_pdf_spans.pdf
+
+根据 span 类型的不同，采用不同颜色线框绘制页面上所有 span。该文件可以用于质检，可以快速排查出文本丢失、行间公式未识别等问题。
+
+![span 页面示例](images/spans_example.png)
+
+### some_pdf_model.json
+
+#### 结构定义
+
+```python
+from pydantic import BaseModel, Field
+from enum import IntEnum
+
+class CategoryType(IntEnum):
+     title = 0               # 标题
+     plain_text = 1          # 文本
+     abandon = 2             # 包括页眉页脚页码和页面注释
+     figure = 3              # 图片
+     figure_caption = 4      # 图片描述
+     table = 5               # 表格
+     table_caption = 6       # 表格描述
+     table_footnote = 7      # 表格注释
+     isolate_formula = 8     # 行间公式
+     formula_caption = 9     # 行间公式的标号
+
+     embedding = 13          # 行内公式
+     isolated = 14           # 行间公式
+     text = 15               # ocr 识别结果
+
+
+class PageInfo(BaseModel):
+    page_no: int = Field(description="页码序号，第一页的序号是 0", ge=0)
+    height: int = Field(description="页面高度", gt=0)
+    width: int = Field(description="页面宽度", ge=0)
+
+class ObjectInferenceResult(BaseModel):
+    category_id: CategoryType = Field(description="类别", ge=0)
+    poly: list[float] = Field(description="四边形坐标, 分别是 左上，右上，右下，左下 四点的坐标")
+    score: float = Field(description="推理结果的置信度")
+    latex: str | None = Field(description="latex 解析结果", default=None)
+    html: str | None = Field(description="html 解析结果", default=None)
+
+class PageInferenceResults(BaseModel):
+     layout_dets: list[ObjectInferenceResult] = Field(description="页面识别结果", ge=0)
+     page_info: PageInfo = Field(description="页面元信息")
+
+
+# 所有页面的推理结果按照页码顺序依次放到列表中即为 minerU 推理结果
+inference_result: list[PageInferenceResults] = []
+
+```
+
+poly 坐标的格式 \[x0, y0, x1, y1, x2, y2, x3, y3\], 分别表示左上、右上、右下、左下四点的坐标
+![poly 坐标示意图](images/poly.png)
+
+#### 示例数据
+
+```json
+[
+    {
+        "layout_dets": [
+            {
+                "category_id": 2,
+                "poly": [
+                    99.1906967163086,
+                    100.3119125366211,
+                    730.3707885742188,
+                    100.3119125366211,
+                    730.3707885742188,
+                    245.81326293945312,
+                    99.1906967163086,
+                    245.81326293945312
+                ],
+                "score": 0.9999997615814209
+            }
+        ],
+        "page_info": {
+            "page_no": 0,
+            "height": 2339,
+            "width": 1654
+        }
+    },
+    {
+        "layout_dets": [
+            {
+                "category_id": 5,
+                "poly": [
+                    99.13092803955078,
+                    2210.680419921875,
+                    497.3183898925781,
+                    2210.680419921875,
+                    497.3183898925781,
+                    2264.78076171875,
+                    99.13092803955078,
+                    2264.78076171875
+                ],
+                "score": 0.9999997019767761
+            }
+        ],
+        "page_info": {
+            "page_no": 1,
+            "height": 2339,
+            "width": 1654
+        }
+    }
+]
+```
+
+### some_pdf_middle.json
+
+| 字段名         | 解释                                                               |
+| :------------- | :----------------------------------------------------------------- |
+| pdf_info       | list，每个元素都是一个dict,这个dict是每一页pdf的解析结果，详见下表 |
+| \_parse_type   | ocr \| txt，用来标识本次解析的中间态使用的模式                     |
+| \_version_name | string, 表示本次解析使用的 magic-pdf 的版本号                      |
+
+<br>
+
+**pdf_info**
+字段结构说明
+
+| 字段名              | 解释                                                                 |
+| :------------------ | :------------------------------------------------------------------- |
+| preproc_blocks      | pdf预处理后，未分段的中间结果                                        |
+| layout_bboxes       | 布局分割的结果，含有布局的方向（垂直、水平），和bbox，按阅读顺序排序 |
+| page_idx            | 页码，从0开始                                                        |
+| page_size           | 页面的宽度和高度                                                     |
+| \_layout_tree       | 布局树状结构                                                         |
+| images              | list，每个元素是一个dict，每个dict表示一个img_block                  |
+| tables              | list，每个元素是一个dict，每个dict表示一个table_block                |
+| interline_equations | list，每个元素是一个dict，每个dict表示一个interline_equation_block   |
+| discarded_blocks    | List, 模型返回的需要drop的block信息                                  |
+| para_blocks         | 将preproc_blocks进行分段之后的结果                                   |
+
+上表中 `para_blocks` 是个dict的数组，每个dict是一个block结构，block最多支持一次嵌套
+
+<br>
+
+**block**
+
+外层block被称为一级block，一级block中的字段包括
+
+| 字段名 | 解释                                            |
+| :----- | :---------------------------------------------- |
+| type   | block类型（table\|image）                       |
+| bbox   | block矩形框坐标                                 |
+| blocks | list，里面的每个元素都是一个dict格式的二级block |
+
+<br>
+一级block只有"table"和"image"两种类型，其余block均为二级block
+
+二级block中的字段包括
+
+| 字段名 | 解释                                                         |
+| :----- | :----------------------------------------------------------- |
+| type   | block类型                                                    |
+| bbox   | block矩形框坐标                                              |
+| lines  | list，每个元素都是一个dict表示的line，用来描述一行信息的构成 |
+
+二级block的类型详解
+
+| type               | desc           |
+| :----------------- | :------------- |
+| image_body         | 图像的本体     |
+| image_caption      | 图像的描述文本 |
+| image_footnote     | 图像的脚注   |
+| table_body         | 表格本体    |
+| table_caption      | 表格的描述文本 |
+| table_footnote     | 表格的脚注   |
+| text               | 文本块     |
+| title              | 标题块     |
+| index              | 目录块     |
+| list               | 列表块     |
+| interline_equation | 行间公式块   |
+
+<br>
+
+**line**
+
+line 的 字段格式如下
+
+| 字段名 | 解释                                                                 |
+| :----- | :------------------------------------------------------------------- |
+| bbox   | line的矩形框坐标                                                     |
+| spans  | list，每个元素都是一个dict表示的span，用来描述一个最小组成单元的构成 |
+
+<br>
+
+**span**
+
+| 字段名              | 解释                                                                             |
+| :------------------ | :------------------------------------------------------------------------------- |
+| bbox                | span的矩形框坐标                                                                 |
+| type                | span的类型                                                                       |
+| content \| img_path | 文本类型的span使用content，图表类使用img_path 用来存储实际的文本或者截图路径信息 |
+
+span 的类型有如下几种
+
+| type               | desc     |
+| :----------------- | :------- |
+| image              | 图片     |
+| table              | 表格     |
+| text               | 文本     |
+| inline_equation    | 行内公式 |
+| interline_equation | 行间公式 |
+
+**总结**
+
+span是所有元素的最小存储单元
+
+para_blocks内存储的元素为区块信息
+
+区块结构为
+
+一级block(如有)->二级block->line->span
+
+#### 示例数据
+
+```json
+{
+    "pdf_info": [
+        {
+            "preproc_blocks": [
+                {
+                    "type": "text",
+                    "bbox": [
+                        52,
+                        61.956024169921875,
+                        294,
+                        82.99800872802734
+                    ],
+                    "lines": [
+                        {
+                            "bbox": [
+                                52,
+                                61.956024169921875,
+                                294,
+                                72.0000228881836
+                            ],
+                            "spans": [
+                                {
+                                    "bbox": [
+                                        54.0,
+                                        61.956024169921875,
+                                        296.2261657714844,
+                                        72.0000228881836
+                                    ],
+                                    "content": "dependent on the service headway and the reliability of the departure ",
+                                    "type": "text",
+                                    "score": 1.0
+                                }
+                            ]
+                        }
+                    ]
+                }
+            ],
+            "layout_bboxes": [
+                {
+                    "layout_bbox": [
+                        52,
+                        61,
+                        294,
+                        731
+                    ],
+                    "layout_label": "V",
+                    "sub_layout": []
+                }
+            ],
+            "page_idx": 0,
+            "page_size": [
+                612.0,
+                792.0
+            ],
+            "_layout_tree": [],
+            "images": [],
+            "tables": [],
+            "interline_equations": [],
+            "discarded_blocks": [],
+            "para_blocks": [
+                {
+                    "type": "text",
+                    "bbox": [
+                        52,
+                        61.956024169921875,
+                        294,
+                        82.99800872802734
+                    ],
+                    "lines": [
+                        {
+                            "bbox": [
+                                52,
+                                61.956024169921875,
+                                294,
+                                72.0000228881836
+                            ],
+                            "spans": [
+                                {
+                                    "bbox": [
+                                        54.0,
+                                        61.956024169921875,
+                                        296.2261657714844,
+                                        72.0000228881836
+                                    ],
+                                    "content": "dependent on the service headway and the reliability of the departure ",
+                                    "type": "text",
+                                    "score": 1.0
+                                }
+                            ]
+                        }
+                    ]
+                }
+            ]
+        }
+    ],
+    "_parse_type": "txt",
+    "_version_name": "0.6.1"
+}
+```

third_party/MinerU/magic-pdf.template.json

@@ -0,0 +1,54 @@
+{
+    "bucket_info":{
+        "bucket-name-1":["ak", "sk", "endpoint"],
+        "bucket-name-2":["ak", "sk", "endpoint"]
+    },
+    "models-dir":"/tmp/models",
+    "layoutreader-model-dir":"/tmp/layoutreader",
+    "device-mode":"cpu",
+    "layout-config": {
+        "model": "doclayout_yolo"
+    },
+    "formula-config": {
+        "mfd_model": "yolo_v8_mfd",
+        "mfr_model": "unimernet_small",
+        "enable": true
+    },
+    "table-config": {
+        "model": "rapid_table",
+        "sub_model": "slanet_plus",
+        "enable": true,
+        "max_time": 400
+    },
+    "latex-delimiter-config": {
+        "display": {
+            "left": "$$",
+            "right": "$$"
+        },
+        "inline": {
+            "left": "$",
+            "right": "$"
+        }
+    },
+    "llm-aided-config": {
+        "formula_aided": {
+            "api_key": "your_api_key",
+            "base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1",
+            "model": "qwen2.5-7b-instruct",
+            "enable": false
+        },
+        "text_aided": {
+            "api_key": "your_api_key",
+            "base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1",
+            "model": "qwen2.5-7b-instruct",
+            "enable": false
+        },
+        "title_aided": {
+            "api_key": "your_api_key",
+            "base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1",
+            "model": "qwen2.5-32b-instruct",
+            "enable": false
+        }
+    },
+    "config_version": "1.2.1"
+}

