Skip to main content

Update Log

Menghuan1918July 12, 2024About 5 min

V0.4.9

✨ New Features

  • Introduced a new built-in upload tool: MinIO. You can easily upload images from Markdown documents (whether online links or local links) to MinIO and replace them with links generated by MinIO. 🔍View usage example by @Huxb12138 in #51

🚀 Other

V0.4.8

✨ New Features

  • Added a new oss_choose option to the PDF conversion function, supporting Doc2X's new interface for uploading files via OSS, significantly improving upload speed. The default value is always (all files are uploaded via OSS). 🔍View synchronous interface documentation, 📦View asynchronous interface documentation
  • Introduced the ability to output multiple formats simultaneously (without consuming additional credits). Note that due to export interface rate limitations, enabling this feature will slightly extend conversion time. 🔍View details

🚀 Other

  • More detailed network error detection
  • Deprecated the ocr option as the upstream API no longer provides an ocr toggle option (it is now forced on by default).
  • Adapted to new error codes

V0.4.8b3

⚠️ This is a beta version, which may contain instabilities and potential issues. It is recommended to use it cautiously in production environments. To install this version, use the following command: pip install pdfdeal==0.4.8b3

✨ New Features

  • Added the ability to output multiple formats (without consuming additional credits). Note that due to export interface rate limitations, this will significantly extend conversion time. 🔍View details

🚀 Other

  • More detailed network error detection

V0.4.8b2

⚠️ This is a beta version, which may contain instabilities and potential issues. It is recommended to use it cautiously in production environments. To install this version, use the following command: pip install pdfdeal==0.4.8b2

🚀 Other

  • Deprecated the ocr option as the upstream API no longer provides an ocr toggle option (it is now forced on by default).
  • Adapted to new error codes.

V0.4.8b1

⚠️ This is a beta version, which may contain instabilities and potential issues. It is recommended to use it cautiously in production environments. To install this version, use the following command: pip install pdfdeal==0.4.8b1

✨ New Features

V0.4.7

✨ New Features

  • Enabled HTTP/2 support for all requests, which theoretically can improve file transfer performance

🔧 Bug Fixes

  • Fixed a bug where the image replacement tool mds_replace_imgs was not functioning
  • Fixed a potential deadlock issue when full_speed is enabled

🚀 Other

  • Reintroduced Ruff in GitHub Actions for code checking and formatting verification
  • Added prompts for API key authentication failures

V0.4.6

✨ New Features

  • Introduced the full_speed beta feature, which automatically detects the highest available concurrency limit.

🔧 Bug Fixes

  • Corrected spelling errors in function comments.

🚀 Other

  • More detailed error descriptions are now provided, with trace-id included whenever possible to facilitate problem localization.
  • The experimental retry option has been removed due to not meeting expected results.

V0.4.5

🔧 Bug Fixes

  • Fixed the issue where the error for exceeding the page limit could not be handled

V0.4.4

🔧 Bug Fixes

  • Fixed the issue of too small request intervals

V0.4.3

Note

The V1 interface of Doc2X will soon be deprecated! Please migrate to the V2 interface as soon as possible. Check the Interface Migration Guide to determine if any code changes are necessary.

In most cases, you will not need to change any code, as the 0.4.X version aims to maintain upward compatibility with the 0.3.1 version.

🔧 Bug Fixes

  • Fixed a potential deadlock issue
  • Significantly improved concurrency performance

🚀 Other

  • Synchronized new error codes for Doc2X
  • Improved package dependencies

V0.4.2

Warning

The V1 interface of Doc2X will soon be deprecated! Please migrate to the V2 interface as soon as possible. Check the Interface Migration Guide to determine if any code changes are necessary.

In most cases, you will not need to change any code, as the 0.4.X version aims to maintain upward compatibility with the 0.3.1 version.

  • Fixed an issue where requests could get stuck indefinitely in poor network environments
  • Fixed a problem with wrong parameters in CLI programs.

V0.4.1

Warning

The V1 interface of Doc2X will soon be deprecated! Please migrate to the V2 interface as soon as possible. Check the Interface Migration Guide to determine if any code changes are necessary.

In most cases, you will not need to change any code, as the 0.4.X version aims to maintain upward compatibility with the 0.3.1 version.

