First step in two-part process to rename directories to kallithea.
This first step is to change all references in the files where they refer
to the old directory name.

Imported some of the GPLv3'd changes from RhodeCode v2.2.5.

This imports changes between changesets 21af6c4eab3d and 6177597791c2 in
RhodeCode's original repository, including only changes to Python files and HTML.

RhodeCode clearly licensed its changes to these files under GPLv3
in their /LICENSE file, which states the following:
The Python code and integrated HTML are licensed under the GPLv3 license.

(See:
https://code.rhodecode.com/rhodecode/files/v2.2.5/LICENSE
or
http://web.archive.org/web/20140512193334/https://code.rhodecode.com/rhodecode/files/f3b123159901f15426d18e3dc395e8369f70ebe0/LICENSE
for an online copy of that LICENSE file)

Conservancy reviewed these changes and confirmed that they can be licensed as
a whole to the Kallithea project under GPLv3-only.

While some of the contents committed herein are clearly licensed
GPLv3-or-later, on the whole we must assume the are GPLv3-only, since the
statement above from RhodeCode indicates that they intend GPLv3-only as their
license, per GPLv3§14 and other relevant sections of GPLv3.

celery tasks: use None as special value

Use a more common pattern.

moved out password reset tasks from celery, it doesn't make any sense to keep them there, additionally they are broken
in when executing _(), and url() calls. This fixes issue #572

moved out password reset tasks from celery, it doesn't make any sense to keep them there, additionally they are broken
in when executing _(), and url() calls. This fixes issue #572

added recursion limit for stats gathering, sometimes it did >1000 loops which lead to python throwing max recursion depth exceeded error.
ref #642

Refactored create fork function to use new RepoModel functions instead of old
soon-to-be-destroyed functions

New repo model create function
- fixed some sessions calls

utils/conf
- created temporary utils2
- made config.conf for storing some configurations
- fixed some dependency import problems
- code cleanup
- rc-extensions now properly work for celery

garden
- unified logging formatting to use only %

backported fixes for statistics from beta branch, try to fix issue #271
- version bump

- fixed issue with missing commits on some repos commands
- fixed issue with parsing commits for stats when there were only two commits

commit less models
- models don't do any commits(with few exceptions)
- all db transactions should be handled by higher level modules like controllers, celery tasks

implements #193 journal stores information about deleting of repos
- only admin journal does that now
- added new info about creation of repo as fork

- fixes celery sqlalchemy session issues for async forking
- summary page css fixes
- speed optimizations

- refactoring to overcome poor usage of global pylons config
- db transaction fixes
- fixed tests
- garden

#235 forking page repo group selection
- group is selected as parent repo
- refactoring of session behavior to get more atomic operations in models

Notification fixes
- email prefix added to .ini files
- html templates emails
- rewrote email system to use some parts from pyramid_mailer

fixed issues with python2.5
added compat module to rhodecode to have one point of fetching backward incompatible libs.

fixes #200, rewrote the whole caching mechanism to get rid of such problems. Now cached instances are attached
to db repository instance, and then fetched from cache. Also made all current test work.

fixed issue #165 trending source files are now stored in cache as ext only, and translated to description only when displaying, so future changes of mappings will take affect right away.
Also changes the way how map is generating, ie. autogenerate from pygments lexers, and than complete with user defined mappings.

Added api_key into user, api key get's generated again after password change
updated ini files

Added branch filter to repo pager
fixed stop condition for building statistics

moved statistics parse_limit into .ini files
decreased celery max task per child to 1 for better memory usage
stats task take usage of new slice option of repository

made rhodecode work with celery 2.2, made some tasks optimizations(forget results)

added celeryconfig.py with just the definitions of hosts, it seams just this is needed to get celery working nice, all other config options are taken from .ini files. This is a temp workaround until i get the proper soltuion to this problem.

complete rewrite of paster commands,
used fancy celery-pylons baseCommand to implement better support for paster commands
fixed make-index, and paster celeryd. Moved config's add_cache to utils

fixed cache problem,
updated docs

Celery is configured by the .ini files and run from paster now
removed celeryconfig, added homebrew celery-pylons,
added paster celeryd command, fixed tasks to use pylons configs, sqlalchemy sessions

fixed Example celery config to ampq,
fixed session problems on models and celery,
implemented one base model as a parent for other

fixes #69 password confirmation for register dialog.
Fixes appcrash when using some special characters in register field
fixes app crash when no email config is enabled

Code refactor for auth func, preparing for ldap support
css updates.
turned off graph,and branches for git changelog

#50 on point cache invalidation changes.
Created cacheInvalidation table
cleaned up sa sessions from models, since it wasn't really needed.

Refactor codes for scm model
Some test updates, added test for admin user controller

fixed annotation bug, added history to annotation.
multiple fixes for raw_id length
removed unneded function from index daemon.

Fixed age, for new vcs implementation. Removed all obsolete date formatters
Added simplegit middleware.
fixed deps
added scm type icon to main page

Code refactoring,models renames
cleaned up sqlalchemy sessions,
added cache support to most queries in models
fixed test.ini file