third_party/MinerU/magic_pdf.egg-info/PKG-INFO

@@ -0,0 +1,674 @@
+Metadata-Version: 2.4
+Name: magic_pdf
+Version: 1.3.11
+Summary: A practical tool for converting PDF to Markdown
+License: AGPL-3.0
+Project-URL: Home, https://mineru.net/
+Project-URL: Repository, https://github.com/opendatalab/MinerU
+Keywords: magic-pdf, mineru, MinerU, convert, pdf, markdown
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Requires-Python: >=3.10,<3.14
+Description-Content-Type: text/markdown
+License-File: LICENSE.md
+Requires-Dist: boto3>=1.28.43
+Requires-Dist: Brotli>=1.1.0
+Requires-Dist: click>=8.1.7
+Requires-Dist: fast-langdetect<0.3.0,>=0.2.3
+Requires-Dist: loguru>=0.6.0
+Requires-Dist: numpy>=1.21.6
+Requires-Dist: pydantic<2.11,>=2.7.2
+Requires-Dist: PyMuPDF<1.25.0,>=1.24.9
+Requires-Dist: scikit-learn>=1.0.2
+Requires-Dist: torch!=2.5.0,!=2.5.1,<3,>=2.2.2
+Requires-Dist: torchvision
+Requires-Dist: transformers!=4.51.0,<5.0.0,>=4.49.0
+Requires-Dist: pdfminer.six==20250506
+Requires-Dist: tqdm>=4.67.1
+Provides-Extra: lite
+Requires-Dist: paddleocr==2.7.3; extra == "lite"
+Requires-Dist: paddlepaddle==3.0.0b1; platform_system == "Linux" and extra == "lite"
+Requires-Dist: paddlepaddle==2.6.1; (platform_system == "Windows" or platform_system == "Darwin") and extra == "lite"
+Provides-Extra: full
+Requires-Dist: matplotlib<4,>=3.10; extra == "full"
+Requires-Dist: ultralytics<9,>=8.3.48; extra == "full"
+Requires-Dist: doclayout_yolo==0.0.2b1; extra == "full"
+Requires-Dist: dill<1,>=0.3.8; extra == "full"
+Requires-Dist: rapid_table<2.0.0,>=1.0.5; extra == "full"
+Requires-Dist: PyYAML<7,>=6.0.2; extra == "full"
+Requires-Dist: ftfy<7,>=6.3.1; extra == "full"
+Requires-Dist: openai<2,>=1.70.0; extra == "full"
+Requires-Dist: shapely<3,>=2.0.7; extra == "full"
+Requires-Dist: pyclipper<2,>=1.3.0; extra == "full"
+Requires-Dist: omegaconf<3,>=2.3.0; extra == "full"
+Provides-Extra: full-old-linux
+Requires-Dist: matplotlib<=3.10.1,>=3.10; extra == "full-old-linux"
+Requires-Dist: ultralytics<=8.3.104,>=8.3.48; extra == "full-old-linux"
+Requires-Dist: doclayout_yolo==0.0.2b1; extra == "full-old-linux"
+Requires-Dist: dill==0.3.8; extra == "full-old-linux"
+Requires-Dist: PyYAML==6.0.2; extra == "full-old-linux"
+Requires-Dist: ftfy==6.3.1; extra == "full-old-linux"
+Requires-Dist: openai==1.71.0; extra == "full-old-linux"
+Requires-Dist: shapely==2.1.0; extra == "full-old-linux"
+Requires-Dist: pyclipper==1.3.0.post6; extra == "full-old-linux"
+Requires-Dist: omegaconf==2.3.0; extra == "full-old-linux"
+Requires-Dist: albumentations==1.4.20; extra == "full-old-linux"
+Requires-Dist: rapid_table==1.0.3; extra == "full-old-linux"
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: keywords
+Dynamic: license
+Dynamic: license-file
+Dynamic: project-url
+Dynamic: provides-extra
+Dynamic: requires-dist
+Dynamic: requires-python
+Dynamic: summary
+
+<div align="center" xmlns="http://www.w3.org/1999/html">
+<!-- logo -->
+<p align="center">
+  <img src="docs/images/MinerU-logo.png" width="300px" style="vertical-align:middle;">
+</p>
+
+<!-- icon -->
+
+[![stars](https://img.shields.io/github/stars/opendatalab/MinerU.svg)](https://github.com/opendatalab/MinerU)
+[![forks](https://img.shields.io/github/forks/opendatalab/MinerU.svg)](https://github.com/opendatalab/MinerU)
+[![open issues](https://img.shields.io/github/issues-raw/opendatalab/MinerU)](https://github.com/opendatalab/MinerU/issues)
+[![issue resolution](https://img.shields.io/github/issues-closed-raw/opendatalab/MinerU)](https://github.com/opendatalab/MinerU/issues)
+[![PyPI version](https://img.shields.io/pypi/v/magic-pdf)](https://pypi.org/project/magic-pdf/)
+[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/magic-pdf)](https://pypi.org/project/magic-pdf/)
+[![Downloads](https://static.pepy.tech/badge/magic-pdf)](https://pepy.tech/project/magic-pdf)
+[![Downloads](https://static.pepy.tech/badge/magic-pdf/month)](https://pepy.tech/project/magic-pdf)
+
+[![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://mineru.net/OpenSourceTools/Extractor?source=github)
+[![HuggingFace](https://img.shields.io/badge/Demo_on_HuggingFace-yellow.svg?logo=data:image/png;base64,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&labelColor=white)](https://huggingface.co/spaces/opendatalab/MinerU)
+[![ModelScope](https://img.shields.io/badge/Demo_on_ModelScope-purple?logo=data:image/svg+xml;base64,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&labelColor=white)](https://www.modelscope.cn/studios/OpenDataLab/MinerU)
+[![Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/gist/myhloli/3b3a00a4a0a61577b6c30f989092d20d/mineru_demo.ipynb)
+[![Paper](https://img.shields.io/badge/Paper-arXiv-green)](https://arxiv.org/abs/2409.18839)
+
+
+<a href="https://trendshift.io/repositories/11174" target="_blank"><img src="https://trendshift.io/api/badge/repositories/11174" alt="opendatalab%2FMinerU | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+
+<!-- language -->
+
+[English](README.md) | [简体中文](README_zh-CN.md)
+
+<!-- hot link -->
+
+<p align="center">
+<a href="https://github.com/opendatalab/PDF-Extract-Kit">PDF-Extract-Kit: High-Quality PDF Extraction Toolkit</a>🔥🔥🔥
+<br>
+<br>
+<a href="https://mineru.net/client?source=github">
+Easier to use: Just grab MinerU Desktop. No coding, no login, just a simple interface and smooth interactions. Enjoy it without any fuss!</a>🚀🚀🚀
+
+</p>
+
+<!-- join us -->
+
+<p align="center">
+    👋 join us on <a href="https://discord.gg/Tdedn9GTXq" target="_blank">Discord</a> and <a href="http://mineru.space/s/V85Yl" target="_blank">WeChat</a>
+</p>
+
+</div>
+
+# Changelog
+- 2025/04/29 1.3.10 Released
+  - Support for custom formula delimiters can be achieved by modifying the `latex-delimiter-config` item in the `magic-pdf.json` file under the user directory.
+- 2025/04/27 1.3.9 Released  
+  - Optimized the formula parsing function to improve the success rate of formula rendering
+- 2025/04/23 1.3.8 Released
+  - The default `ocr` model (`ch`) has been updated to `PP-OCRv4_server_rec_doc` (model update required)
+    - `PP-OCRv4_server_rec_doc` is trained on a mix of more Chinese document data and PP-OCR training data, enhancing recognition capabilities for some traditional Chinese characters, Japanese, and special characters. It supports over 15,000 recognizable characters, improving text recognition in documents while also boosting general text recognition.
+    - [Performance comparison between PP-OCRv4_server_rec_doc, PP-OCRv4_server_rec, and PP-OCRv4_mobile_rec](https://paddlepaddle.github.io/PaddleX/latest/en/module_usage/tutorials/ocr_modules/text_recognition.html#ii-supported-model-list)
+    - Verified results show that the `PP-OCRv4_server_rec_doc` model significantly improves accuracy in both single-language (`Chinese`, `English`, `Japanese`, `Traditional Chinese`) and mixed-language scenarios, with speed comparable to `PP-OCRv4_server_rec`, making it suitable for most use cases.
+    - In a small number of pure English scenarios, the `PP-OCRv4_server_rec_doc` model may encounter word concatenation issues, whereas `PP-OCRv4_server_rec` performs better in such cases. Therefore, we have retained the `PP-OCRv4_server_rec` model, which users can invoke by passing the parameter `lang='ch_server'`(python api) or `--lang ch_server`(cli).
+- 2025/04/22 1.3.7 Released
+  - Fixed the issue where the `lang` parameter was ineffective during table parsing model initialization.
+  - Fixed the significant slowdown in OCR and table parsing speed in `cpu` mode.
+- 2025/04/16 1.3.4 Released
+  - Slightly improved the speed of OCR detection by removing some unused blocks.
+  - Fixed page-level sorting errors caused by footnotes in certain cases.
+- 2025/04/12 1.3.2 released
+  - Fixed the issue of incompatible dependency package versions when installing in Python 3.13 environment on Windows systems.
+  - Optimized memory usage during batch inference.
+  - Improved the parsing effect of tables rotated by 90 degrees.
+  - Enhanced the parsing accuracy for large tables in financial report samples.
+  - Fixed the occasional word concatenation issue in English text areas when OCR language is not specified.(The model needs to be updated)
+- 2025/04/08 1.3.1 released, fixed some compatibility issues
+  - Supported Python 3.13
+  - Made the final adaptation for some outdated Linux systems (e.g., CentOS 7), and no further support will be guaranteed for subsequent versions. [Installation Instructions](https://github.com/opendatalab/MinerU/issues/1004)
+- 2025/04/03 1.3.0 released, in this version we made many optimizations and improvements:
+  - Installation and compatibility optimization
+    - By removing the use of `layoutlmv3` in layout, resolved compatibility issues caused by `detectron2`.
+    - Torch version compatibility extended to 2.2~2.6 (excluding 2.5).
+    - CUDA compatibility supports 11.8/12.4/12.6/12.8 (CUDA version determined by torch), resolving compatibility issues for some users with 50-series and H-series GPUs.
+    - Python compatible versions expanded to 3.10~3.12, solving the problem of automatic downgrade to 0.6.1 during installation in non-3.10 environments.
+    - Offline deployment process optimized; no internet connection required after successful deployment to download any model files.
+  - Performance optimization
+    - By supporting batch processing of multiple PDF files ([script example](demo/batch_demo.py)), improved parsing speed for small files in batches (compared to version 1.0.1, formula parsing speed increased by over 1400%, overall parsing speed increased by over 500%).
+    - Optimized loading and usage of the mfr model, reducing GPU memory usage and improving parsing speed (requires re-execution of the [model download process](docs/how_to_download_models_en.md) to obtain incremental updates of model files).
+    - Optimized GPU memory usage, requiring only a minimum of 6GB to run this project.
+    - Improved running speed on MPS devices.
+  - Parsing effect optimization
+    - Updated the mfr model to `unimernet(2503)`, solving the issue of lost line breaks in multi-line formulas.
+  - Usability Optimization
+    - By using `paddleocr2torch`, completely replaced the use of the `paddle` framework and `paddleocr` in the project, resolving conflicts between `paddle` and `torch`, as well as thread safety issues caused by the `paddle` framework.
+    - Added a real-time progress bar during the parsing process to accurately track progress, making the wait less painful.
+<details>
+<summary>2025/03/03 1.2.1 released</summary>
+<ul>
+  <li>Fixed the impact on punctuation marks during full-width to half-width conversion of letters and numbers</li>
+  <li>Fixed caption matching inaccuracies in certain scenarios</li>
+  <li>Fixed formula span loss issues in certain scenarios</li>
+</ul>
+</details>
+
+<details>
+<summary>2025/02/24 1.2.0 released</summary>
+<p>This version includes several fixes and improvements to enhance parsing efficiency and accuracy:</p>
+<ul>
+  <li><strong>Performance Optimization</strong>
+    <ul>
+      <li>Increased classification speed for PDF documents in auto mode.</li>
+    </ul>
+  </li>
+  <li><strong>Parsing Optimization</strong>
+    <ul>
+      <li>Improved parsing logic for documents containing watermarks, significantly enhancing the parsing results for such documents.</li>
+      <li>Enhanced the matching logic for multiple images/tables and captions within a single page, improving the accuracy of image-text matching in complex layouts.</li>
+    </ul>
+  </li>
+  <li><strong>Bug Fixes</strong>
+    <ul>
+      <li>Fixed an issue where image/table spans were incorrectly filled into text blocks under certain conditions.</li>
+      <li>Resolved an issue where title blocks were empty in some cases.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2025/01/22 1.1.0 released</summary>
+<p>In this version we have focused on improving parsing accuracy and efficiency:</p>
+<ul>
