Merge pull request #879 from opendatalab/release-0.9.1

Release 0.9.1

Merge pull request #879 from opendatalab/release-0.9.1
Release 0.9.1
069bcfe6 · Xiaomeng Zhao · GitHub · 8ee1da82 · bff7bd93 · 069bcfe6
Unverified Commit 069bcfe6 authored Nov 06, 2024 by Xiaomeng Zhao Committed by GitHub Nov 06, 2024
20 changed files
--- a/.gitattributes
+++ b/.gitattributes
+*.js linguist-vendored
+*.mjs linguist-vendored
+*.html linguist-documentation
+*.css linguist-vendored
+*.scss linguist-vendored
\ No newline at end of file
--- a/.github/ISSUE_TEMPLATE/bug_report.yml
+++ b/.github/ISSUE_TEMPLATE/bug_report.yml
@@ -78,9 +78,9 @@ body:
      #multiple: false
      options:
        -
-        - "0.6.x"
        - "0.7.x"
        - "0.8.x"
+        - "0.9.x"
    validations:
      required: true

--- a/.readthedocs.yaml
+++ b/.readthedocs.yaml
@@ -10,7 +10,7 @@ formats:
 python:
  install:
-    - requirements: docs/zh_cn/requirements.txt
+    - requirements: next_docs/zh_cn/requirements.txt
 sphinx:
-  configuration: docs/zh_cn/conf.py
+  configuration: next_docs/zh_cn/conf.py
--- a/README.md
+++ b/README.md
@@ -17,7 +17,7 @@
 [![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://opendatalab.com/OpenSourceTools/Extractor/PDF)
 [![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,PHN2ZyB3aWR0aD0iMjIzIiBoZWlnaHQ9IjIwMCIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj4KCiA8Zz4KICA8dGl0bGU+TGF5ZXIgMTwvdGl0bGU+CiAgPHBhdGggaWQ9InN2Z18xNCIgZmlsbD0iIzYyNGFmZiIgZD0ibTAsODkuODRsMjUuNjUsMGwwLDI1LjY0OTk5bC0yNS42NSwwbDAsLTI1LjY0OTk5eiIvPgogIDxwYXRoIGlkPSJzdmdfMTUiIGZpbGw9IiM2MjRhZmYiIGQ9Im05OS4xNCwxMTUuNDlsMjUuNjUsMGwwLDI1LjY1bC0yNS42NSwwbDAsLTI1LjY1eiIvPgogIDxwYXRoIGlkPSJzdmdfMTYiIGZpbGw9IiM2MjRhZmYiIGQ9Im0xNzYuMDksMTQxLjE0bC0yNS42NDk5OSwwbDAsMjIuMTlsNDcuODQsMGwwLC00Ny44NGwtMjIuMTksMGwwLDI1LjY1eiIvPgogIDxwYXRoIGlkPSJzdmdfMTciIGZpbGw9IiMzNmNmZDEiIGQ9Im0xMjQuNzksODkuODRsMjUuNjUsMGwwLDI1LjY0OTk5bC0yNS42NSwwbDAsLTI1LjY0OTk5eiIvPgogIDxwYXRoIGlkPSJzdmdfMTgiIGZpbGw9IiMzNmNmZDEiIGQ9Im0wLDY0LjE5bDI1LjY1LDBsMCwyNS42NWwtMjUuNjUsMGwwLC0yNS42NXoiLz4KICA8cGF0aCBpZD0ic3ZnXzE5IiBmaWxsPSIjNjI0YWZmIiBkPSJtMTk4LjI4LDg5Ljg0bDI1LjY0OTk5LDBsMCwyNS42NDk5OWwtMjUuNjQ5OTksMGwwLC0yNS42NDk5OXoiLz4KICA8cGF0aCBpZD0ic3ZnXzIwIiBmaWxsPSIjMzZjZmQxIiBkPSJtMTk4LjI4LDY0LjE5bDI1LjY0OTk5LDBsMCwyNS42NWwtMjUuNjQ5OTksMGwwLC0yNS42NXoiLz4KICA8cGF0aCBpZD0ic3ZnXzIxIiBmaWxsPSIjNjI0YWZmIiBkPSJtMTUwLjQ0LDQybDAsMjIuMTlsMjUuNjQ5OTksMGwwLDI1LjY1bDIyLjE5LDBsMCwtNDcuODRsLTQ3Ljg0LDB6Ii8+CiAgPHBhdGggaWQ9InN2Z18yMiIgZmlsbD0iIzM2Y2ZkMSIgZD0ibTczLjQ5LDg5Ljg0bDI1LjY1LDBsMCwyNS42NDk5OWwtMjUuNjUsMGwwLC0yNS42NDk5OXoiLz4KICA8cGF0aCBpZD0ic3ZnXzIzIiBmaWxsPSIjNjI0YWZmIiBkPSJtNDcuODQsNjQuMTlsMjUuNjUsMGwwLC0yMi4xOWwtNDcuODQsMGwwLDQ3Ljg0bDIyLjE5LDBsMCwtMjUuNjV6Ii8+CiAgPHBhdGggaWQ9InN2Z18yNCIgZmlsbD0iIzYyNGFmZiIgZD0ibTQ3Ljg0LDExNS40OWwtMjIuMTksMGwwLDQ3Ljg0bDQ3Ljg0LDBsMCwtMjIuMTlsLTI1LjY1LDBsMCwtMjUuNjV6Ii8+CiA8L2c+Cjwvc3ZnPg==&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/papayalove/b5f4913389e7ff9883c6b687de156e78/mineru_demo.ipynb)
+[![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)
@@ -42,6 +42,7 @@
 </div>
 # Changelog
+- 2024/11/06 0.9.1 released. Integrated the [StructTable-InternVL2-1B](https://huggingface.co/U4R/StructTable-InternVL2-1B) model for table recognition functionality.
 - 2024/10/31 0.9.0 released. This is a major new version with extensive code refactoring, addressing numerous issues, improving performance, reducing hardware requirements, and enhancing usability:
  - Refactored the sorting module code to use [layoutreader](https://github.com/ppaanngggg/layoutreader) for reading order sorting, ensuring high accuracy in various layouts.
  - Refactored the paragraph concatenation module to achieve good results in cross-column, cross-page, cross-figure, and cross-table scenarios.

--- a/README_ja-JP.md
+++ b/README_ja-JP.md
@@ -18,6 +18,9 @@
 <a href="https://trendshift.io/repositories/11174" target="_blank"><img src="https://trendshift.io/api/badge/repositories/11174" alt="opendatalab%2FMinerU | Trendshift" style="width: 200px; height: 55px;"/></a>
+<div align="center" style="color: red; background-color: #ffdddd; padding: 10px; border: 1px solid red; border-radius: 5px;">
+  <strong>NOTE：</strong> このドキュメントはすでに古くなっています。最新版のドキュメントを参照してください。
+</div>
 [English](README.md) | [简体中文](README_zh-CN.md) | [日本語](README_ja-JP.md)

--- a/README_zh-CN.md
+++ b/README_zh-CN.md
@@ -17,7 +17,7 @@
 [![OpenDataLab](https://img.shields.io/badge/Demo_on_OpenDataLab-blue?logo=data:image/svg+xml;base64,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&labelColor=white)](https://opendatalab.com/OpenSourceTools/Extractor/PDF)
 [![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/papayalove/b5f4913389e7ff9883c6b687de156e78/mineru_demo.ipynb)
+[![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)
@@ -43,6 +43,7 @@
 # 更新记录
+- 2024/11/06 0.9.1发布，为表格识别功能接入了[StructTable-InternVL2-1B](https://huggingface.co/U4R/StructTable-InternVL2-1B)模型
 - 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：
  - 重构排序模块代码，使用 [layoutreader](https://github.com/ppaanngggg/layoutreader) 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
  - 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

--- a/docs/FAQ_en_us.md
+++ b/docs/FAQ_en_us.md
@@ -57,3 +57,10 @@ pip install paddlepaddle-gpu==3.0.0b1 -i https://www.paddlepaddle.org.cn/package
 ```
 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
--- a/docs/FAQ_zh_cn.md
+++ b/docs/FAQ_zh_cn.md
@@ -59,3 +59,9 @@ pip install paddlepaddle-gpu==3.0.0b1 -i https://www.paddlepaddle.org.cn/package
 ```
 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
\ No newline at end of file
--- a/docs/README_Ubuntu_CUDA_Acceleration_en_US.md
+++ b/docs/README_Ubuntu_CUDA_Acceleration_en_US.md
@@ -93,7 +93,7 @@ Download a sample file from the repository and test it.
