jEdit (#157) - Successful:
Changes for Build #157
[jarekczek] Load properties from extras fileset, not src. Increased verbosity.
[jarekczek] Added a check for failure in plugin properties discovery
[k_satoda] Put hard coded workarounds not to cause unexpected soft wrap at
starts/ends of quotes with some non-ASCII quotation marks. The cause of
the problem seem to be in the implementation of BreakIterator, and seems
not to be fixed in near future. (originally reported as SF.net bug
#3488310, analyzed in a derived FR #3489523)
[k_satoda] Gathered various ways of width accumulation into a simple loop after
getting GlyphVector. The complexity is anyway linear. Then, there is
no reason to have different ways in different places.
[k_satoda] Fixed wrong separation in composed character sequence which happened
when font substitution is used on that sequence. (SF.net bug #3495002;
the problem was introduced in r21213)
[k_satoda] Added (prev > 0x7f) as a condition to accept a line break to be
consistent with (next > 0x7f).
With a non-ASCII character "*", rejecting a break between "*a" is not
consistent while accepting a break between "a*".
[jarekczek] build.support: Added ability to automatically download the plugins
being dependencies specified in props file of another plugin, FR #3488089.
Functionality of parsing plugins' props files added by the way to
By the way added build-helper.xml ant library which allows fast inclusion
of libraries, for example necessary to run scripts. This functionality
is based on ivy.
The way plugins are build is not changed yet.
[k_satoda] Reduced the calls to layoutGlyphVector() in case of the font
substitution logic found a character which can't be displayed with the
main font, and also with substitution fonts. Before this change, it
could be called for every single character.
A new private helper class Chunk.FontSubstitution was introduced to make
further tweaks easier without worry of this kind of problem.
[k_satoda] Added a check if font substitution should switch back to the main font
to obey the desired behavior written in the comment for
Chunk.layoutGlyphs(); "All characters that can be rendered with the main
font will be".
The problem was reported in SF.net patch #3488967, but the fix was
restructured by me to be fairly readable.
[kpouer] Moved localization options to general option pane
[k_satoda] Avoid unnecessary use (including allocation) of ArrayList<GlyphVector>
in major cases where font substitution doesn't occur at all.
[kpouer] include localization package
[kpouer] renamed i18n to localization
[kpouer] renamed i18n package into localization
[kpouer] Renamed initi18nProperties into initLocalizationProperties
[k_satoda] Improved readability (naming, comments, etc) of font substitution logic
to make further changes clearer. No behavior change was intended.
[kpouer] added a new line at the end of the file
[k_satoda] Removed unnecessary use of List on glyphs field in Chunk for a bit less
use of memory and simpler code.
[jarekczek] Fixed bug #3493415, an error messsage on double-clicking multiple directories.
Artifacts can be downloaded here: