The gettext.install() functions supplies a way to force it to return
unicode which helps with compatibility between Python 2 & 3. To make it
work properly when merging those returned string with strings in the
different modules of gitinspector we also had to do some changes to make
sure all string literals are in unicode.
The name of the author is now highlighted in the table whenever a pie
piece in the pie chart is hovered over. Currently, it does not work the
other way around (because of a limitation in the flot library).
This particular setup script has been tested with Distribute 0.6.24.
For some strange reason, the generated entry-point script does not work
with Python 3.3 - resulting in import errors. I can only assume this is
a bug in this version of the Distribute package when it's running together
with Python 3.3.
Even though it is supported, PHP is not part of the default file extensions
used during analysis. To include PHP files in the statistical analysis,
the extension needs to be supplied to gitinspector using the -f flag.
The required Python version is 2.6. The check is done after all the
command-line arguments have been parsed. In this way, it is still possible
to get help and version information from the script even if the script is
not able to execute properly.
As we don't really need it, storing this was just eating up unnecessary
memory. On very big repositories (with many commits) removing it makes a
big difference.
Instead of manually adding this functionality with our own code; we use
the JQuery tablesorter plugin which adds support for sorting with a few
simple rows of code.
The printed data was using instead of spaces when printing out all
extensions found. This resulted in the browser not being able to properly
separate rows.
In practice, this means that the tables are initially collapsed and hide
all authors with a workload below one percent. The user can expand a
table in order to show all authors found during statistical analysis.
In practice, this means that whenever an author is not part of a
collection of periods he will not be printed out.
This cleans up the output and also speeds up analysis.
The output was simply wrong and was outputting extra "<authors>" tags all
over the place. Also, whenever there was very little work, this was not
included in the timeline.
Multi-line POD comments in Perl are not supported for now; they are a
little hairy and probably require a little extra work in order to be
parsed correctly.