 ```sh
 wget https://github.com/opendatalab/MinerU/raw/master/demo/small_ocr.pdf
-magic-pdf -p small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
 ```
 ### 9. Test CUDA Acceleration
@@ -108,7 +108,7 @@ If your graphics card has at least **8GB** of VRAM, follow these steps to test C
   ```
 2. Test CUDA acceleration with the following command:
   ```sh
-   magic-pdf -p small_ocr.pdf
+   magic-pdf -p small_ocr.pdf -o ./output
   ```
 ### 10. Enable CUDA Acceleration for OCR
@@ -119,5 +119,5 @@ If your graphics card has at least **8GB** of VRAM, follow these steps to test C
   ```
 2. Test OCR acceleration with the following command:
   ```sh
-   magic-pdf -p small_ocr.pdf
+   magic-pdf -p small_ocr.pdf -o ./output
   ```
--- a/docs/README_Ubuntu_CUDA_Acceleration_zh_CN.md
+++ b/docs/README_Ubuntu_CUDA_Acceleration_zh_CN.md
@@ -11,7 +11,6 @@ nvidia-smi
 注意:`CUDA Version` 显示的版本号应 >= 12.1，如显示的版本号小于12.1，请升级驱动
 ```plaintext
-```
 +---------------------------------------------------------------------------------------+
 | NVIDIA-SMI 537.34                 Driver Version: 537.34       CUDA Version: 12.2     |
 |-----------------------------------------+----------------------+----------------------+
@@ -93,7 +92,7 @@ pip install -U magic-pdf[full] --extra-index-url https://wheels.myhloli.com -i h
 ```bash
 wget https://gitee.com/myhloli/MinerU/raw/master/demo/small_ocr.pdf
-magic-pdf -p small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
 ```
 ## 9. 测试CUDA加速
@@ -111,7 +110,7 @@ magic-pdf -p small_ocr.pdf
 **2.运行以下命令测试cuda加速效果**
 ```bash
-magic-pdf -p small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
 ```
 > 提示：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，`layout detection cost` 和 `mfr time` 应提速10倍以上。
@@ -127,7 +126,7 @@ python -m pip install paddlepaddle-gpu==3.0.0b1 -i https://www.paddlepaddle.org.
 **2.运行以下命令测试ocr加速效果**
 ```bash
-magic-pdf -p small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
 ```
 > 提示：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，`ocr cost`应提速10倍以上。
--- a/docs/README_Windows_CUDA_Acceleration_en_US.md
+++ b/docs/README_Windows_CUDA_Acceleration_en_US.md
@@ -53,7 +53,7 @@ Download a sample file from the repository and test it.
 ```powershell
  wget https://github.com/opendatalab/MinerU/raw/master/demo/small_ocr.pdf -O small_ocr.pdf
-  magic-pdf -p small_ocr.pdf
+  magic-pdf -p small_ocr.pdf -o ./output
 ```
 ### 8. Test CUDA Acceleration
@@ -86,7 +86,7 @@ If your graphics card has at least 8GB of VRAM, follow these steps to test CUDA-
 3. **Run the following command to test CUDA acceleration**:
   ```
-   magic-pdf -p small_ocr.pdf
+   magic-pdf -p small_ocr.pdf -o ./output
   ```
 ### 9. Enable CUDA Acceleration for OCR
@@ -97,5 +97,5 @@ If your graphics card has at least 8GB of VRAM, follow these steps to test CUDA-
   ```
 2. **Run the following command to test OCR acceleration**:
   ```
-   magic-pdf -p small_ocr.pdf
+   magic-pdf -p small_ocr.pdf -o ./output
   ```
--- a/docs/README_Windows_CUDA_Acceleration_zh_CN.md
+++ b/docs/README_Windows_CUDA_Acceleration_zh_CN.md
@@ -55,7 +55,7 @@ pip install -U magic-pdf[full] --extra-index-url https://wheels.myhloli.com -i h
 ```powershell
 wget https://github.com/opendatalab/MinerU/raw/master/demo/small_ocr.pdf -O small_ocr.pdf
- magic-pdf -p small_ocr.pdf
+ magic-pdf -p small_ocr.pdf -o ./output
 ```
 ## 8. 测试CUDA加速
@@ -87,7 +87,7 @@ pip install --force-reinstall torch==2.3.1 torchvision==0.18.1 --index-url https
 **3.运行以下命令测试cuda加速效果**
 ```bash
-magic-pdf -p small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
 ```
 > 提示：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，`layout detection time` 和 `mfr time` 应提速10倍以上。
@@ -103,7 +103,7 @@ pip install paddlepaddle-gpu==2.6.1
 **2.运行以下命令测试ocr加速效果**
 ```bash
-magic-pdf -p small_ocr.pdf
+magic-pdf -p small_ocr.pdf -o ./output
 ```
 > 提示：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，`ocr time`应提速10倍以上。
--- a/magic_pdf/data/data_reader_writer/multi_bucket_s3.py
+++ b/magic_pdf/data/data_reader_writer/multi_bucket_s3.py
+import os
 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
@@ -7,30 +8,34 @@ from magic_pdf.libs.path_utils import (parse_s3_range_params, parse_s3path,
 class MultiS3Mixin:
-    def __init__(self, default_bucket: str, s3_configs: list[S3Config]):
+    def __init__(self, default_prefix: str, s3_configs: list[S3Config]):
        """Initialized with multiple s3 configs.
