CHANGES
=======

2.0.4
-----

* Hopefully fix travis pypi deploy

2.0.3
-----

* Sort documents in batches as celery's accumulate function does not preserve ordering
* Auto-deploy to pypi using travis
* Intra-XML image path syntax for tesseract segmentation tests
* Remove kraken jpg test as those are never bitonal
* disable ocropy tests for now

2.0.2
-----

* Correct docstring of pybossa plugin
* Adapt get\_state to new 2.x state format
* Forgot to add lies here too back in august
* More helpful exception when executing docless batch
* make tesseract plugin log output
* Add syriac model to script selection
* more hacky switches
* Add hacky switch in frontend
* Update dependencies also in template
* Disable auto queue processing in frontend
* Update web frontend dependencies
* Bump requirements to 4.1 for Task.replace bugfixes
* Wrap add\_task and add\_metadata in promises
* Pipeline complete Batch.run() invocation
* Check bitonality of input images in ocr.kraken
* Resync batch definition before execution
* Unescape URLs in get\_storage\_path\_url
* Correct error output in CLI for remote batches
* Fix classifier identifiers in html
* Bump pillow version as they broke TIFF conversion
* Fix exception in Batch.get\_tasks()
* Allow verbatim inclusion of model repository in kraken plugin
* Remove deprecate file access methods
* Use send\_from\_directory in API server
* Add more metadata fields to web frontend
* Fix arabic font selector
* Add arabic selections to web frontend
* Fix href in frontned
* Fix display of output in frontend
* Fix status output in CLI
* Minor adjustments to introspection functions
* Fix tesseract exception
* Reset segment scope when scoping a line
* Add task/document removal routines to Batch and API
* Fix syntax error in web frontend
* Convert color space to RGB for any\_to\_png
* Correct variable names in pybossa plugin
* Create one pybossa project per batch
* Make pybossa plugin actually work
* pybossa plugin archival fix
* Fix return value of tesseract segmentation
* Import exception to api
* Change output format of segmentations tasks
* circumvent celery.accumulate arbitraryness
* Changed pybossa plugin to submit lines as individual tasks
* Fix document iterator in pybossa plugin
* kraken doesn't allow tuples in segmentation bboxes anymore
* Adapt to new kraken version
* Correct accrued style errors/pep8ify