✨ New Features

  • The pdf2file function has added an experimental option retry, which determines whether to retry failed conversions. It is disabled by default. This feature will be further improved in future versions, 🔍view.
  • When a single file path is provided, pdf2file will automatically retain the original file name, 🔍view.
  • Updated the CLI program to support the new V2 interface.

🔧 Bug Fixes

  • Fixed the issue where custom export file names were not effective when a single file path was provided.
  • Fixed the issue where downloading converted files could hang in poor network conditions.

🚀 Other

  • Support for Python 3.13, with related tests added in GitHub Action.
  • Experimental support for Python 3.13t (nogil).

V0.4.0

Caution

The V1 interface of Doc2X will be deprecated in the near future! Please migrate to the V2 interface as soon as possible. Please check the [Interface Migration Guide](. /v1tov2.md) to see if any code changes are required for your scenario.

In most cases, you will not need to change any code, and the 0.4.X version will be as upwardly compatible with the 0.3.1 version as possible.

✨ Feature Changes

  • Support for Doc2X V2 interface
  • The pdf2file interface will automatically recognise the input as a folder path / file path / list form file path and process it, view
  • pdf2file will automatically maintain the original file structure, no longer need to manually intervene, View
  • Improve the error reporting hints, now its going to try to provide a solution for the errors reported

🚀 Other

  • Optimise package dependencies, now only two small packages are needed, httpx and pypdf.
  • Easier debug logging switches.

V0.3.1

Caution

After version 0.3.1 the output has been updated to logging, which by default only outputs Warning and above. If you want to see the processing, set the logging level to INFO:

import logging
httpx_logger = logging.getLogger("httpx")
httpx_logger.setLevel(logging.WARNING)
logging.basicConfig(level=logging.INFO)

🚀 Others

  • Changed the package information output to the logging module, which doesn't output a bunch of stuff anymore

V0.3.0

✨ Feature Changes

🔧 BUG Fixes

  • Fixed terminology for switching state prompts
  • Fixed not being able to print error report stacks

🚀 Others

  • Documentation page improved for Linux users (more friendly font designation)
  • Added demonstration of use with RAG applications (e.g. Fastgpt, Dify, etc.)

V0.2.5

✨ Feature Changes

  • Added built-in upload tool: S3

🔧 Bug Fixes

  • MD Document Image Uploader could not handle relative path images.

🚀 Others

  • Introduced Ruff in GitHub Action for code checking and code formatting.

V0.2.4

✨ Feature changes

  • New MD Document Auto Split tool
  • New MD document image upload tool
  • New built-in upload tool: AliCloud OSS
  • CLI tool will keep the source name of the file (instead of UUID name).

🔧 BUG Fixes

  • Fixed an issue where an error message was not displayed when a status request failed

V0.2.3

🔧 BUG Fixes

  • Fixed issue that did not work in Jupyter Notebook
  • Fixed rate limiter not working in pdfdeal function.

V0.2.2

✨ Feature Changes

  • CLI command line program doc2x supports automatic decompression of downloaded zip files

🔧 BUG Fixes

  • The CLI command line program doc2x does not save the key locally in some cases.
  • The Replace image links in Markdown files with links to local files function saves images in the wrong format (saves jpg images as png).

V0.2.1

✨ Feature Changes

  • Updated to adapt new doc2x rate limiting rules from requests per minute RPM -> simultaneous task requests.

🔧 BUG Fixes

  • CLI command line program doc2x does not save error logs, only prints them in the terminal.

V0.2.0

Caution

This version has major interface updates (impact range: all)

  • Function return parameters have changed, please check update details to see how to migrate

✨ Feature Changes

  • Added CLI command line program doc2x, for quickly using doc2x to batch process PDF or image files, please refer to here for usage
  • Adaptation of CLI commands to graphrag has been added, please refer to here for usage
  • Updated Doc2X document translation functions to use see here
  • Enhanced exception handling
  • Function return parameters have changed, will return more detailed content
  • Decoupled various parts of the processing process

🔧 BUG Fixes

  • [Doc2X] When using personal API, if the input file has multiple corrupted files, it may cause an infinite loop
  • [FileTool] The get_files function cannot accept pdf output format

🚀 Others

  • Documentation updated to a separate repository pdfdeal-docs
  • Updated unit tests
Default footer
Copyright © 2024 Menghuan1918