For some reason, BeautifulSoup wants now fromEncoding='UTF-8' in some unidentified cases.

Fix south migration problem caused by 0003_update_keys_to_textversion using TextVersion model which is updated only in 0010_auto__add_field_textversion_category_1__add_field_textversion_category, same with 0007_populate_id_key_in_comments using Comment model updated in 0011_auto__add_field_comment_category.

Reorder classes in models.py.

Definitely this is not a good idea to use ApplicationConfiguration in TextVersion model as default for category_x fields.

Reorders stuffs in models

Reorder ConfigurationManager model, because it broke co-ment lite

Do not proposed undefined categories in filter.

Fixes test for categories defined in workspace while None in textversion.

Fixes lower() applied to None string, and not empty string as changeset 506 has mistakenly fixed.

Fixes one FR translation.

Fixes lower() applied to None string.

Reapply changeset 504.

Adds a category to comments, painted with colored vertical bar.

Add can_edit_comment_own to Individual student (similar to changeset 359 for generic model).

do not skip span for newline textnodes otherwise compute_new_comment_positions() will return bad results for pandoc texts.

Prevents sumitting comment several time while waiting for server response.

Added tag 2.5.0 for changeset 805e08ea57b1

Fixes crash when editing imported text with orphan comment.

One debug line forgotten in previous commit.

One debug line forgotten in previous commit.

Fixes crash when editing, apparently when comment spans on multiple text nodes.

Forgot a condition when using libreoffice as a fallback when abiword crashes.

Use libreoffice to convert html to pdf when abiword crashes.

Do not list all comments when exporting filtered comments.

Since scrollIntoView scroll the embed iframe *and* the parent window, restore the position of the toolbar.

Since scrollIntoView scroll the embed iframe *and* the parent window, save and restore the position of the toolbar.

Do not reset filters when adding a comment.

New arrow-down adn arrow-up icons.

Reverts arrow-down icon.

Adds ompressed version for changeset 485.

Clean versions list/comparison/view and add original version.

New icons.

Preselect embeded code to copy.

Do not try to process custom_css if none has be defined

Add custom CSS and fonts.

No need to get content with display:none on parent view frame

Replace pre_edit stage by an input checkbox to increase performances (compute_new_comment_positions run once).

Add GZipMiddleware to compress exchanged data.

Remove some unused comments.

Add some MIME instructions for font to silence Chrome warnings.

Applies new fonts to interface.

New fonts.

fix switch between different styles.

Do not spannify empty text nodes, prevents abiword crash.

Use co-ment user name when co-ment is directly accessed, even if DECORATED_CREATORS is True (i.e. normaly accessed by API).

Update norwegian translation

Fix filtering of teachers' comments for individual students in role_teacher role model when using DECORATED_CREATORS.

When building toc, delete <a> elements, including when written in uppercase (<A>).

Stupid IE does not know getElementsByClassName.

Stupid IE counts extra element in array when ending by a comma.

indexOf method of Array is unknown by stupid IE.

Fix export in XML of dates when timezone is negative.

For some reasons BeautifulSouf does not convert entities with convertEntities=["xml","html"], use convertEntities=BeautifulSoup.ALL_ENTITIES instead.

Fix import for use with api (with co_ment Drupal module for eg.).

Fix Italian translation.

Adds Italian translation.

Added import from XML file, including text, comments and attachments.

Fixed renderContents on BeautifilSoup if there is no body in source html

When extracting body of a document in HTML source format (i.e. without pandoc conversion before display), use BeautifilSoup, which is more tolerant, instead of xml.dom.minidom, which crashes too often.

Fixed export with abiword by running tidy on html

Fixed bad conversion of & to &amp;

Added export of attachements as inline b64 images for appropriate formats.

Cleaned export.

Added export to XML for re-import (nb. without attachements).

forget a coma in list of input format for abiword conversion.

Change Title: to Subject: in English to not mix with author's title.

Fix #77 html code inside document viewer.

In role_teacher model, individual students can see their own comments but also teacher's ones, whereas individual students' comments cannot be seen by students.

Avoid warnings for deprecated instructions in djan 1.3

Avoid warnings for deprecated instructions in djan 1.3

Avoid warnings for deprecated instructions in djan 1.3

Update settings_local_sample with DECORATED_CREATORS and USE_ABI parameters

Exports to epub only with pandoc 1.9 and upper versions.

Adds export to epub.

Takes into account various releases of pandoc.

Fixes export with abiword when attached images are served by co-ment

Adds generic getElementsByClassName for IE compatibility.

Reverts to changeset 435, and just add {% csrf_token %} to template forgot_pw.html, since CSRF protection seems to be only here (surely because of django.contrib.auth.views).

Fix side effects with csrf token, add csrf_token to every post form + add ajax (see http://stackoverflow.com/questions/5100539/django-csrf-check-failing-with-an-ajax-post-request), remove django.middleware.csrf.CsrfViewMiddleware ?!

Avoids HTTP 403 on login and reset/change password, seehttps://docs.djangoproject.com/en/dev/ref/contrib/csrf/#ref-contrib-csrf.

Avoids HTTP 403 on reset password, seehttps://docs.djangoproject.com/en/dev/ref/contrib/csrf/#ref-contrib-csrf.

Fixed comments filter by name when name is empty

Fixed bad string for export to ms-word 2007.

Added export in .doc and .docx formats.

Respect order by scope in show all comments or show all detached comments.

Fixed sort with y is NoneType.

CSS for filter.

Fixed author when there is none.

Still use --sanitize-html for pandoc 1.5.1.1.

Smaller size for text input form elements in filter.

