Po parsing now works directly with an array of lines. The previous

Po parsing now works directly with an array of lines. The previous
version split the entire file contents into an array of lines, only
to join subarrays into strings, which it would pass to the unit parser,
which in its turn would split the strings into arrays of lines. This
inefficiency was remedied by simply factoring out the line array
handling code of the unit parsing code and calling that directly
on the main array of lines (from the file parsing method).

git-svn-id: http://translate.svn.sourceforge.net/svnroot/translate/src/trunk/translate@7518 54714841-351b-0410-a198-e36a94b762f5