+  <li><strong>Model capability upgrade</strong> (requires re-executing the <a href="https://github.com/opendatalab/MinerU/blob/master/docs/how_to_download_models_en.md">model download process</a> to obtain incremental updates of model files)
+    <ul>
+      <li>The layout recognition model has been upgraded to the latest <code>doclayout_yolo(2501)</code> model, improving layout recognition accuracy.</li>
+      <li>The formula parsing model has been upgraded to the latest <code>unimernet(2501)</code> model, improving formula recognition accuracy.</li>
+    </ul>
+  </li>
+  <li><strong>Performance optimization</strong>
+    <ul>
+      <li>On devices that meet certain configuration requirements (16GB+ VRAM), by optimizing resource usage and restructuring the processing pipeline, overall parsing speed has been increased by more than 50%.</li>
+    </ul>
+  </li>
+  <li><strong>Parsing effect optimization</strong>
+    <ul>
+      <li>Added a new heading classification feature (testing version, enabled by default) to the online demo (<a href="https://mineru.net/OpenSourceTools/Extractor">mineru.net</a>/<a href="https://huggingface.co/spaces/opendatalab/MinerU">huggingface</a>/<a href="https://www.modelscope.cn/studios/OpenDataLab/MinerU">modelscope</a>), which supports hierarchical classification of headings, thereby enhancing document structuring.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2025/01/10 1.0.1 released</summary>
+<p>This is our first official release, where we have introduced a completely new API interface and enhanced compatibility through extensive refactoring, as well as a brand new automatic language identification feature:</p>
+<ul>
+  <li><strong>New API Interface</strong>
+    <ul>
+      <li>For the data-side API, we have introduced the Dataset class, designed to provide a robust and flexible data processing framework. This framework currently supports a variety of document formats, including images (.jpg and .png), PDFs, Word documents (.doc and .docx), and PowerPoint presentations (.ppt and .pptx). It ensures effective support for data processing tasks ranging from simple to complex.</li>
+      <li>For the user-side API, we have meticulously designed the MinerU processing workflow as a series of composable Stages. Each Stage represents a specific processing step, allowing users to define new Stages according to their needs and creatively combine these stages to customize their data processing workflows.</li>
+    </ul>
+  </li>
+  <li><strong>Enhanced Compatibility</strong>
+    <ul>
+      <li>By optimizing the dependency environment and configuration items, we ensure stable and efficient operation on ARM architecture Linux systems.</li>
+      <li>We have deeply integrated with Huawei Ascend NPU acceleration, providing autonomous and controllable high-performance computing capabilities. This supports the localization and development of AI application platforms in China. <a href="https://github.com/opendatalab/MinerU/blob/master/docs/README_Ascend_NPU_Acceleration_zh_CN.md">Ascend NPU Acceleration</a></li>
+    </ul>
+  </li>
+  <li><strong>Automatic Language Identification</strong>
+    <ul>
+      <li>By introducing a new language recognition model, setting the <code>lang</code> configuration to <code>auto</code> during document parsing will automatically select the appropriate OCR language model, improving the accuracy of scanned document parsing.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2024/11/22 0.10.0 released</summary>
+<p>Introducing hybrid OCR text extraction capabilities:</p>
+<ul>
+  <li>Significantly improved parsing performance in complex text distribution scenarios such as dense formulas, irregular span regions, and text represented by images.</li>
+  <li>Combines the dual advantages of accurate content extraction and faster speed in text mode, and more precise span/line region recognition in OCR mode.</li>
+</ul>
+</details>
+
+<details>
+<summary>2024/11/15 0.9.3 released</summary>
+<p>Integrated <a href="https://github.com/RapidAI/RapidTable">RapidTable</a> for table recognition, improving single-table parsing speed by more than 10 times, with higher accuracy and lower GPU memory usage.</p>
+</details>
+
+<details>
+<summary>2024/11/06 0.9.2 released</summary>
+<p>Integrated the <a href="https://huggingface.co/U4R/StructTable-InternVL2-1B">StructTable-InternVL2-1B</a> model for table recognition functionality.</p>
+</details>
+
+<details>
+<summary>2024/10/31 0.9.0 released</summary>
+<p>This is a major new version with extensive code refactoring, addressing numerous issues, improving performance, reducing hardware requirements, and enhancing usability:</p>
+<ul>
+  <li>Refactored the sorting module code to use <a href="https://github.com/ppaanngggg/layoutreader">layoutreader</a> for reading order sorting, ensuring high accuracy in various layouts.</li>
+  <li>Refactored the paragraph concatenation module to achieve good results in cross-column, cross-page, cross-figure, and cross-table scenarios.</li>
+  <li>Refactored the list and table of contents recognition functions, significantly improving the accuracy of list blocks and table of contents blocks, as well as the parsing of corresponding text paragraphs.</li>
+  <li>Refactored the matching logic for figures, tables, and descriptive text, greatly enhancing the accuracy of matching captions and footnotes to figures and tables, and reducing the loss rate of descriptive text to near zero.</li>
+  <li>Added multi-language support for OCR, supporting detection and recognition of 84 languages. For the list of supported languages, see <a href="https://paddlepaddle.github.io/PaddleOCR/latest/en/ppocr/blog/multi_languages.html#5-support-languages-and-abbreviations">OCR Language Support List</a>.</li>
+  <li>Added memory recycling logic and other memory optimization measures, significantly reducing memory usage. The memory requirement for enabling all acceleration features except table acceleration (layout/formula/OCR) has been reduced from 16GB to 8GB, and the memory requirement for enabling all acceleration features has been reduced from 24GB to 10GB.</li>
+  <li>Optimized configuration file feature switches, adding an independent formula detection switch to significantly improve speed and parsing results when formula detection is not needed.</li>
+  <li>Integrated <a href="https://github.com/opendatalab/PDF-Extract-Kit">PDF-Extract-Kit 1.0</a>:
+    <ul>
+      <li>Added the self-developed <code>doclayout_yolo</code> model, which speeds up processing by more than 10 times compared to the original solution while maintaining similar parsing effects, and can be freely switched with <code>layoutlmv3</code> via the configuration file.</li>
+      <li>Upgraded formula parsing to <code>unimernet 0.2.1</code>, improving formula parsing accuracy while significantly reducing memory usage.</li>
+      <li>Due to the repository change for <code>PDF-Extract-Kit 1.0</code>, you need to re-download the model. Please refer to <a href="https://github.com/opendatalab/MinerU/blob/master/docs/how_to_download_models_en.md">How to Download Models</a> for detailed steps.</li>
+    </ul>
+  </li>
+</ul>
+</details>
+
+<details>
+<summary>2024/09/27 Version 0.8.1 released</summary>
+<p>Fixed some bugs, and providing a <a href="https://github.com/opendatalab/MinerU/blob/master/projects/web_demo/README.md">localized deployment version</a> of the <a href="https://opendatalab.com/OpenSourceTools/Extractor/PDF/">online demo</a> and the <a href="https://github.com/opendatalab/MinerU/blob/master/projects/web/README.md">front-end interface</a>.</p>
+</details>
+
+<details>
+<summary>2024/09/09 Version 0.8.0 released</summary>
+<p>Supporting fast deployment with Dockerfile, and launching demos on Huggingface and Modelscope.</p>
+</details>
+
+<details>
+<summary>2024/08/30 Version 0.7.1 released</summary>
+<p>Add paddle tablemaster table recognition option</p>
+</details>
+
+<details>
+<summary>2024/08/09 Version 0.7.0b1 released</summary>
+<p>Simplified installation process, added table recognition functionality</p>
+</details>
+
+<details>
+<summary>2024/08/01 Version 0.6.2b1 released</summary>
+<p>Optimized dependency conflict issues and installation documentation</p>
+</details>
+
+<details>
+<summary>2024/07/05 Initial open-source release</summary>
+</details>
+
+<!-- TABLE OF CONTENT -->
+
+<details open="open">
+  <summary><h2 style="display: inline-block">Table of Contents</h2></summary>
+  <ol>
+    <li>
+      <a href="#mineru">MinerU</a>
+      <ul>
+        <li><a href="#project-introduction">Project Introduction</a></li>
+        <li><a href="#key-features">Key Features</a></li>
+        <li><a href="#quick-start">Quick Start</a>
+            <ul>
+            <li><a href="#online-demo">Online Demo</a></li>
+            <li><a href="#quick-cpu-demo">Quick CPU Demo</a></li>
+            <li><a href="#using-gpu">Using GPU</a></li>
+            <li><a href="#using-npu">Using NPU</a></li>
+            </ul>
+        </li>
+        <li><a href="#usage">Usage</a>
+            <ul>
+            <li><a href="#command-line">Command Line</a></li>
+            <li><a href="#api">API</a></li>
+            <li><a href="#deploy-derived-projects">Deploy Derived Projects</a></li>
+            <li><a href="#development-guide">Development Guide</a></li>
+            </ul>
+        </li>
+      </ul>
+    </li>
+    <li><a href="#todo">TODO</a></li>
+    <li><a href="#known-issues">Known Issues</a></li>
+    <li><a href="#faq">FAQ</a></li>
+    <li><a href="#all-thanks-to-our-contributors">All Thanks To Our Contributors</a></li>
+    <li><a href="#license-information">License Information</a></li>
+    <li><a href="#acknowledgments">Acknowledgments</a></li>
+    <li><a href="#citation">Citation</a></li>
+    <li><a href="#star-history">Star History</a></li>
+    <li><a href="#magic-doc">Magic-doc</a></li>
+    <li><a href="#magic-html">Magic-html</a></li>
+    <li><a href="#links">Links</a></li>
+  </ol>
+</details>
+
+# MinerU
+
+## Project Introduction
+
+MinerU is a tool that converts PDFs into machine-readable formats (e.g., markdown, JSON), allowing for easy extraction into any format.
+MinerU was born during the pre-training process of [InternLM](https://github.com/InternLM/InternLM). We focus on solving symbol conversion issues in scientific literature and hope to contribute to technological development in the era of large models.
+Compared to well-known commercial products, MinerU is still young. If you encounter any issues or if the results are not as expected, please submit an issue on [issue](https://github.com/opendatalab/MinerU/issues) and **attach the relevant PDF**.
+
+https://github.com/user-attachments/assets/4bea02c9-6d54-4cd6-97ed-dff14340982c
+
+## Key Features
+
+- Remove headers, footers, footnotes, page numbers, etc., to ensure semantic coherence.
+- Output text in human-readable order, suitable for single-column, multi-column, and complex layouts.
+- Preserve the structure of the original document, including headings, paragraphs, lists, etc.
+- Extract images, image descriptions, tables, table titles, and footnotes.
+- Automatically recognize and convert formulas in the document to LaTeX format.
+- Automatically recognize and convert tables in the document to HTML format.
+- Automatically detect scanned PDFs and garbled PDFs and enable OCR functionality.
+- OCR supports detection and recognition of 84 languages.
+- Supports multiple output formats, such as multimodal and NLP Markdown, JSON sorted by reading order, and rich intermediate formats.
+- Supports various visualization results, including layout visualization and span visualization, for efficient confirmation of output quality.
+- Supports running in a pure CPU environment, and also supports GPU(CUDA)/NPU(CANN)/MPS acceleration
+- Compatible with Windows, Linux, and Mac platforms.
+
+## Quick Start
+
+If you encounter any installation issues, please first consult the <a href="#faq">FAQ</a>. </br>
+If the parsing results are not as expected, refer to the <a href="#known-issues">Known Issues</a>. </br>