Added 'me' and 'author' in filter by name.

Fix some javascript error with IE8 (getElementsByClassName() is unknown).

Removes <a> tags inside toc + adjusts height of toc.

Better positionning of table of contents.

Hides button if table of contents is empty.

Adds a table of content.

Prevents javascript error, when iframe is not ready and therefore iframe_onload function is not available.

Still use --sanitize-html for pandoc 1.5.1.1.

Just a typo.

After investigations, commit the proposed patch by kklimonda presented with the following introduction : cm/templates/site/login_form.html template uses a relative url for the form action, which seems to break deployment in the subfolder when trailing slash is missing (the form redirects to the upper folder breaking login). Using {% url login %} (as in the attached patch) instead would fix the issue.

Adds rtfas input formats when converting with abiword.

Adds api/convert to return HTML document from legacy format + Saves attached images when just converting to HTML, without creating a text + Returns nb of detached and removed commentswhen updating text.

Adds image button in rich text editor for HTML.

Fix mismatching filter with tag.

Adds comments' tags when exporting

When exporting, expand entities to keep correct offsets of comments.

Allows multiple co-mentable texts to be isplayed in the same HTML page (with Drupal 7 co_ment module).

Requires django > 1.3

Requires django > 1.3

When last_version is deleted, do not delete text and previous versions in cascade, but really delete text only when there is no previous version, otherwise update text.last_version with previous version.

Merge some previous commits.

Fixes missing closing bracket for column spearator in diff pages.

Start dev of release 1.4.0

Add a tag corresponding to the last release : 1.3.0

Add a VERSION.txt file, to hold the version number of the designated comt release

Update contributors and licence files, adding Gibus and I in it, and the content of the AGPL licence.

Do not transform commentator's displayed name in lower case when posting comment.

repeat 397 (css div.frame)

nothing changed, but hg push needs this commit

Fix selection of amath elements

add some css for sharing-thebook

Add libjs-asciimathml.

Some CSS for footnotes.

Some CSS styles

Default height from embed ifrmae changed from 200px to 600px.

add_comment_button should be hidden with initially_hidden css class if can_create_comment permission is false => do not set this css property directly on html element.

commit c_client.min.js (compressed version of previous commits with scripts/compressjs.py).

Allow permalink on comments even when embedfing iframe.

throbber should be upper

add throbber image

add throbber to wait until text is loaded

better use images for arrows on toggle bar to avoid font size dofferences and ut8 non-compatible browsers

fix a french translation

add arrows on toggle bars

better positioning of 'new comment' form

:active on whole 'add comment' button, not just text

better design for 'add comment' button

rounded corners for 'add comment' button

commit c_client.min.js (compressed version of previous commits with scripts/compressjs.py).

restore image for old 'add comment button' even if unused

title when hovering 'add comment' button

replace 'add comment' button

semicolon useless in python

Add a title attribute on comment scopes to indicate the number of comments in the scope. Also ensure that the last c-count-x defined color class is set even if there are more than 25 comments in the scope.

Correct typo in help messages. Rebuild all locales, eliminating the induced fuzzy strings.

Revision of french translation. No more empty strings (-24)

Revision of spanish translation. No more fuzzy strings (-63). No more empty translations (-91). Fuzzy strings are automatically added when source string change in code. Fuzzy strings must be regularly reviewed.

One french fuzzy string uncomented.

Add spaces before item.count(s) in lists of com state/author/tags/time in com filtering facilities.

markdown2pdf is buggy => convert to HTML and use abiword to export in PDF

For some reasons, abiwords can read background style attribute but not background-color

For some reasons, abiwords can read background style attribute but not background-color

tidyfy html before conversion with abiword

add python pexpect in dependencies

add python pexpect in dependencies

Delete temporary files

Add abiword as an alternative to open office for conversions

Add can_delete_comment_own to Commentator

when trying markdown2pdf without xetex option, use custom-header for propoer input encoding and so on

if markdown2pdf fails try without -xetex option

--xetex option in markdown2pdf seems to cause "Invalid or incomplete multibyte or wide character" errors

if DECORATED_CREATORS take into account the "fake" username if has_own_perm()

merge with changeset 352:7bc8d8bfe443

merge

do not use pandoc to convert from html to html

use different tempfiles for in and out + use xtex for markdown2pdf

replace msdos \r\n

workspace in parameter for memoize

add piwik tracking

prevent uno custom __import__ from messing with south import machinery (to discover south enabled dj apps)

3.0.0 -> 3.0.0b1

3.0.0 -> 3.0.0b1

memoize get_activity

open source -> free software

Missing French translation for some javascript.

Coding style for js: expand tabs

Be more strict by controlling that value ofcomment_op url parameter is set to "reply"

Open automatically reply to comment form when comment_op url parameter is set to "reply", *even if there is no reply yet* (therefore no animation should have been triggered)

Focus on content textarea when opening reply to comment form

Add url parameter (comment_op=reply) to open automatically reply to comment form

api doc

add comment_key optional parameter for /api/comments/ to fetch a single comment

Added tag v1.0.5 for changeset f080de4cad76

Fix browsing by modifications, take into account every comment in a thread (instead of only the first two levels in each thread).

commit c_client.min.js (compressed version of previous commits with scripts/compressjs.py).

focus on comment specified as url parameter *even if it is not the last comment of a thread*

oops forgot "else" on last commit

When comment_id_key is passed as URL parameter, focus on this comment, instead of the top comment of the related thread.

Select only comments from last version.

save prototype for django new version (1.2.3)

Merge with addba77f2f901c689d9abed3d4ce2c5d25fef5f1

i18n compile