Version 0.10

  • Added a proper setup views for the document grouping functionality.

  • Document grouping is now called smart linking as it relates better to how it actually works. The data base schema was changed and users must do the required:

    $ ./manager syncdb

    for the new tables to be created.

  • Grappelli is no longer required as can be uninstalled.

  • New smarter document preview widget that doesn’t allow zooming or viewing unknown or invalid documents.

  • New office document converter, requires:

  • The new office documents converter won’t convert files with the extension .docx because these files are recognized as zip files instead. This is an issue of the libmagic library.

  • New configuration option added CONVERTER_UNOCONV_USE_PIPE that controls how unoconv handles the communication with LibreOffice. The default of True causes unoconv to use pipes, this approach is slower than using TCP/IP ports but it is more stable.

  • Initial REST API that exposes documents properties and one method, this new API is used by the new smart document widget and requires the package djangorestframework, users must issue a:

    $ pip install -r requirements/production.txt

    to install this new requirement.

  • MIME type detection and caching performance updates.

  • Updated the included version of jQuery to 1.7

  • Updated the included version of JqueryAsynchImageLoader to 0.9.7

  • Document image serving response now specifies a MIME type for increased browser compatibility.

  • Small change in the scheduler that increases stability.

  • Russian translation updates Sergey Glita

  • Improved and generalized the OCR queue locking mechanism, this should eliminate any possibility of race conditions between Mayan EDMS OCR nodes.

  • Added support for signals to the OCR queue, this results in instant OCR processing upon submittal of a document to the OCR queue, this works in addition to the current polling processing which eliminates the possibility of stale documents in the OCR queue.

  • Added multiple document OCR submit link

  • Re enabled Tesseract language specific OCR processing and added a one (1) time language neutral retry for failed language specific OCR