+There are three different ways to experience MinerU:
+
+- [Online Demo (No Installation Required)](#online-demo)
+- [Quick CPU Demo (Windows, Linux, Mac)](#quick-cpu-demo)
+- Accelerate inference by using CUDA/CANN/MPS
+  - [Linux/Windows + CUDA](#Using-GPU)
+  - [Linux + CANN](#using-npu)
+  - [MacOS + MPS](#using-mps)
+
+> [!WARNING]
+> **Pre-installation Notice—Hardware and Software Environment Support**
+>
+> To ensure the stability and reliability of the project, we only optimize and test for specific hardware and software environments during development. This ensures that users deploying and running the project on recommended system configurations will get the best performance with the fewest compatibility issues.
+>
+> By focusing resources on the mainline environment, our team can more efficiently resolve potential bugs and develop new features.
+>
+> In non-mainline environments, due to the diversity of hardware and software configurations, as well as third-party dependency compatibility issues, we cannot guarantee 100% project availability. Therefore, for users who wish to use this project in non-recommended environments, we suggest carefully reading the documentation and FAQ first. Most issues already have corresponding solutions in the FAQ. We also encourage community feedback to help us gradually expand support.
+
+<table>
+    <tr>
+        <td colspan="3" rowspan="2">Operating System</td>
+    </tr>
+    <tr>
+        <td>Linux after 2019</td>
+        <td>Windows 10 / 11</td>
+        <td>macOS 11+</td>
+    </tr>
+    <tr>
+        <td colspan="3">CPU</td>
+        <td>x86_64 / arm64</td>
+        <td>x86_64(unsupported ARM Windows)</td>
+        <td>x86_64 / arm64</td>
+    </tr>
+    <tr>
+        <td colspan="3">Memory Requirements</td>
+        <td colspan="3">16GB or more, recommended 32GB+</td>
+    </tr>
+    <tr>
+        <td colspan="3">Storage Requirements</td>
+        <td colspan="3">20GB or more, with a preference for SSD</td>
+    </tr>
+    <tr>
+        <td colspan="3">Python Version</td>
+        <td colspan="3">3.10~3.13</td>
+    </tr>
+    <tr>
+        <td colspan="3">Nvidia Driver Version</td>
+        <td>latest (Proprietary Driver)</td>
+        <td>latest</td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td colspan="3">CUDA Environment</td>
+        <td colspan="2"><a href="https://pytorch.org/get-started/locally/">Refer to the PyTorch official website</a></td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td colspan="3">CANN Environment(NPU support)</td>
+        <td>8.0+(Ascend 910b)</td>
+        <td>None</td>
+        <td>None</td>
+    </tr>
+    <tr>
+        <td rowspan="2">GPU/MPS Hardware Support List</td>
+        <td colspan="2">GPU VRAM 6GB or more</td>
+        <td colspan="2">All GPUs with Tensor Cores produced from Volta(2017) onwards.<br>
+        More than 6GB VRAM </td>
+        <td rowspan="2">Apple silicon</td>
+    </tr>
+</table>
+
+### Online Demo
+
+Synced with dev branch updates:
+
+[![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://mineru.net/OpenSourceTools/Extractor?source=github)
+[![HuggingFace](https://img.shields.io/badge/Demo_on_HuggingFace-yellow.svg?logo=data:image/png;base64,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&labelColor=white)](https://huggingface.co/spaces/opendatalab/MinerU)
+[![ModelScope](https://img.shields.io/badge/Demo_on_ModelScope-purple?logo=data:image/svg+xml;base64,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&labelColor=white)](https://www.modelscope.cn/studios/OpenDataLab/MinerU)
+
+### Quick CPU Demo
+
+#### 1. Install magic-pdf
+
+```bash
+conda create -n mineru 'python=3.12' -y
+conda activate mineru
+pip install -U "magic-pdf[full]"
+```
+
+#### 2. Download model weight files
+
+Refer to [How to Download Model Files](docs/how_to_download_models_en.md) for detailed instructions.
+
+#### 3. Modify the Configuration File for Additional Configuration
+
+After completing the [2. Download model weight files](#2-download-model-weight-files) step, the script will automatically generate a `magic-pdf.json` file in the user directory and configure the default model path.
+You can find the `magic-pdf.json` file in your 【user directory】.
+
+> [!TIP]
+> The user directory for Windows is "C:\\Users\\username", for Linux it is "/home/username", and for macOS it is "/Users/username".
+
+You can modify certain configurations in this file to enable or disable features, such as table recognition:
+
+
+> [!NOTE]
+> If the following items are not present in the JSON, please manually add the required items and remove the comment content (standard JSON does not support comments).
+
+```json
+{
+    // other config
+    "layout-config": {
+        "model": "doclayout_yolo" 
+    },
+    "formula-config": {
+        "mfd_model": "yolo_v8_mfd",
+        "mfr_model": "unimernet_small",
+        "enable": true  // The formula recognition feature is enabled by default. If you need to disable it, please change the value here to "false".
+    },
+    "table-config": {
+        "model": "rapid_table", 
+        "sub_model": "slanet_plus",
+        "enable": true, // The table recognition feature is enabled by default. If you need to disable it, please change the value here to "false".
+        "max_time": 400
+    }
+}
+```
+
+### Using GPU
+
+If your device supports CUDA and meets the GPU requirements of the mainline environment, you can use GPU acceleration. Please select the appropriate guide based on your system:
+
+- [Ubuntu 22.04 LTS + GPU](docs/README_Ubuntu_CUDA_Acceleration_en_US.md)
+- [Windows 10/11 + GPU](docs/README_Windows_CUDA_Acceleration_en_US.md)
+- Quick Deployment with Docker
+> [!IMPORTANT]
+> Docker requires a GPU with at least 6GB of VRAM, and all acceleration features are enabled by default.
+>
+> Before running this Docker, you can use the following command to check if your device supports CUDA acceleration on Docker.
+> 
+> ```bash
+> docker run --rm --gpus=all nvidia/cuda:12.1.0-base-ubuntu22.04 nvidia-smi
+> ```
+  ```bash
+  wget https://github.com/opendatalab/MinerU/raw/master/docker/global/Dockerfile -O Dockerfile
+  docker build -t mineru:latest .
+  docker run -it --name mineru --gpus=all mineru:latest /bin/bash -c "echo 'source /opt/mineru_venv/bin/activate' >> ~/.bashrc && exec bash"
+  magic-pdf --help
+  ```
+
+### Using NPU
+
+If your device has NPU acceleration hardware, you can follow the tutorial below to use NPU acceleration:
+
+[Ascend NPU Acceleration](docs/README_Ascend_NPU_Acceleration_zh_CN.md)
+
+### Using MPS
+
+If your device uses Apple silicon chips, you can enable MPS acceleration for your tasks.
+
+You can enable MPS acceleration by setting the `device-mode` parameter to `mps` in the `magic-pdf.json` configuration file.
+
+```json
+{
+    // other config
+    "device-mode": "mps"
+}
+```
+
+
+## Usage
+
+### Command Line
+
+[Using MinerU via Command Line](https://mineru.readthedocs.io/en/latest/user_guide/usage/command_line.html)
+
+> [!TIP]
+> For more information about the output files, please refer to the [Output File Description](docs/output_file_en_us.md).
+
+### API
+
+[Using MinerU via Python API](https://mineru.readthedocs.io/en/latest/user_guide/usage/api.html)
+
+
+### Deploy Derived Projects
+
+Derived projects include secondary development projects based on MinerU by project developers and community developers,  
+such as application interfaces based on Gradio, RAG based on llama, web demos similar to the official website, lightweight multi-GPU load balancing client/server ends, etc.
+These projects may offer more features and a better user experience.  
+For specific deployment methods, please refer to the [Derived Project README](projects/README.md)
+
+
+### Development Guide
+
+TODO
+
+# TODO
+
+- [x] Reading order based on the model  
+- [x] Recognition of `index` and `list` in the main text  
+- [x] Table recognition
+- [x] Heading Classification
+- [ ] Code block recognition in the main text
+- [ ] [Chemical formula recognition](docs/chemical_knowledge_introduction/introduction.pdf)
+- [ ] Geometric shape recognition
+
+# Known Issues
+
+- Reading order is determined by the model based on the spatial distribution of readable content, and may be out of order in some areas under extremely complex layouts.
+- Vertical text is not supported.
+- Tables of contents and lists are recognized through rules, and some uncommon list formats may not be recognized.
+- Code blocks are not yet supported in the layout model.
+- Comic books, art albums, primary school textbooks, and exercises cannot be parsed well.
+- Table recognition may result in row/column recognition errors in complex tables.
+- OCR recognition may produce inaccurate characters in PDFs of lesser-known languages (e.g., diacritical marks in Latin script, easily confused characters in Arabic script).
+- Some formulas may not render correctly in Markdown.
+
+# FAQ
+
+[FAQ in Chinese](docs/FAQ_zh_cn.md)
+
+[FAQ in English](docs/FAQ_en_us.md)
+
+# All Thanks To Our Contributors
+
+<a href="https://github.com/opendatalab/MinerU/graphs/contributors">
+  <img src="https://contrib.rocks/image?repo=opendatalab/MinerU" />
+</a>
+
+# License Information
+
+[LICENSE.md](LICENSE.md)
+
+This project currently uses PyMuPDF to achieve advanced functionality. However, since it adheres to the AGPL license, it may impose restrictions on certain usage scenarios. In future iterations, we plan to explore and replace it with a more permissive PDF processing library to enhance user-friendliness and flexibility.
+
+# Acknowledgments
+
+- [PDF-Extract-Kit](https://github.com/opendatalab/PDF-Extract-Kit)
+- [DocLayout-YOLO](https://github.com/opendatalab/DocLayout-YOLO)
+- [StructEqTable](https://github.com/UniModal4Reasoning/StructEqTable-Deploy)
+- [RapidTable](https://github.com/RapidAI/RapidTable)
+- [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)
+- [RapidOCR](https://github.com/RapidAI/RapidOCR)
+- [PaddleOCR2Pytorch](https://github.com/frotms/PaddleOCR2Pytorch)
+- [PyMuPDF](https://github.com/pymupdf/PyMuPDF)
+- [layoutreader](https://github.com/ppaanngggg/layoutreader)
+- [fast-langdetect](https://github.com/LlmKira/fast-langdetect)
+- [pdfminer.six](https://github.com/pdfminer/pdfminer.six)
+
+# Citation
+
+```bibtex
+@misc{wang2024mineruopensourcesolutionprecise,
+      title={MinerU: An Open-Source Solution for Precise Document Content Extraction}, 
+      author={Bin Wang and Chao Xu and Xiaomeng Zhao and Linke Ouyang and Fan Wu and Zhiyuan Zhao and Rui Xu and Kaiwen Liu and Yuan Qu and Fukai Shang and Bo Zhang and Liqun Wei and Zhihao Sui and Wei Li and Botian Shi and Yu Qiao and Dahua Lin and Conghui He},
+      year={2024},
+      eprint={2409.18839},
+      archivePrefix={arXiv},
+      primaryClass={cs.CV},
+      url={https://arxiv.org/abs/2409.18839}, 
+}
+
+@article{he2024opendatalab,
+  title={Opendatalab: Empowering general artificial intelligence with open datasets},
+  author={He, Conghui and Li, Wei and Jin, Zhenjiang and Xu, Chao and Wang, Bin and Lin, Dahua},
+  journal={arXiv preprint arXiv:2407.13773},
+  year={2024}
+}
+```
+
+# Star History
+
+<a>
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date&theme=dark" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date" />
+   <img alt="Star History Chart" src="https://api.star-history.com/svg?repos=opendatalab/MinerU&type=Date" />
+ </picture>
+</a>
+
+# Magic-doc
+
+[Magic-Doc](https://github.com/InternLM/magic-doc) Fast speed ppt/pptx/doc/docx/pdf extraction tool
+
+# Magic-html
+
+[Magic-HTML](https://github.com/opendatalab/magic-html) Mixed web page extraction tool
+
+# Links
+
+- [LabelU (A Lightweight Multi-modal Data Annotation Tool)](https://github.com/opendatalab/labelU)
+- [LabelLLM (An Open-source LLM Dialogue Annotation Platform)](https://github.com/opendatalab/LabelLLM)
+- [PDF-Extract-Kit (A Comprehensive Toolkit for High-Quality PDF Content Extraction)](https://github.com/opendatalab/PDF-Extract-Kit)