        Args:
-            default_bucket (str): the default bucket name of the relative path
+            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: default bucket config not in s3_configs.
-            InvalidConfig: bucket name not unique in s3_configs
+            InvalidConfig: bucket name not unique in s3_configs.
-            InvalidConfig: default bucket must be provided
+            InvalidConfig: default bucket must be provided.
        """
-        if len(default_bucket) == 0:
+        if len(default_prefix) == 0:
-            raise InvalidConfig('default_bucket must be provided')
+            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 == default_bucket:
+            if conf.bucket_name == self.default_bucket:
                found_default_bucket_config = True
                break
        if not found_default_bucket_config:
            raise InvalidConfig(
-                f'default_bucket: {default_bucket} config must be provided in s3_configs: {s3_configs}'
+                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])
@@ -39,7 +44,6 @@ class MultiS3Mixin:
                f'the bucket_name in s3_configs: {s3_configs} must be unique'
            )
-        self.default_bucket = default_bucket
        self.s3_configs = s3_configs
        self._s3_clients_h: dict = {}
@@ -47,14 +51,14 @@ class MultiS3Mixin:
 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 path, also support range read.
+        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
+            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
+            for example: s3://bucket_name/path?0,100.
        Returns:
-            bytes: the content of s3 file
+            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):
@@ -84,21 +88,22 @@ class MultiBucketS3DataReader(DataReader, MultiS3Mixin):
    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 path.
+        for each request based on the bucket.
        Args:
-            path (str): the file path
+            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
+            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)
+            path = os.path.join(self.default_prefix, path)
        return s3_reader.read_at(path, offset, limit)
@@ -123,15 +128,16 @@ class MultiBucketS3DataWriter(DataWriter, MultiS3Mixin):
    def write(self, path: str, data: bytes) -> None:
        """Write file with data, also select diffect bucket client for each
-        request based on the path.
+        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
+            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)
+            path = os.path.join(self.default_prefix, path)
        return s3_writer.write(path, data)
--- a/magic_pdf/data/data_reader_writer/s3.py
+++ b/magic_pdf/data/data_reader_writer/s3.py
@@ -6,6 +6,7 @@ from magic_pdf.data.schemas import S3Config
 class S3DataReader(MultiBucketS3DataReader):
    def __init__(
        self,
+        default_prefix_without_bucket: str,
        bucket: str,
        ak: str,
        sk: str,
@@ -15,6 +16,7 @@ class S3DataReader(MultiBucketS3DataReader):
        """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
@@ -23,7 +25,7 @@ class S3DataReader(MultiBucketS3DataReader):
            refer to https://boto3.amazonaws.com/v1/documentation/api/1.9.42/guide/s3.html
        """
        super().__init__(
-            bucket,
+            f'{bucket}/{default_prefix_without_bucket}',
            [
                S3Config(
                    bucket_name=bucket,
@@ -39,6 +41,7 @@ class S3DataReader(MultiBucketS3DataReader):
 class S3DataWriter(MultiBucketS3DataWriter):
    def __init__(
        self,
+        default_prefix_without_bucket: str,
        bucket: str,
        ak: str,
        sk: str,
@@ -48,6 +51,7 @@ class S3DataWriter(MultiBucketS3DataWriter):
        """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
@@ -56,7 +60,7 @@ class S3DataWriter(MultiBucketS3DataWriter):
            refer to https://boto3.amazonaws.com/v1/documentation/api/1.9.42/guide/s3.html
        """
        super().__init__(
-            bucket,
+            f'{bucket}/{default_prefix_without_bucket}',
            [
                S3Config(
                    bucket_name=bucket,

--- a/magic_pdf/data/io/__init__.py
+++ b/magic_pdf/data/io/__init__.py
+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']
\ No newline at end of file
--- a/magic_pdf/data/io/base.py
+++ b/magic_pdf/data/io/base.py
