Initial draft of a changeset count column

Fix update query to use single quotes

Don't discard changesets with no branch

Touch last_updated when attaching changesets

Fix #145: Full import creates bad history entries

Fix #144: Proper changeset processing order ...

When commiting changesets that resolve an issue, the changeset data
should be saved to the database before beginning to modify the issue
data.  This is to fix a problem of generating an email with a change
history that did not show the attached changesets, which could be
confusing in certain cases.

Fix #110: Non-unique index name in schema block 13

Fix #131: Reversed history entry

Fix incorrect paramater name and PHP notice

Fix #143: Lingering hardcoded strings

Issue #133: Repo config support for API tokens

Fix #142: Added German translation strings ...

Signed-off-by: John Reese <jreese@leetcode.net>

Bug fixes for resolving commits

F/U 6999df13: Use proper repo object

Include repo name in history entries for clarity

Fix #135: Added French translation for core+websvn ...

Signed-off-by: John Reese <jreese@leetcode.net>

Fix #134: Lingering hardcoded language strings ...

Signed-off-by: John Reese <jreese@leetcode.net>

Fix #136: Only output cell if using porting status ...

Signed-off-by: John Reese <jreese@leetcode.net>

Fix SVN debugging output to include final </pre>

Follow up to previous commit to fix versions

Separate remote import page from user imports ...

Due to complexities of trying to have both user-initiated imports and
remote imports handled in the same page, the import process has been
split into two separate use cases.  pages/import.php is now specifically
for use by the remote import process, while pages/repo_import_latest.php
remains for usage by authenticated MantisBT users.

This will require that all cronjobs or post-commit hooks that triggered
a remote import be updated to point to the new import url.  Framework
version has been bumped accordingly.

Version removed from lang/ by translator request

Fix incorrect SVN log parsing regexes ...

When importing a repository with non-standard paths, the parsing regexes
were incorrectly looking for filenames without a leading slash, and also
not looking for filenames with a slash after the branch/tag path.

Also, when importing standard repositories where the trunk/branches/tags
folders were not named in complete lower case, the regexes and logic for
finding those paths has been updated to be case-insensitive, and to
prefer using the same case as found in the repository for trunk.

Update JS behavior to use portable jQuery methods

Fix #126: Sort SVN changesets by numeric revision ...

To better handle merged SVN repositories, it is necessary to retrieve
the latest changeset from the database based on integer revision number
rather than by timestamp.  This will prevent an "infinite loop" of
repeatedly importing the same changesets in cases where the latest
revisions were actually committed before earlier changesets.

Using ANSI SQL's `CAST(? AS DECIMAL)` as it seems to be compatible with
the most database platforms.

Fix #129: Set default table collation to utf8

Add a bit of debug output for SVN imports/checkins

Fix #122: Implement method to edit changeset data ...

Enhances the existing update method for porting status with a full edit
page allowing the user to update the following changeset data:

- Author's user ID
- Committer's user ID
- Changeset branch
- Porting status
- Commit message

Also fixes an issue where users without update access can still see the
dropdown and update form for porting status.

Refactor loading changesets from database rows ...

This both simplifies the codebase by reducing duplication, and also
fixes a bug where the committer info was never loaded from the database
rows because the load_* methods had never been updated with the
appropriate row/property setters.

Added named link to changeset rows for hyperlinks