third_party/MinerU/magic_pdf.egg-info/SOURCES.txt

@@ -0,0 +1,208 @@
+LICENSE.md
+README.md
+setup.py
+magic_pdf/__init__.py
+magic_pdf/pdf_parse_union_core_v2.py
+magic_pdf.egg-info/PKG-INFO
+magic_pdf.egg-info/SOURCES.txt
+magic_pdf.egg-info/dependency_links.txt
+magic_pdf.egg-info/entry_points.txt
+magic_pdf.egg-info/not-zip-safe
+magic_pdf.egg-info/requires.txt
+magic_pdf.egg-info/top_level.txt
+magic_pdf/config/__init__.py
+magic_pdf/config/constants.py
+magic_pdf/config/drop_reason.py
+magic_pdf/config/drop_tag.py
+magic_pdf/config/enums.py
+magic_pdf/config/exceptions.py
+magic_pdf/config/make_content_config.py
+magic_pdf/config/model_block_type.py
+magic_pdf/config/ocr_content_type.py
+magic_pdf/data/__init__.py
+magic_pdf/data/batch_build_dataset.py
+magic_pdf/data/dataset.py
+magic_pdf/data/read_api.py
+magic_pdf/data/schemas.py
+magic_pdf/data/utils.py
+magic_pdf/data/data_reader_writer/__init__.py
+magic_pdf/data/data_reader_writer/base.py
+magic_pdf/data/data_reader_writer/filebase.py
+magic_pdf/data/data_reader_writer/multi_bucket_s3.py
+magic_pdf/data/data_reader_writer/s3.py
+magic_pdf/data/io/__init__.py
+magic_pdf/data/io/base.py
+magic_pdf/data/io/http.py
+magic_pdf/data/io/s3.py
+magic_pdf/dict2md/__init__.py
+magic_pdf/dict2md/ocr_mkcontent.py
+magic_pdf/filter/__init__.py
+magic_pdf/filter/pdf_classify_by_type.py
+magic_pdf/filter/pdf_meta_scan.py
+magic_pdf/integrations/__init__.py
+magic_pdf/integrations/rag/__init__.py
+magic_pdf/integrations/rag/api.py
+magic_pdf/integrations/rag/type.py
+magic_pdf/integrations/rag/utils.py
+magic_pdf/libs/__init__.py
+magic_pdf/libs/boxbase.py
+magic_pdf/libs/clean_memory.py
+magic_pdf/libs/commons.py
+magic_pdf/libs/config_reader.py
+magic_pdf/libs/convert_utils.py
+magic_pdf/libs/coordinate_transform.py
+magic_pdf/libs/draw_bbox.py
+magic_pdf/libs/hash_utils.py
+magic_pdf/libs/json_compressor.py
+magic_pdf/libs/language.py
+magic_pdf/libs/local_math.py
+magic_pdf/libs/markdown_utils.py
+magic_pdf/libs/path_utils.py
+magic_pdf/libs/pdf_check.py
+magic_pdf/libs/pdf_image_tools.py
+magic_pdf/libs/performance_stats.py
+magic_pdf/libs/safe_filename.py
+magic_pdf/libs/version.py
+magic_pdf/model/__init__.py
+magic_pdf/model/batch_analyze.py
+magic_pdf/model/doc_analyze_by_custom_model.py
+magic_pdf/model/magic_model.py
+magic_pdf/model/model_list.py
+magic_pdf/model/pdf_extract_kit.py
+magic_pdf/model/pp_structure_v2.py
+magic_pdf/model/sub_modules/__init__.py
+magic_pdf/model/sub_modules/model_init.py
+magic_pdf/model/sub_modules/model_utils.py
+magic_pdf/model/sub_modules/language_detection/__init__.py
+magic_pdf/model/sub_modules/language_detection/utils.py
+magic_pdf/model/sub_modules/language_detection/yolov11/YOLOv11.py
+magic_pdf/model/sub_modules/language_detection/yolov11/__init__.py
+magic_pdf/model/sub_modules/layout/__init__.py
+magic_pdf/model/sub_modules/layout/doclayout_yolo/DocLayoutYOLO.py
+magic_pdf/model/sub_modules/layout/doclayout_yolo/__init__.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/__init__.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/backbone.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/beit.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/deit.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/model_init.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/rcnn_vl.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/visualizer.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/__init__.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/data/__init__.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/data/cord.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/data/data_collator.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/data/funsd.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/data/image_utils.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/data/xfund.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/models/__init__.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/models/layoutlmv3/__init__.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/models/layoutlmv3/configuration_layoutlmv3.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/models/layoutlmv3/modeling_layoutlmv3.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/models/layoutlmv3/tokenization_layoutlmv3.py
+magic_pdf/model/sub_modules/layout/layoutlmv3/layoutlmft/models/layoutlmv3/tokenization_layoutlmv3_fast.py
+magic_pdf/model/sub_modules/mfd/__init__.py
+magic_pdf/model/sub_modules/mfd/yolov8/YOLOv8.py
+magic_pdf/model/sub_modules/mfd/yolov8/__init__.py
+magic_pdf/model/sub_modules/mfr/__init__.py
+magic_pdf/model/sub_modules/mfr/unimernet/Unimernet.py
+magic_pdf/model/sub_modules/mfr/unimernet/__init__.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/__init__.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/modeling_unimernet.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_mbart/__init__.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_mbart/configuration_unimer_mbart.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_mbart/modeling_unimer_mbart.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_mbart/tokenization_unimer_mbart.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_swin/__init__.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_swin/configuration_unimer_swin.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_swin/image_processing_unimer_swin.py
+magic_pdf/model/sub_modules/mfr/unimernet/unimernet_hf/unimer_swin/modeling_unimer_swin.py
+magic_pdf/model/sub_modules/ocr/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/ocr_utils.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorch_paddle.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/base_ocr_v20.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/data/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/data/imaug/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/data/imaug/operators.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/common.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/architectures/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/architectures/base_model.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/det_mobilenet_v3.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/rec_hgnet.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/rec_lcnetv3.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/rec_mobilenet_v3.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/rec_mv1_enhance.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/backbones/rec_svtrnet.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/heads/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/heads/cls_head.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/heads/det_db_head.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/heads/rec_ctc_head.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/heads/rec_multi_head.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/necks/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/necks/db_fpn.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/necks/intracl.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/modeling/necks/rnn.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/postprocess/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/postprocess/cls_postprocess.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/postprocess/db_postprocess.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/postprocess/rec_postprocess.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/arch_config.yaml
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/models_config.yml
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/arabic_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/chinese_cht_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/cyrillic_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/devanagari_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/en_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/japan_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/ka_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/korean_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/latin_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/ppocr_keys_v1.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/ppocrv4_doc_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/ta_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/pytorchocr/utils/resources/dict/te_dict.txt
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/infer/__init__.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/infer/predict_cls.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/infer/predict_det.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/infer/predict_rec.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/infer/predict_system.py
+magic_pdf/model/sub_modules/ocr/paddleocr2pytorch/tools/infer/pytorchocr_utility.py
+magic_pdf/model/sub_modules/reading_oreder/__init__.py
+magic_pdf/model/sub_modules/reading_oreder/layoutreader/__init__.py
+magic_pdf/model/sub_modules/reading_oreder/layoutreader/helpers.py
+magic_pdf/model/sub_modules/reading_oreder/layoutreader/xycut.py
+magic_pdf/model/sub_modules/table/__init__.py
+magic_pdf/model/sub_modules/table/table_utils.py
+magic_pdf/model/sub_modules/table/rapidtable/__init__.py
+magic_pdf/model/sub_modules/table/rapidtable/rapid_table.py
+magic_pdf/operators/__init__.py
+magic_pdf/operators/models.py
+magic_pdf/operators/pipes.py
+magic_pdf/post_proc/__init__.py
+magic_pdf/post_proc/llm_aided.py
+magic_pdf/post_proc/para_split_v3.py
+magic_pdf/pre_proc/__init__.py
+magic_pdf/pre_proc/construct_page_dict.py
+magic_pdf/pre_proc/cut_image.py
+magic_pdf/pre_proc/ocr_detect_all_bboxes.py
+magic_pdf/pre_proc/ocr_dict_merge.py
+magic_pdf/pre_proc/ocr_span_list_modify.py
+magic_pdf/pre_proc/remove_bbox_overlap.py
+magic_pdf/resources/fasttext-langdetect/lid.176.ftz
+magic_pdf/resources/model_config/model_configs.yaml
+magic_pdf/resources/slanet_plus/slanet-plus.onnx
+magic_pdf/resources/yolov11-langdetect/yolo_v11_ft.pt
+magic_pdf/spark/__init__.py
+magic_pdf/spark/spark_api.py
+magic_pdf/tools/__init__.py
+magic_pdf/tools/cli.py
+magic_pdf/tools/cli_dev.py
+magic_pdf/tools/common.py
+magic_pdf/utils/__init__.py
+magic_pdf/utils/annotations.py
+magic_pdf/utils/office_to_pdf.py

third_party/MinerU/magic_pdf.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

third_party/MinerU/magic_pdf.egg-info/entry_points.txt

@@ -0,0 +1,3 @@
+[console_scripts]
+magic-pdf = magic_pdf.tools.cli:cli
+magic-pdf-dev = magic_pdf.tools.cli_dev:cli

third_party/MinerU/magic_pdf.egg-info/not-zip-safe

@@ -0,0 +1 @@
+

third_party/MinerU/magic_pdf.egg-info/requires.txt

@@ -0,0 +1,50 @@
+boto3>=1.28.43
+Brotli>=1.1.0
+click>=8.1.7
+fast-langdetect<0.3.0,>=0.2.3
+loguru>=0.6.0
+numpy>=1.21.6
+pydantic<2.11,>=2.7.2
+PyMuPDF<1.25.0,>=1.24.9
+scikit-learn>=1.0.2
+torch!=2.5.0,!=2.5.1,<3,>=2.2.2
+torchvision
+transformers!=4.51.0,<5.0.0,>=4.49.0
+pdfminer.six==20250506
+tqdm>=4.67.1
+
+[full]
+matplotlib<4,>=3.10
+ultralytics<9,>=8.3.48
+doclayout_yolo==0.0.2b1
+dill<1,>=0.3.8
+rapid_table<2.0.0,>=1.0.5
+PyYAML<7,>=6.0.2
+ftfy<7,>=6.3.1
+openai<2,>=1.70.0
+shapely<3,>=2.0.7
+pyclipper<2,>=1.3.0
+omegaconf<3,>=2.3.0
+
+[full_old_linux]
+matplotlib<=3.10.1,>=3.10
+ultralytics<=8.3.104,>=8.3.48
+doclayout_yolo==0.0.2b1
+dill==0.3.8
+PyYAML==6.0.2
+ftfy==6.3.1
+openai==1.71.0
+shapely==2.1.0
+pyclipper==1.3.0.post6
+omegaconf==2.3.0
+albumentations==1.4.20
+rapid_table==1.0.3
+
+[lite]
+paddleocr==2.7.3
+
+[lite:platform_system == "Linux"]
+paddlepaddle==3.0.0b1
+
+[lite:platform_system == "Windows" or platform_system == "Darwin"]
+paddlepaddle==2.6.1

third_party/MinerU/magic_pdf.egg-info/top_level.txt

@@ -0,0 +1 @@
+magic_pdf

third_party/MinerU/magic_pdf/config/constants.py

@@ -0,0 +1,60 @@
+"""span维度自定义字段."""
+# span是否是跨页合并的
+CROSS_PAGE = 'cross_page'
+
+"""
+block维度自定义字段
+"""
+# block中lines是否被删除
+LINES_DELETED = 'lines_deleted'
+
+# table recognition max time default value
+TABLE_MAX_TIME_VALUE = 400
+
+# pp_table_result_max_length
+TABLE_MAX_LEN = 480
+
+# table master structure dict
+TABLE_MASTER_DICT = 'table_master_structure_dict.txt'
+
+# table master dir
+TABLE_MASTER_DIR = 'table_structure_tablemaster_infer/'
+
+# pp detect model dir
+DETECT_MODEL_DIR = 'ch_PP-OCRv4_det_infer'
+
+# pp rec model dir
+REC_MODEL_DIR = 'ch_PP-OCRv4_rec_infer'
+
+# pp rec char dict path
+REC_CHAR_DICT = 'ppocr_keys_v1.txt'
+
+# pp rec copy rec directory
+PP_REC_DIRECTORY = '.paddleocr/whl/rec/ch/ch_PP-OCRv4_rec_infer'
+
+# pp rec copy det directory
+PP_DET_DIRECTORY = '.paddleocr/whl/det/ch/ch_PP-OCRv4_det_infer'
+
+
+class MODEL_NAME:
+    # pp table structure algorithm
+    TABLE_MASTER = 'tablemaster'
+    # struct eqtable
+    STRUCT_EQTABLE = 'struct_eqtable'
+
+    DocLayout_YOLO = 'doclayout_yolo'
+
+    LAYOUTLMv3 = 'layoutlmv3'
+
+    YOLO_V8_MFD = 'yolo_v8_mfd'
+
+    UniMerNet_v2_Small = 'unimernet_small'
+
+    RAPID_TABLE = 'rapid_table'
+
+    YOLO_V11_LangDetect = 'yolo_v11n_langdetect'
+
+
+PARSE_TYPE_TXT = 'txt'
+PARSE_TYPE_OCR = 'ocr'
+

third_party/MinerU/magic_pdf/config/drop_reason.py

@@ -0,0 +1,35 @@
+class DropReason:
+    TEXT_BLCOK_HOR_OVERLAP = 'text_block_horizontal_overlap'  # 文字块有水平互相覆盖，导致无法准确定位文字顺序
+    USEFUL_BLOCK_HOR_OVERLAP = (
+        'useful_block_horizontal_overlap'  # 需保留的block水平覆盖
+    )
+    COMPLICATED_LAYOUT = 'complicated_layout'  # 复杂的布局，暂时不支持
+    TOO_MANY_LAYOUT_COLUMNS = 'too_many_layout_columns'  # 目前不支持分栏超过2列的
+    COLOR_BACKGROUND_TEXT_BOX = 'color_background_text_box'  # 含有带色块的PDF，色块会改变阅读顺序，目前不支持带底色文字块的PDF。
+    HIGH_COMPUTATIONAL_lOAD_BY_IMGS = (
+        'high_computational_load_by_imgs'  # 含特殊图片，计算量太大，从而丢弃
+    )
+    HIGH_COMPUTATIONAL_lOAD_BY_SVGS = (
+        'high_computational_load_by_svgs'  # 特殊的SVG图，计算量太大，从而丢弃
+    )
+    HIGH_COMPUTATIONAL_lOAD_BY_TOTAL_PAGES = 'high_computational_load_by_total_pages'  # 计算量超过负荷，当前方法下计算量消耗过大
+    MISS_DOC_LAYOUT_RESULT = 'missing doc_layout_result'  # 版面分析失败
+    Exception = '_exception'  # 解析中发生异常
+    ENCRYPTED = 'encrypted'  # PDF是加密的
+    EMPTY_PDF = 'total_page=0'  # PDF页面总数为0
+    NOT_IS_TEXT_PDF = 'not_is_text_pdf'  # 不是文字版PDF，无法直接解析
+    DENSE_SINGLE_LINE_BLOCK = 'dense_single_line_block'  # 无法清晰的分段
+    TITLE_DETECTION_FAILED = 'title_detection_failed'  # 探测标题失败
+    TITLE_LEVEL_FAILED = (
+        'title_level_failed'  # 分析标题级别失败（例如一级、二级、三级标题）
+    )
+    PARA_SPLIT_FAILED = 'para_split_failed'  # 识别段落失败
+    PARA_MERGE_FAILED = 'para_merge_failed'  # 段落合并失败
+    NOT_ALLOW_LANGUAGE = 'not_allow_language'  # 不支持的语种
+    SPECIAL_PDF = 'special_pdf'
+    PSEUDO_SINGLE_COLUMN = 'pseudo_single_column'  # 无法精确判断文字分栏
+    CAN_NOT_DETECT_PAGE_LAYOUT = 'can_not_detect_page_layout'  # 无法分析页面的版面
+    NEGATIVE_BBOX_AREA = 'negative_bbox_area'  # 缩放导致 bbox 面积为负
+    OVERLAP_BLOCKS_CAN_NOT_SEPARATION = (
+        'overlap_blocks_can_t_separation'  # 无法分离重叠的block
+    )