@@ -29,7 +29,7 @@ class IOReader(ABC):
        pass
-class IOWriter:
+class IOWriter(ABC):
    @abstractmethod
    def write(self, path: str, data: bytes) -> None:

--- a/magic_pdf/data/schemas.py
+++ b/magic_pdf/data/schemas.py
@@ -3,6 +3,8 @@ from pydantic import BaseModel, Field
 class S3Config(BaseModel):
+    """S3 config
+    """
    bucket_name: str = Field(description='s3 bucket name', min_length=1)
    access_key: str = Field(description='s3 access key', min_length=1)
    secret_key: str = Field(description='s3 secret key', min_length=1)
@@ -11,5 +13,7 @@ class S3Config(BaseModel):
 class PageInfo(BaseModel):
+    """The width and height of page
+    """
    w: float = Field(description='the width of page')
    h: float = Field(description='the height of page')
--- a/magic_pdf/dict2md/ocr_mkcontent.py
+++ b/magic_pdf/dict2md/ocr_mkcontent.py
@@ -119,6 +119,16 @@ def detect_language(text):
        return 'empty'
+# 连写字符拆分
+def __replace_ligatures(text: str):
+    text = re.sub(r'ﬁ', 'fi', text)  # 替换 fi 连写符
+    text = re.sub(r'ﬂ', 'fl', text)  # 替换 fl 连写符
+    text = re.sub(r'ﬀ', 'ff', text)  # 替换 ff 连写符
+    text = re.sub(r'ﬃ', 'ffi', text)  # 替换 ffi 连写符
+    text = re.sub(r'ﬄ', 'ffl', text)  # 替换 ffl 连写符
+    return text
 def merge_para_with_text(para_block):
    para_text = ''
    for i, line in enumerate(para_block['lines']):
@@ -141,22 +151,34 @@ def merge_para_with_text(para_block):
            if span_type == ContentType.Text:
                content = ocr_escape_special_markdown_char(span['content'])
            elif span_type == ContentType.InlineEquation:
-                content = f" ${span['content']}$ "
+                content = f"${span['content']}$"
            elif span_type == ContentType.InterlineEquation:
                content = f"\n$$\n{span['content']}\n$$\n"
+            content = content.strip()
            if content != '':
                langs = ['zh', 'ja', 'ko']
                if line_lang in langs:  # 遇到一些一个字一个span的文档，这种单字语言判断不准，需要用整行文本判断
-                    para_text += content  # 中文/日语/韩文语境下，content间不需要空格分隔
+                    if span_type in [ContentType.Text, ContentType.InterlineEquation]:
-                elif line_lang == 'en':
+                        para_text += content  # 中文/日语/韩文语境下，content间不需要空格分隔
-                    # 如果是前一行带有-连字符，那么末尾不应该加空格
+                    elif span_type == ContentType.InlineEquation:
-                    if __is_hyphen_at_line_end(content):
+                        para_text += f" {content} "
-                        para_text += content[:-1]
-                    else:
-                        para_text += content + ' '
                else:
-                    para_text += content + ' '  # 西方文本语境下 content间需要空格分隔
+                    if span_type in [ContentType.Text, ContentType.InlineEquation]:
+                        # 如果是前一行带有-连字符，那么末尾不应该加空格
+                        if __is_hyphen_at_line_end(content):
+                            para_text += content[:-1]
+                        elif len(content) == 1 and content not in ['A', 'I', 'a', 'i']:
+                            para_text += content
+                        else:  # 西方文本语境下 content间需要空格分隔
+                            para_text += f"{content} "
+                    elif span_type == ContentType.InterlineEquation:
+                        para_text += content
+            else:
+                continue
+    # 连写字符拆分
+    para_text = __replace_ligatures(para_text)
    return para_text

--- a/magic_pdf/model/pdf_extract_kit.py
+++ b/magic_pdf/model/pdf_extract_kit.py
@@ -38,15 +38,13 @@ except ImportError as e:
 from magic_pdf.model.pek_sub_modules.layoutlmv3.model_init import Layoutlmv3_Predictor
 from magic_pdf.model.pek_sub_modules.post_process import latex_rm_whitespace
 from magic_pdf.model.pek_sub_modules.self_modify import ModifiedPaddleOCR
-# from magic_pdf.model.pek_sub_modules.structeqtable.StructTableModel import StructTableModel
+from magic_pdf.model.pek_sub_modules.structeqtable.StructTableModel import StructTableModel
 from magic_pdf.model.ppTableModel import ppTableModel
 def table_model_init(table_model_type, model_path, max_time, _device_='cpu'):
    if table_model_type == MODEL_NAME.STRUCT_EQTABLE:
-        # table_model = StructTableModel(model_path, max_time=max_time, device=_device_)
+        table_model = StructTableModel(model_path, max_time=max_time)
-        logger.error("StructEqTable is under upgrade, the current version does not support it.")
