Hi Rene, Rene Engelhard <r...@debian.org> ezt írta (időpont: 2020. aug. 7., P, 17:31):
> Hi, > > my LO 7.0.0 build on Debian buster failed: > > [...] > Test name: SwUiWriterTest::testTdf133589 > equality assertion failed > - Expected: 𐳥𐳋𐳓𐳉𐳗 > - Actual : székely 0 > > Failures !!! > Run: 305 Failure total: 1 Failures: 1 Errors: 0 > > This looks like libnumbertext needs to be 1.0.6 which does have this new > "old hungarian" feature... (yes, "of course" --with-system-libnumbertext > and libnumbertext in Debian buster is 1.0.5). > Ah, indeed. I didn't think of this option... > > $ git diff > diff --git a/configure.ac b/configure.ac > index 59ec7ec58bf1..d5782d32be4b 100644 > --- a/configure.ac > +++ b/configure.ac > @@ -10509,7 +10509,7 @@ AC_SUBST(SYSTEM_LIBEXTTEXTCAT_DATA) > dnl =================================================================== > dnl Checking for libnumbertext > dnl =================================================================== > -libo_CHECK_SYSTEM_MODULE([libnumbertext],[LIBNUMBERTEXT],[libnumbertext > >= 1.0.0]) > +libo_CHECK_SYSTEM_MODULE([libnumbertext],[LIBNUMBERTEXT],[libnumbertext > >= 1.0.6]) > if test "$with_system_libnumbertext" = "yes"; then > SYSTEM_LIBNUMBERTEXT_DATA=file://`$PKG_CONFIG --variable=pkgdatadir > libnumbertext` > SYSTEM_LIBNUMBERTEXT=YES > > The other alternative is to do it like the DLP libraries and only do the > test needing it if libnumbertext is 1.0.6 and thus support "old > hungarian"... > > Which one would you prefer? The first one of course is trivial :-) > I prefer the trivial one, if it's not a problem for you. I gave a +2 for it, thanks! :) A question related to this issue: I started to add Old Hungarian numbering to Writer (https://gerrit.libreoffice.org/c/core/+/99950) and the UI will contain Old Hungarian numbers, see the attached screenshot, so I plan to update the noto-fonts-20171024.tar.gz tarball, adding also the fortunately small (45 kB) NotoSansOldHungarian-Regular.otf to help to fix the potential font issue [note: the screenshot contains another problem: showing squares at the place of 1-character non-Unicode BMP text spans]. There is a bug report for Linux Mint ( https://github.com/linuxmint/cinnamon/issues/9458), also an older one for Debian ( https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=908817). I am going to file a bug for my LibreOffice patch related to the numbering, maybe for Noto font update in LibreOffice. Do you suggest something similar for Debian (Ubuntu) etc., or it is not necessary? Thanks, László > > Regards, > > Rene >
_______________________________________________ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