third_party/MinerU/magic_pdf/config/drop_tag.py

@@ -0,0 +1,19 @@
+
+COLOR_BG_HEADER_TXT_BLOCK = 'color_background_header_txt_block'
+PAGE_NO = 'page-no'  # 页码
+CONTENT_IN_FOOT_OR_HEADER = 'in-foot-header-area'  # 页眉页脚内的文本
+VERTICAL_TEXT = 'vertical-text'  # 垂直文本
+ROTATE_TEXT = 'rotate-text'  # 旋转文本
+EMPTY_SIDE_BLOCK = 'empty-side-block'  # 边缘上的空白没有任何内容的block
+ON_IMAGE_TEXT = 'on-image-text'  # 文本在图片上
+ON_TABLE_TEXT = 'on-table-text'  # 文本在表格上
+
+
+class DropTag:
+    PAGE_NUMBER = 'page_no'
+    HEADER = 'header'
+    FOOTER = 'footer'
+    FOOTNOTE = 'footnote'
+    NOT_IN_LAYOUT = 'not_in_layout'
+    SPAN_OVERLAP = 'span_overlap'
+    BLOCK_OVERLAP = 'block_overlap'

third_party/MinerU/magic_pdf/config/enums.py

@@ -0,0 +1,7 @@
+
+import enum
+
+
+class SupportedPdfParseMethod(enum.Enum):
+    OCR = 'ocr'
+    TXT = 'txt'

third_party/MinerU/magic_pdf/config/exceptions.py

@@ -0,0 +1,39 @@
+
+class FileNotExisted(Exception):
+
+    def __init__(self, path):
+        self.path = path
+
+    def __str__(self):
+        return f'File {self.path} does not exist.'
+
+
+class InvalidConfig(Exception):
+    def __init__(self, msg):
+        self.msg = msg
+
+    def __str__(self):
+        return f'Invalid config: {self.msg}'
+
+
+class InvalidParams(Exception):
+    def __init__(self, msg):
+        self.msg = msg
+
+    def __str__(self):
+        return f'Invalid params: {self.msg}'
+
+
+class EmptyData(Exception):
+    def __init__(self, msg):
+        self.msg = msg
+
+    def __str__(self):
+        return f'Empty data: {self.msg}'
+
+class CUDA_NOT_AVAILABLE(Exception):
+    def __init__(self, msg):
+        self.msg = msg
+
+    def __str__(self):
+        return f'CUDA not available: {self.msg}'

third_party/MinerU/magic_pdf/config/make_content_config.py

@@ -0,0 +1,11 @@
+class MakeMode:
+    MM_MD = 'mm_markdown'
+    NLP_MD = 'nlp_markdown'
+    STANDARD_FORMAT = 'standard_format'
+
+
+class DropMode:
+    WHOLE_PDF = 'whole_pdf'
+    SINGLE_PAGE = 'single_page'
+    NONE = 'none'
+    NONE_WITH_REASON = 'none_with_reason'

third_party/MinerU/magic_pdf/config/model_block_type.py

@@ -0,0 +1,10 @@
+from enum import Enum
+
+
+class ModelBlockTypeEnum(Enum):
+    TITLE = 0
+    PLAIN_TEXT = 1
+    ABANDON = 2
+    ISOLATE_FORMULA = 8
+    EMBEDDING = 13
+    ISOLATED = 14

third_party/MinerU/magic_pdf/config/ocr_content_type.py

@@ -0,0 +1,40 @@
+class ContentType:
+    Image = 'image'
+    Table = 'table'
+    Text = 'text'
+    InlineEquation = 'inline_equation'
+    InterlineEquation = 'interline_equation'
+
+
+class BlockType:
+    Image = 'image'
+    ImageBody = 'image_body'
+    ImageCaption = 'image_caption'
+    ImageFootnote = 'image_footnote'
+    Table = 'table'
+    TableBody = 'table_body'
+    TableCaption = 'table_caption'
+    TableFootnote = 'table_footnote'
+    Text = 'text'
+    Title = 'title'
+    InterlineEquation = 'interline_equation'
+    Footnote = 'footnote'
+    Discarded = 'discarded'
+    List = 'list'
+    Index = 'index'
+
+
+class CategoryId:
+    Title = 0
+    Text = 1
+    Abandon = 2
+    ImageBody = 3
+    ImageCaption = 4
+    TableBody = 5
+    TableCaption = 6
+    TableFootnote = 7
+    InterlineEquation_Layout = 8
+    InlineEquation = 13
+    InterlineEquation_YOLO = 14
+    OcrText = 15
+    ImageFootnote = 101

third_party/MinerU/magic_pdf/data/batch_build_dataset.py

@@ -0,0 +1,167 @@
+import concurrent.futures
+
+import fitz
+
+from magic_pdf.data.dataset import PymuDocDataset
+from magic_pdf.data.utils import fitz_doc_to_image  # PyMuPDF
+
+
+def partition_array_greedy(arr, k):
+    """Partition an array into k parts using a simple greedy approach.
+
+    Parameters:
+    -----------
+    arr : list
+        The input array of integers
+    k : int
+        Number of partitions to create
+
+    Returns:
+    --------
+    partitions : list of lists
+        The k partitions of the array
+    """
+    # Handle edge cases
+    if k <= 0:
+        raise ValueError('k must be a positive integer')
+    if k > len(arr):
+        k = len(arr)  # Adjust k if it's too large
+    if k == 1:
+        return [list(range(len(arr)))]
+    if k == len(arr):
+        return [[i] for i in range(len(arr))]
+
+    # Sort the array in descending order
+    sorted_indices = sorted(range(len(arr)), key=lambda i: arr[i][1], reverse=True)
+
+    # Initialize k empty partitions
+    partitions = [[] for _ in range(k)]
+    partition_sums = [0] * k
+
+    # Assign each element to the partition with the smallest current sum
+    for idx in sorted_indices:
+        # Find the partition with the smallest sum
+        min_sum_idx = partition_sums.index(min(partition_sums))
+
+        # Add the element to this partition
+        partitions[min_sum_idx].append(idx)  # Store the original index
+        partition_sums[min_sum_idx] += arr[idx][1]
+
+    return partitions
+
+
+def process_pdf_batch(pdf_jobs, idx):
+    """Process a batch of PDF pages using multiple threads.
+
+    Parameters:
+    -----------
+    pdf_jobs : list of tuples
+        List of (pdf_path, page_num) tuples
+    output_dir : str or None
+        Directory to save images to
+    num_threads : int
+        Number of threads to use
+    **kwargs :
+        Additional arguments for process_pdf_page
+
+    Returns:
+    --------
+    images : list
+        List of processed images
+    """
+    images = []
+
+    for pdf_path, _ in pdf_jobs:
+        doc = fitz.open(pdf_path)
+        tmp = []
+        for page_num in range(len(doc)):
+            page = doc[page_num]
+            tmp.append(fitz_doc_to_image(page))
+        images.append(tmp)
+    return (idx, images)
+
+
+def batch_build_dataset(pdf_paths, k, lang=None):
+    """Process multiple PDFs by partitioning them into k balanced parts and
+    processing each part in parallel.
+
+    Parameters:
+    -----------
+    pdf_paths : list
+        List of paths to PDF files
+    k : int
+        Number of partitions to create
+    output_dir : str or None
+        Directory to save images to
+    threads_per_worker : int
+        Number of threads to use per worker
+    **kwargs :
+        Additional arguments for process_pdf_page
+
+    Returns:
+    --------
+    all_images : list
+        List of all processed images
+    """
+
+    results = []
+    for pdf_path in pdf_paths:
+        with open(pdf_path, 'rb') as f:
+            pdf_bytes = f.read()
+        dataset = PymuDocDataset(pdf_bytes, lang=lang)
+        results.append(dataset)
+    return results
+
+
+    #
+    # # Get page counts for each PDF
+    # pdf_info = []
+    # total_pages = 0
+    #
+    # for pdf_path in pdf_paths:
+    #     try:
+    #         doc = fitz.open(pdf_path)
+    #         num_pages = len(doc)
+    #         pdf_info.append((pdf_path, num_pages))
+    #         total_pages += num_pages
+    #         doc.close()
+    #     except Exception as e:
+    #         print(f'Error opening {pdf_path}: {e}')
+    #
+    # # Partition the jobs based on page countEach job has 1 page
+    # partitions = partition_array_greedy(pdf_info, k)
+    #
+    # # Process each partition in parallel
+    # all_images_h = {}
+    #
+    # with concurrent.futures.ProcessPoolExecutor(max_workers=k) as executor:
+    #     # Submit one task per partition
+    #     futures = []
+    #     for sn, partition in enumerate(partitions):
+    #         # Get the jobs for this partition
+    #         partition_jobs = [pdf_info[idx] for idx in partition]
+    #
+    #         # Submit the task
+    #         future = executor.submit(
+    #             process_pdf_batch,
+    #             partition_jobs,
+    #             sn
+    #         )
+    #         futures.append(future)
+    #     # Process results as they complete
+    #     for i, future in enumerate(concurrent.futures.as_completed(futures)):
+    #         try:
+    #             idx, images = future.result()
+    #             all_images_h[idx] = images
+    #         except Exception as e:
+    #             print(f'Error processing partition: {e}')
+    # results = [None] * len(pdf_paths)
+    # for i in range(len(partitions)):
+    #     partition = partitions[i]
+    #     for j in range(len(partition)):
+    #         with open(pdf_info[partition[j]][0], 'rb') as f:
+    #             pdf_bytes = f.read()
+    #         dataset = PymuDocDataset(pdf_bytes, lang=lang)
+    #         dataset.set_images(all_images_h[i][j])
+    #         results[partition[j]] = dataset
+    # return results

third_party/MinerU/magic_pdf/data/data_reader_writer/__init__.py

@@ -0,0 +1,12 @@
+from magic_pdf.data.data_reader_writer.filebase import \
+    FileBasedDataReader  # noqa: F401
+from magic_pdf.data.data_reader_writer.filebase import \
+    FileBasedDataWriter  # noqa: F401
+from magic_pdf.data.data_reader_writer.multi_bucket_s3 import \
+    MultiBucketS3DataReader  # noqa: F401
+from magic_pdf.data.data_reader_writer.multi_bucket_s3 import \
+    MultiBucketS3DataWriter  # noqa: F401
+from magic_pdf.data.data_reader_writer.s3 import S3DataReader  # noqa: F401
+from magic_pdf.data.data_reader_writer.s3 import S3DataWriter  # noqa: F401
+from magic_pdf.data.data_reader_writer.base import DataReader  # noqa: F401
+from magic_pdf.data.data_reader_writer.base import DataWriter  # noqa: F401

third_party/MinerU/magic_pdf/data/data_reader_writer/base.py

@@ -0,0 +1,63 @@
+
+from abc import ABC, abstractmethod
+
+
+class DataReader(ABC):
+
+    def read(self, path: str) -> bytes:
+        """Read the file.
+
+        Args:
+            path (str): file path to read
+
+        Returns:
+            bytes: the content of the file
+        """
+        return self.read_at(path)
+
+    @abstractmethod
+    def read_at(self, path: str, offset: int = 0, limit: int = -1) -> bytes:
+        """Read the file at offset and limit.
+
+        Args:
+            path (str): the file path
+            offset (int, optional): the number of bytes skipped. Defaults to 0.
+            limit (int, optional): the length of bytes want to read. Defaults to -1.
+
+        Returns:
+            bytes: the content of the file
+        """
+        pass
+
+
+class DataWriter(ABC):
+    @abstractmethod
+    def write(self, path: str, data: bytes) -> None:
+        """Write the data to the file.
+
+        Args:
+            path (str): the target file where to write
+            data (bytes): the data want to write
+        """
+        pass
+
+    def write_string(self, path: str, data: str) -> None:
+        """Write the data to file, the data will be encoded to bytes.
+
+        Args:
+            path (str): the target file where to write
+            data (str): the data want to write
+        """
+
+        def safe_encode(data: str, method: str):
+            try:
+                bit_data = data.encode(encoding=method, errors='replace')
+                return bit_data, True
+            except:  # noqa
+                return None, False
+
+        for method in ['utf-8', 'ascii']:
+            bit_data, flag = safe_encode(data, method)
+            if flag:
+                self.write(path, bit_data)
+                break

