addition to the existing 'pyang' checker.
Added modal overlay displays showing the yang check results every place the
yin/yang symbol is shown (red or green) to indicate the presencee and result
of yang checks. Added a Yang Validation: line in the document
meta-information section on the document's page in the datatracker.
Added the result of the xym extaction to the yang check results, to make
extration failures visible.
Added the version of the used xym, pyang, and yanglint commands to the check
results.
Added an action to move successfully extracted and validated modules to the
module library directories immediately on submission.
Added the xym and pyang repositories as svn:external components, rather than
listing them in requirements.txt, as there has been delays of many months
between essential features in the repositories, and an actual release. We may
get occasional buildbot failures if broken code is pulled in from the
repository, but better that than the functionality failure of severely
outdated componets.
Added a new management command to re-run yang validation for active drafts for
which yang modules were found at submission time, in order to pick up imported
models which may have arrived in the model libraries after the draft's
submission. Run daily from bin/daily.
Added a table to hold version information for external commands. The yang
checker output should include the version information of the used checkers,
but seems unnecessary to run each command with its --version switch every
time we check a module...
Added a new management command to collect version information for external
commands on demand. To be run daily from bin/daily.
Added tests to verify that xym, pyang and yanglint information is available
on the submission confirmation page, and updated the yang module contained in
the test document to validate under both pyang and yanglint.
Updated admin.py and resource.py files as needed.
- Legacy-Id: 13630
using the htmlization code previously developed for tools.ietf.org. As the
generation of the htmlized page is a bit too costly to do on the fly for
often-referenced drafts and RFCs, the part of each page which contains the
htmlized document is cached on file with a cache time of 2 weeks.
Changed all links which pointed to the htmlized version on tools to instead
point at the datatracker htmlized document.
Tweaked some URLs which didn't permit retrieval of intermediate-rev-charters.
Narrowed the pattern for document names to disallow dots in names, and
instead explicitly enumerated the few historical draftw with dots in the
name.
Added a file-system cache for the htmlized documents, and specified a
max_entries value for caches, overriding the default 300 entries.
Tweaked the code for new author email entries to provide a time if missing
in an updated entry.
Changed links in various email templates which pointed at tools.ietf.org
pages to instead point at datatracker pages, where appropriate.
Changed the search result rows to provide links to both the current meta-
information document pages (with a (i) info symbol) and to the new htmlized
document pages.
- Legacy-Id: 13040
in order to autogenerate dotted path url pattern names. Updated a number
of url reverses to use dotted path, and removed explicit url pattern names
as needed.
Changed some imports to prevent import of ietf.urls before django
initialization was complete.
Changed 3 cases of form classes being curried to functions; django 1.10
didn't accept that.
Started converting old-style middleware classes to new-style middleware
functions (incomplete).
Tweaked a nomcom decorator to preserve function names and attributes, like
a good decorator should.
Replaced the removed django templatetag 'removetags' with our own version
which uses bleach, and does sanitizing in addition to removing explicitly
mentionied html tags.
Rewrote the filename argument handling in a management command which had
broken with the upgrade.
- Legacy-Id: 12818
From a user perspective: Use friendlier URLs for lists. Reuse the
search results table for displaying lists. Simplify the management
pages and improve the search rule UI to help fill in the values and
validating them, instead of just providing a text field. Fixes#1874.
Add an explicit button for adding individual documents. Include all
changes in the document change streams, not just some changes. Fix a
concurrency issue that allows changed documents to escape the search
rules. Don't create an empty list just be logging in.
From a code maintenance perspective: Clean up the models. Replace the
background caching scheme with direct queries. Get rid of a big chunk
of code. Speed up the code that adds track buttons to search results.
Add tests of all community views. Fixes#1422. Also fix some minor
bugs and oddities here and there.
There's still some work to do with respect to integrating the group
lists better.
- Legacy-Id: 10921
Commit r10578 broke search result table display due to the added column. Fix
this. (Also update vnu.jar. Unrelated, but wasn't careful when committing.)
- Legacy-Id: 10689
Note: SVN reference [10685] has been migrated to Git commit 820d406e33
The previous fix for the search result table width was incorrect.
- Legacy-Id: 10649
Note: SVN reference [10622] has been migrated to Git commit b23952c615
Fix display of AD and shepherd names, and also fix the respective mailto: links.
- Legacy-Id: 9865
Note: SVN reference [9852] has been migrated to Git commit 573496df73
Color agenda/documents rows with position taken when viewing as an AD. Fixes bug #1679.
- Legacy-Id: 9657
Note: SVN reference [9637] has been migrated to Git commit edcaf46138
tabs) and kill the scripts block in favour of just using a js block with
a script tag (which is easier to understand for context-sensitive
modes such as web-mode in Emacs). Also fix a couple of details, e.g.
missing semicolons in JS snippets.
- Legacy-Id: 9096
