[Koha-bugs] [Bug 35024] New: PO files wrap differently depending on versions

[bugzilla-daemon at bugs.koha-community.org]

https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=35024

            Bug ID: 35024
           Summary: PO files wrap differently depending on versions
 Change sponsored?: ---
           Product: Koha
           Version: unspecified
          Hardware: All
                OS: All
            Status: ASSIGNED
          Severity: normal
          Priority: P5 - low
         Component: Architecture, internals, and plumbing
          Assignee: jonathan.druart+koha at gmail.com
          Reporter: jonathan.druart+koha at gmail.com
        QA Contact: testopia at bugs.koha-community.org

Looks like (thanks Julian!) it's coming from libunistring, between 0.9.10 and
1.0 I am seeing a change in how the strings in the PO files are wrapped.

Debian 11 (using libunistring 0.9):
"%s %sPrintemps%s %s %sÉté%s %s %sAutomne%s %s %sHiver%s %s %sPrtps%s %s %sÉté"
"%s %s %sAut%s %s %sHiv%s %s %s %s"

Debian 12, Ubuntu 22.04(using libunistring 1.0)
"%s %sPrintemps%s %s %sÉté%s %s %sAutomne%s %s %sHiver%s %s %sPrtps%s %s "
"%sÉté%s %s %sAut%s %s %sHiv%s %s %s %s"

It *could* come from this changes:
https://git.savannah.gnu.org/gitweb/?p=libunistring.git;a=blob;f=NEWS;h=5a43ddd7011d62a952733f6c0b7ad52aa4f385c7;hb=HEAD
   * The functions u*_possible_linebreaks and u*_width_linebreaks now make it
     easier to work with strings that contain CR-LF sequences: In this case,
     in the returned array, it will return UC_BREAK_CR_BEFORE_LF followed by
     UC_BREAK_MANDATORY (instead of twice UC_BREAK_MANDATORY).

The command used is:
% msgmerge --backup=off --quiet -F --update
misc/translator/po/fr-FR-staff-prog.po misc/translator/Koha-staff-prog.pot

No matter the value of --width.

-- 
You are receiving this mail because:
You are watching all bug changes.