third_party/MinerU/magic_pdf/data/data_reader_writer/filebase.py

@@ -0,0 +1,62 @@
+import os
+
+from magic_pdf.data.data_reader_writer.base import DataReader, DataWriter
+
+
+class FileBasedDataReader(DataReader):
+    def __init__(self, parent_dir: str = ''):
+        """Initialized with parent_dir.
+
+        Args:
+            parent_dir (str, optional): the parent directory that may be used within methods. Defaults to ''.
+        """
+        self._parent_dir = parent_dir
+
+    def read_at(self, path: str, offset: int = 0, limit: int = -1) -> bytes:
+        """Read at offset and limit.
+
+        Args:
+            path (str): the path of file, if the path is relative path, it will be joined with parent_dir.
+            offset (int, optional): the number of bytes skipped. Defaults to 0.
+            limit (int, optional): the length of bytes want to read. Defaults to -1.
+
+        Returns:
+            bytes: the content of file
+        """
+        fn_path = path
+        if not os.path.isabs(fn_path) and len(self._parent_dir) > 0:
+            fn_path = os.path.join(self._parent_dir, path)
+
+        with open(fn_path, 'rb') as f:
+            f.seek(offset)
+            if limit == -1:
+                return f.read()
+            else:
+                return f.read(limit)
+
+
+class FileBasedDataWriter(DataWriter):
+    def __init__(self, parent_dir: str = '') -> None:
+        """Initialized with parent_dir.
+
+        Args:
+            parent_dir (str, optional): the parent directory that may be used within methods. Defaults to ''.
+        """
+        self._parent_dir = parent_dir
+
+    def write(self, path: str, data: bytes) -> None:
+        """Write file with data.
+
+        Args:
+            path (str): the path of file, if the path is relative path, it will be joined with parent_dir.
+            data (bytes): the data want to write
+        """
+        fn_path = path
+        if not os.path.isabs(fn_path) and len(self._parent_dir) > 0:
+            fn_path = os.path.join(self._parent_dir, path)
+
+        if not os.path.exists(os.path.dirname(fn_path)) and os.path.dirname(fn_path) != "":
+            os.makedirs(os.path.dirname(fn_path), exist_ok=True)
+
+        with open(fn_path, 'wb') as f:
+            f.write(data)

third_party/MinerU/magic_pdf/data/data_reader_writer/multi_bucket_s3.py

@@ -0,0 +1,145 @@
+
+from magic_pdf.config.exceptions import InvalidConfig, InvalidParams
+from magic_pdf.data.data_reader_writer.base import DataReader, DataWriter
+from magic_pdf.data.io.s3 import S3Reader, S3Writer
+from magic_pdf.data.schemas import S3Config
+from magic_pdf.libs.path_utils import (parse_s3_range_params, parse_s3path,
+                                       remove_non_official_s3_args)
+
+
+class MultiS3Mixin:
+    def __init__(self, default_prefix: str, s3_configs: list[S3Config]):
+        """Initialized with multiple s3 configs.
+
+        Args:
+            default_prefix (str): the default prefix of the relative path. for example, {some_bucket}/{some_prefix} or {some_bucket}
+            s3_configs (list[S3Config]): list of s3 configs, the bucket_name must be unique in the list.
+
+        Raises:
+            InvalidConfig: default bucket config not in s3_configs.
+            InvalidConfig: bucket name not unique in s3_configs.
+            InvalidConfig: default bucket must be provided.
+        """
+        if len(default_prefix) == 0:
+            raise InvalidConfig('default_prefix must be provided')
+
+        arr = default_prefix.strip('/').split('/')
+        self.default_bucket = arr[0]
+        self.default_prefix = '/'.join(arr[1:])
+
+        found_default_bucket_config = False
+        for conf in s3_configs:
+            if conf.bucket_name == self.default_bucket:
+                found_default_bucket_config = True
+                break
+
+        if not found_default_bucket_config:
+            raise InvalidConfig(
+                f'default_bucket: {self.default_bucket} config must be provided in s3_configs: {s3_configs}'
+            )
+
+        uniq_bucket = set([conf.bucket_name for conf in s3_configs])
+        if len(uniq_bucket) != len(s3_configs):
+            raise InvalidConfig(
+                f'the bucket_name in s3_configs: {s3_configs} must be unique'
+            )
+
+        self.s3_configs = s3_configs
+        self._s3_clients_h: dict = {}
+
+
+class MultiBucketS3DataReader(DataReader, MultiS3Mixin):
+    def read(self, path: str) -> bytes:
+        """Read the path from s3, select diffect bucket client for each request
+        based on the bucket, also support range read.
+
+        Args:
+            path (str): the s3 path of file, the path must be in the format of s3://bucket_name/path?offset,limit.
+            for example: s3://bucket_name/path?0,100.
+
+        Returns:
+            bytes: the content of s3 file.
+        """
+        may_range_params = parse_s3_range_params(path)
+        if may_range_params is None or 2 != len(may_range_params):
+            byte_start, byte_len = 0, -1
+        else:
+            byte_start, byte_len = int(may_range_params[0]), int(may_range_params[1])
+        path = remove_non_official_s3_args(path)
+        return self.read_at(path, byte_start, byte_len)
+
+    def __get_s3_client(self, bucket_name: str):
+        if bucket_name not in set([conf.bucket_name for conf in self.s3_configs]):
+            raise InvalidParams(
+                f'bucket name: {bucket_name} not found in s3_configs: {self.s3_configs}'
+            )
+        if bucket_name not in self._s3_clients_h:
+            conf = next(
+                filter(lambda conf: conf.bucket_name == bucket_name, self.s3_configs)
+            )
+            self._s3_clients_h[bucket_name] = S3Reader(
+                bucket_name,
+                conf.access_key,
+                conf.secret_key,
+                conf.endpoint_url,
+                conf.addressing_style,
+            )
+        return self._s3_clients_h[bucket_name]
+
+    def read_at(self, path: str, offset: int = 0, limit: int = -1) -> bytes:
+        """Read the file with offset and limit, select diffect bucket client
+        for each request based on the bucket.
+
+        Args:
+            path (str): the file path.
+            offset (int, optional): the number of bytes skipped. Defaults to 0.
+            limit (int, optional): the number of bytes want to read. Defaults to -1 which means infinite.
+
+        Returns:
+            bytes: the file content.
+        """
+        if path.startswith('s3://'):
+            bucket_name, path = parse_s3path(path)
+            s3_reader = self.__get_s3_client(bucket_name)
+        else:
+            s3_reader = self.__get_s3_client(self.default_bucket)
+            if self.default_prefix:
+                path = self.default_prefix + '/' + path
+        return s3_reader.read_at(path, offset, limit)
+
+
+class MultiBucketS3DataWriter(DataWriter, MultiS3Mixin):
+    def __get_s3_client(self, bucket_name: str):
+        if bucket_name not in set([conf.bucket_name for conf in self.s3_configs]):
+            raise InvalidParams(
+                f'bucket name: {bucket_name} not found in s3_configs: {self.s3_configs}'
+            )
+        if bucket_name not in self._s3_clients_h:
+            conf = next(
+                filter(lambda conf: conf.bucket_name == bucket_name, self.s3_configs)
+            )
+            self._s3_clients_h[bucket_name] = S3Writer(
+                bucket_name,
+                conf.access_key,
+                conf.secret_key,
+                conf.endpoint_url,
+                conf.addressing_style,
+            )
+        return self._s3_clients_h[bucket_name]
+
+    def write(self, path: str, data: bytes) -> None:
+        """Write file with data, also select diffect bucket client for each
+        request based on the bucket.
+
+        Args:
+            path (str): the path of file, if the path is relative path, it will be joined with parent_dir.
+            data (bytes): the data want to write.
+        """
+        if path.startswith('s3://'):
+            bucket_name, path = parse_s3path(path)
+            s3_writer = self.__get_s3_client(bucket_name)
+        else:
+            s3_writer = self.__get_s3_client(self.default_bucket)
+            if self.default_prefix:
+                path = self.default_prefix + '/' + path
+        return s3_writer.write(path, data)

third_party/MinerU/magic_pdf/data/data_reader_writer/s3.py

@@ -0,0 +1,73 @@
+from magic_pdf.data.data_reader_writer.multi_bucket_s3 import (
+    MultiBucketS3DataReader, MultiBucketS3DataWriter)
+from magic_pdf.data.schemas import S3Config
+
+
+class S3DataReader(MultiBucketS3DataReader):
+    def __init__(
+        self,
+        default_prefix_without_bucket: str,
+        bucket: str,
+        ak: str,
+        sk: str,
+        endpoint_url: str,
+        addressing_style: str = 'auto',
+    ):
+        """s3 reader client.
+
+        Args:
+            default_prefix_without_bucket: prefix that not contains bucket
+            bucket (str): bucket name
+            ak (str): access key
+            sk (str): secret key
+            endpoint_url (str): endpoint url of s3
+            addressing_style (str, optional): Defaults to 'auto'. Other valid options here are 'path' and 'virtual'
+            refer to https://boto3.amazonaws.com/v1/documentation/api/1.9.42/guide/s3.html
+        """
+        super().__init__(
+            f'{bucket}/{default_prefix_without_bucket}',
+            [
+                S3Config(
+                    bucket_name=bucket,
+                    access_key=ak,
+                    secret_key=sk,
+                    endpoint_url=endpoint_url,
+                    addressing_style=addressing_style,
+                )
+            ],
+        )
+
+
+class S3DataWriter(MultiBucketS3DataWriter):
+    def __init__(
+        self,
+        default_prefix_without_bucket: str,
+        bucket: str,
+        ak: str,
+        sk: str,
+        endpoint_url: str,
+        addressing_style: str = 'auto',
+    ):
+        """s3 writer client.
+
+        Args:
+            default_prefix_without_bucket: prefix that not contains bucket
+            bucket (str): bucket name
+            ak (str): access key
+            sk (str): secret key
+            endpoint_url (str): endpoint url of s3
+            addressing_style (str, optional): Defaults to 'auto'. Other valid options here are 'path' and 'virtual'
+            refer to https://boto3.amazonaws.com/v1/documentation/api/1.9.42/guide/s3.html
+        """
+        super().__init__(
+            f'{bucket}/{default_prefix_without_bucket}',
+            [
+                S3Config(
+                    bucket_name=bucket,
+                    access_key=ak,
+                    secret_key=sk,
+                    endpoint_url=endpoint_url,
+                    addressing_style=addressing_style,
+                )
+            ],
+        )