-        exit(1)
    elif table_model_type == MODEL_NAME.TABLE_MASTER:
        config = {
            "model_dir": model_path,
@@ -463,7 +461,9 @@ class CustomPEKModel:
                html_code = None
                if self.table_model_name == MODEL_NAME.STRUCT_EQTABLE:
                    with torch.no_grad():
-                        latex_code = self.table_model.image2latex(new_image)[0]
+                        table_result = self.table_model.predict(new_image, "html")
+                        if len(table_result) > 0:
+                            html_code = table_result[0]
                else:
                    html_code = self.table_model.img2html(new_image)
@@ -474,14 +474,17 @@ class CustomPEKModel:
                # 判断是否返回正常
                if latex_code:
-                    expected_ending = latex_code.strip().endswith('end{tabular}') or latex_code.strip().endswith(
+                    expected_ending = latex_code.strip().endswith('end{tabular}') or latex_code.strip().endswith('end{table}')
-                        'end{table}')
                    if expected_ending:
                        res["latex"] = latex_code
                    else:
                        logger.warning(f"table recognition processing fails, not found expected LaTeX table end")
                elif html_code:
-                    res["html"] = html_code
+                    expected_ending = html_code.strip().endswith('</html>') or html_code.strip().endswith('</table>')
+                    if expected_ending:
+                        res["html"] = html_code
+                    else:
+                        logger.warning(f"table recognition processing fails, not found expected HTML table end")
                else:
                    logger.warning(f"table recognition processing fails, not get latex or html return")
            logger.info(f"table time: {round(time.time() - table_start, 2)}")

--- a/magic_pdf/model/pek_sub_modules/structeqtable/StructTableModel.py
+++ b/magic_pdf/model/pek_sub_modules/structeqtable/StructTableModel.py
-from loguru import logger
+import re
-try:
+import torch
-    from struct_eqtable.model import StructTable
+from struct_eqtable import build_model
-except ImportError:
-    logger.error("StructEqTable is under upgrade, the current version does not support it.")
-from pypandoc import convert_text
 class StructTableModel:
-    def __init__(self, model_path, max_new_tokens=2048, max_time=400, device = 'cpu'):
+    def __init__(self, model_path, max_new_tokens=1024, max_time=60):
        # init
-        self.model_path = model_path
+        assert torch.cuda.is_available(), "CUDA must be available for StructEqTable model."
-        self.max_new_tokens = max_new_tokens # maximum output tokens length
+        self.model = build_model(
-        self.max_time = max_time # timeout for processing in seconds
+            model_ckpt=model_path,
-        if device == 'cuda':
+            max_new_tokens=max_new_tokens,
-            self.model = StructTable(self.model_path, self.max_new_tokens, self.max_time).cuda()
+            max_time=max_time,
+            lmdeploy=False,
+            flash_attn=False,
+            batch_size=1,
+        ).cuda()
+        self.default_format = "html"
+    def predict(self, images, output_format=None, **kwargs):
+        if output_format is None:
+            output_format = self.default_format
        else:
-            self.model = StructTable(self.model_path, self.max_new_tokens, self.max_time)
+            if output_format not in ['latex', 'markdown', 'html']:
+                raise ValueError(f"Output format {output_format} is not supported.")
+        results = self.model(
+            images, output_format=output_format
+        )
+        if output_format == "html":
+            results = [self.minify_html(html) for html in results]
-    def image2latex(self, image) -> str:
+        return results
-        table_latex = self.model.forward(image)
-        return table_latex
-    def image2html(self, image) -> str:
+    def minify_html(self, html):
-        table_latex = self.image2latex(image)
+        # 移除多余的空白字符
-        table_html = convert_text(table_latex, 'html', format='latex')
+        html = re.sub(r'\s+', ' ', html)
-        return table_html
+        # 移除行尾的空白字符
+        html = re.sub(r'\s*>\s*', '>', html)
+        # 移除标签前的空白字符
+        html = re.sub(r'\s*<\s*', '<', html)
+        return html.strip()
\ No newline at end of file