third_party/MinerU/magic_pdf/data/dataset.py

@@ -0,0 +1,408 @@
+import os
+from abc import ABC, abstractmethod
+from typing import Callable, Iterator
+
+import fitz
+from loguru import logger
+
+from magic_pdf.config.enums import SupportedPdfParseMethod
+from magic_pdf.data.schemas import PageInfo
+from magic_pdf.data.utils import fitz_doc_to_image
+from magic_pdf.filter import classify
+
+
+class PageableData(ABC):
+    @abstractmethod
+    def get_image(self) -> dict:
+        """Transform data to image."""
+        pass
+
+    @abstractmethod
+    def get_doc(self) -> fitz.Page:
+        """Get the pymudoc page."""
+        pass
+
+    @abstractmethod
+    def get_page_info(self) -> PageInfo:
+        """Get the page info of the page.
+
+        Returns:
+            PageInfo: the page info of this page
+        """
+        pass
+
+    @abstractmethod
+    def draw_rect(self, rect_coords, color, fill, fill_opacity, width, overlay):
+        """draw rectangle.
+
+        Args:
+            rect_coords (list[float]): four elements array contain the top-left and bottom-right coordinates, [x0, y0, x1, y1]
+            color (list[float] | None): three element tuple which describe the RGB of the board line, None means no board line
+            fill (list[float] | None): fill the board with RGB, None means will not fill with color
+            fill_opacity (float): opacity of the fill, range from [0, 1]
+            width (float): the width of board
+            overlay (bool): fill the color in foreground or background. True means fill in background.
+        """
+        pass
+
+    @abstractmethod
+    def insert_text(self, coord, content, fontsize, color):
+        """insert text.
+
+        Args:
+            coord (list[float]): four elements array contain the top-left and bottom-right coordinates, [x0, y0, x1, y1]
+            content (str): the text content
+            fontsize (int): font size of the text
+            color (list[float] | None):  three element tuple which describe the RGB of the board line, None will use the default font color!
+        """
+        pass
+
+
+class Dataset(ABC):
+    @abstractmethod
+    def __len__(self) -> int:
+        """The length of the dataset."""
+        pass
+
+    @abstractmethod
+    def __iter__(self) -> Iterator[PageableData]:
+        """Yield the page data."""
+        pass
+
+    @abstractmethod
+    def supported_methods(self) -> list[SupportedPdfParseMethod]:
+        """The methods that this dataset support.
+
+        Returns:
+            list[SupportedPdfParseMethod]: The supported methods, Valid methods are: OCR, TXT
+        """
+        pass
+
+    @abstractmethod
+    def data_bits(self) -> bytes:
+        """The bits used to create this dataset."""
+        pass
+
+    @abstractmethod
+    def get_page(self, page_id: int) -> PageableData:
+        """Get the page indexed by page_id.
+
+        Args:
+            page_id (int): the index of the page
+
+        Returns:
+            PageableData: the page doc object
+        """
+        pass
+
+    @abstractmethod
+    def dump_to_file(self, file_path: str):
+        """Dump the file.
+
+        Args:
+            file_path (str): the file path
+        """
+        pass
+
+    @abstractmethod
+    def apply(self, proc: Callable, *args, **kwargs):
+        """Apply callable method which.
+
+        Args:
+            proc (Callable): invoke proc as follows:
+                proc(self, *args, **kwargs)
+
+        Returns:
+            Any: return the result generated by proc
+        """
+        pass
+
+    @abstractmethod
+    def classify(self) -> SupportedPdfParseMethod:
+        """classify the dataset.
+
+        Returns:
+            SupportedPdfParseMethod: _description_
+        """
+        pass
+
+    @abstractmethod
+    def clone(self):
+        """clone this dataset."""
+        pass
+
+
+class PymuDocDataset(Dataset):
+    def __init__(self, bits: bytes, lang=None):
+        """Initialize the dataset, which wraps the pymudoc documents.
+
+        Args:
+            bits (bytes): the bytes of the pdf
+        """
+        self._raw_fitz = fitz.open('pdf', bits)
+        self._records = [Doc(v) for v in self._raw_fitz]
+        self._data_bits = bits
+        self._raw_data = bits
+        self._classify_result = None
+
+        if lang == '':
+            self._lang = None
+        elif lang == 'auto':
+            from magic_pdf.model.sub_modules.language_detection.utils import \
+                auto_detect_lang
+            self._lang = auto_detect_lang(self._data_bits)
+            logger.info(f'lang: {lang}, detect_lang: {self._lang}')
+        else:
+            self._lang = lang
+            logger.info(f'lang: {lang}')
+
+    def __len__(self) -> int:
+        """The page number of the pdf."""
+        return len(self._records)
+
+    def __iter__(self) -> Iterator[PageableData]:
+        """Yield the page doc object."""
+        return iter(self._records)
+
+    def supported_methods(self) -> list[SupportedPdfParseMethod]:
+        """The method supported by this dataset.
+
+        Returns:
+            list[SupportedPdfParseMethod]: the supported methods
+        """
+        return [SupportedPdfParseMethod.OCR, SupportedPdfParseMethod.TXT]
+
+    def data_bits(self) -> bytes:
+        """The pdf bits used to create this dataset."""
+        return self._data_bits
+
+    def get_page(self, page_id: int) -> PageableData:
+        """The page doc object.
+
+        Args:
+            page_id (int): the page doc index
+
+        Returns:
+            PageableData: the page doc object
+        """
+        return self._records[page_id]
+
+    def dump_to_file(self, file_path: str):
+        """Dump the file.
+
+        Args:
+            file_path (str): the file path
+        """
+
+        dir_name = os.path.dirname(file_path)
+        if dir_name not in ('', '.', '..'):
+            os.makedirs(dir_name, exist_ok=True)
+        self._raw_fitz.save(file_path)
+
+    def apply(self, proc: Callable, *args, **kwargs):
+        """Apply callable method which.
+
+        Args:
+            proc (Callable): invoke proc as follows:
+                proc(dataset, *args, **kwargs)
+
+        Returns:
+            Any: return the result generated by proc
+        """
+        if 'lang' in kwargs and self._lang is not None:
+            kwargs['lang'] = self._lang
+        return proc(self, *args, **kwargs)
+
+    def classify(self) -> SupportedPdfParseMethod:
+        """classify the dataset.
+
+        Returns:
+            SupportedPdfParseMethod: _description_
+        """
+        if self._classify_result is None:
+            self._classify_result = classify(self._data_bits)
+        return self._classify_result
+
+    def clone(self):
+        """clone this dataset."""
+        return PymuDocDataset(self._raw_data)
+
+    def set_images(self, images):
+        for i in range(len(self._records)):
+            self._records[i].set_image(images[i])
+
+class ImageDataset(Dataset):
+    def __init__(self, bits: bytes, lang=None):
+        """Initialize the dataset, which wraps the pymudoc documents.
+
+        Args:
+            bits (bytes): the bytes of the photo which will be converted to pdf first. then converted to pymudoc.
+        """
+        pdf_bytes = fitz.open(stream=bits).convert_to_pdf()
+        self._raw_fitz = fitz.open('pdf', pdf_bytes)
+        self._records = [Doc(v) for v in self._raw_fitz]
+        self._raw_data = bits
+        self._data_bits = pdf_bytes
+
+        if lang == '':
+            self._lang = None
+        elif lang == 'auto':
+            from magic_pdf.model.sub_modules.language_detection.utils import \
+                auto_detect_lang
+            self._lang = auto_detect_lang(self._data_bits)
+            logger.info(f'lang: {lang}, detect_lang: {self._lang}')
+        else:
+            self._lang = lang
+            # logger.info(f'lang: {lang}')
+
+    def __len__(self) -> int:
+        """The length of the dataset."""
+        return len(self._records)
+
+    def __iter__(self) -> Iterator[PageableData]:
+        """Yield the page object."""
+        return iter(self._records)
+
+    def supported_methods(self):
+        """The method supported by this dataset.
+
+        Returns:
+            list[SupportedPdfParseMethod]: the supported methods
+        """
+        return [SupportedPdfParseMethod.OCR]
+
+    def data_bits(self) -> bytes:
+        """The pdf bits used to create this dataset."""
+        return self._data_bits
+
+    def get_page(self, page_id: int) -> PageableData:
+        """The page doc object.
+
+        Args:
+            page_id (int): the page doc index
+
+        Returns:
+            PageableData: the page doc object
+        """
+        return self._records[page_id]
+
+    def dump_to_file(self, file_path: str):
+        """Dump the file.
+
+        Args:
+            file_path (str): the file path
+        """
+        dir_name = os.path.dirname(file_path)
+        if dir_name not in ('', '.', '..'):
+            os.makedirs(dir_name, exist_ok=True)
+        self._raw_fitz.save(file_path)
+
+    def apply(self, proc: Callable, *args, **kwargs):
+        """Apply callable method which.
+
+        Args:
+            proc (Callable): invoke proc as follows:
+                proc(dataset, *args, **kwargs)
+
+        Returns:
+            Any: return the result generated by proc
+        """
+        return proc(self, *args, **kwargs)
+
+    def classify(self) -> SupportedPdfParseMethod:
+        """classify the dataset.
+
+        Returns:
+            SupportedPdfParseMethod: _description_
+        """
+        return SupportedPdfParseMethod.OCR
+
+    def clone(self):
+        """clone this dataset."""
+        return ImageDataset(self._raw_data)
+
+    def set_images(self, images):
+        for i in range(len(self._records)):
+            self._records[i].set_image(images[i])
+
+class Doc(PageableData):
+    """Initialized with pymudoc object."""
+
+    def __init__(self, doc: fitz.Page):
+        self._doc = doc
+        self._img = None
+
+    def get_image(self):
+        """Return the image info.
+
+        Returns:
+            dict: {
+                img: np.ndarray,
+                width: int,
+                height: int
+            }
+        """
+        if self._img is None:
+            self._img = fitz_doc_to_image(self._doc)
+        return self._img
+
+    def set_image(self, img):
+        """
+        Args:
+            img (np.ndarray): the image
+        """
+        if self._img is None:
+            self._img = img
+
+    def get_doc(self) -> fitz.Page:
+        """Get the pymudoc object.
+
+        Returns:
+            fitz.Page: the pymudoc object
+        """
+        return self._doc
+
+    def get_page_info(self) -> PageInfo:
+        """Get the page info of the page.
+
+        Returns:
+            PageInfo: the page info of this page
+        """
+        page_w = self._doc.rect.width
+        page_h = self._doc.rect.height
+        return PageInfo(w=page_w, h=page_h)
+
+    def __getattr__(self, name):
+        if hasattr(self._doc, name):
+            return getattr(self._doc, name)
+
+    def draw_rect(self, rect_coords, color, fill, fill_opacity, width, overlay):
+        """draw rectangle.
+
+        Args:
+            rect_coords (list[float]): four elements array contain the top-left and bottom-right coordinates, [x0, y0, x1, y1]
+            color (list[float] | None): three element tuple which describe the RGB of the board line, None means no board line
+            fill (list[float] | None): fill the board with RGB, None means will not fill with color
+            fill_opacity (float): opacity of the fill, range from [0, 1]
+            width (float): the width of board
+            overlay (bool): fill the color in foreground or background. True means fill in background.
+        """
+        self._doc.draw_rect(
+            rect_coords,
+            color=color,
+            fill=fill,
+            fill_opacity=fill_opacity,
+            width=width,
+            overlay=overlay,
+        )
+
+    def insert_text(self, coord, content, fontsize, color):
+        """insert text.
+
+        Args:
+            coord (list[float]): four elements array contain the top-left and bottom-right coordinates, [x0, y0, x1, y1]
+            content (str): the text content
+            fontsize (int): font size of the text
+            color (list[float] | None):  three element tuple which describe the RGB of the board line, None will use the default font color!
+        """
+        self._doc.insert_text(coord, content, fontsize=fontsize, color=color)

third_party/MinerU/magic_pdf/data/io/__init__.py

@@ -0,0 +1,6 @@
+
+from magic_pdf.data.io.base import IOReader, IOWriter  # noqa: F401
+from magic_pdf.data.io.http import HttpReader, HttpWriter  # noqa: F401
+from magic_pdf.data.io.s3 import S3Reader, S3Writer  # noqa: F401
+
+__all__ = ['IOReader', 'IOWriter', 'HttpReader', 'HttpWriter', 'S3Reader', 'S3Writer']

third_party/MinerU/magic_pdf/data/io/base.py

@@ -0,0 +1,42 @@
+from abc import ABC, abstractmethod
+
+
+class IOReader(ABC):
+    @abstractmethod
+    def read(self, path: str) -> bytes:
+        """Read the file.
+
+        Args:
+            path (str): file path to read
+
+        Returns:
+            bytes: the content of the file
+        """
+        pass
+
+    @abstractmethod
+    def read_at(self, path: str, offset: int = 0, limit: int = -1) -> bytes:
+        """Read at offset and limit.
+
+        Args:
+            path (str): the path of file, if the path is relative path, it will be joined with parent_dir.
+            offset (int, optional): the number of bytes skipped. Defaults to 0.
+            limit (int, optional): the length of bytes want to read. Defaults to -1.
+
+        Returns:
+            bytes: the content of file
+        """
+        pass
+
+
+class IOWriter(ABC):
+
+    @abstractmethod
+    def write(self, path: str, data: bytes) -> None:
+        """Write file with data.
+
+        Args:
+            path (str): the path of file, if the path is relative path, it will be joined with parent_dir.
+            data (bytes): the data want to write
+        """
+        pass

third_party/MinerU/magic_pdf/data/io/http.py

@@ -0,0 +1,37 @@
+
+import io
+
+import requests
+
+from magic_pdf.data.io.base import IOReader, IOWriter
+
+
+class HttpReader(IOReader):
+
+    def read(self, url: str) -> bytes:
+        """Read the file.
+
+        Args:
+            path (str): file path to read
+
+        Returns:
+            bytes: the content of the file
+        """
+        return requests.get(url).content
+
+    def read_at(self, path: str, offset: int = 0, limit: int = -1) -> bytes:
+        """Not Implemented."""
+        raise NotImplementedError
+
+
+class HttpWriter(IOWriter):
+    def write(self, url: str, data: bytes) -> None:
+        """Write file with data.
+
+        Args:
+            path (str): the path of file, if the path is relative path, it will be joined with parent_dir.
+            data (bytes): the data want to write
+        """
+        files = {'file': io.BytesIO(data)}
+        response = requests.post(url, files=files)
+        assert 300 > response.status_code and response.status_code > 199