Age | Commit message (Collapse) | Author |
|
added LanguageTag::equals(), fdo#62278 related
(cherry picked from commit fb26196ce64979ab6dc6c82c0fbdeedb126a3433)
Conflicts:
i18npool/inc/i18npool/languagetag.hxx
i18npool/source/languagetag/languagetag.cxx
resolved fdo#62278 test for LanguageTag equality resolving system locale
(cherry picked from commit dde0dcd553b676abcb41eaf3f9f1bf797907e22f)
Change-Id: I7c6e91518b3afd7a42d21bcae41283b0e00b833a
Reviewed-on: https://gerrit.libreoffice.org/2707
Reviewed-by: Fridrich Strba <fridrich@documentfoundation.org>
Tested-by: Fridrich Strba <fridrich@documentfoundation.org>
|
|
fdo#60259 prevent crash when searching backward for $ anchor regex
Old code wasn't prepared that searching for $ may actually return a
result set pointing behind the search string which it does with the ICU
regex engine.
(cherry picked from commit c00601dab0f5533b152cd63cec0a89bfec1ba95f)
Conflicts:
i18npool/source/search/textsearch.cxx
regex: handle zero-length matches, fdo#60259 related
Also in backward search ignore all zero-length matches except the text
end single $ anchor search. The anchor search is a valid match, treat it
as such in Writer.
This still doesn't solve the backward $ backward search, the convoluted
Writer code in that place apparently never worked, someone more familiar
with those internals should straighten out the mess.
(cherry picked from commit 3bc5cb3c485d67f1ce0541d349d11637f52ebda5)
regex: don't loop 10000 identical matches to find a single $ anchor
(cherry picked from commit ccc349d3abb70ef38cd2b7706da51b060a385908)
make forward replacement of $ work again, fdo#60259 related
broken with 3bc5cb3c485d67f1ce0541d349d11637f52ebda5
(cherry picked from commit d8dcfa0e5dbecf77c4d6a8d49caf61b339cf9b43)
Change-Id: I6b5eb28d0a54ceecb6873a3c05f18f70312ab1a2
Reviewed-on: https://gerrit.libreoffice.org/2679
Reviewed-by: Fridrich Strba <fridrich@documentfoundation.org>
Tested-by: Fridrich Strba <fridrich@documentfoundation.org>
|
|
i#118925# enhance textsearch's match-group references
to work for look-ahead/look-behind
(cherry picked from commit 3b83c404c56e5db5bab29ffee41f02822410d625)
Conflicts:
sw/source/core/crsr/findtxt.cxx
(cherry picked from commit 9a93475d6eba53b2e1fba1585dbd11c94ea4b4a3)
Conflicts:
sw/source/core/crsr/findtxt.cxx
i#120598 better emulation of regexp word-start and word-end operators
The emulation of the word-start and word-end operators provided
the previous regexp engine can be approximated much better
by using the ICU-regexp exngines powerful look-around feature.
Patch-by: Herbert Duerr
Found-by: ldgolds33@yahoo.com
(cherry picked from commit ec7ef30693f10315ce80a8f5d7325a0e40855e66)
(cherry picked from commit 8c26876fea085a1bc847abba63dffa97a9499c1d)
i#121482# fix attributed text search for regular expression patterns
(cherry picked from commit e7fc662799e7e936753e24db8d6d3849c12b3ff4)
(cherry picked from commit e6288a5d889da7db5bf23174f85c29ccfcaa44d5)
i#121482# fix backwards regexp search for matches overlapping search start
(cherry picked from commit 854f4ad6c57be62bd922df08f603d8bfb7b918a5)
Additionally fixed unit test, searching backward shall not produce a
different result from searching forward. (erAck)
(cherry picked from commit b514f0ce86e85d9be269ddf2e797befbbf3423f1)
i#121633# fix search for upper-case character classes
in ignore-case regular expressions
(cherry picked from commit b7ee1803453d3a766ce3a833892e1c208aacf8ff)
(cherry picked from commit 531538892795bec909bb8baff6bbf9e19baf809e)
i#121633# fix ignore-case problem caused by i18nsearch API mess
(cherry picked from commit 7644ec176049937b588fe171a553c9a07b375792)
(cherry picked from commit f932a3f1195290f9aa37b593190bd4c6ac5fe2f6)
Change-Id: I44d6216b12f17d0560c4e8cf355937797ddeee2a
Reviewed-on: https://gerrit.libreoffice.org/2678
Reviewed-by: Fridrich Strba <fridrich@documentfoundation.org>
Tested-by: Fridrich Strba <fridrich@documentfoundation.org>
|
|
Change-Id: Ic07d9147ad559573e4e6a3818ee78b8dfcedf12f
Reviewed-on: https://gerrit.libreoffice.org/2482
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>
(cherry picked from commit b3f43db9e998f5656144349a7bb2cf2cf2be68ba)
|
|
Long year was forced unconditionally for all non-Gregorian calendars, do
this only for ROC calendar.
Change-Id: I39d6eb269b3d38046d32dcf0f66edd5617581b9d
(cherry picked from commit be16dba682b8ab1c244f8d24f1bab3a539096962)
Reviewed-on: https://gerrit.libreoffice.org/2474
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
Change-Id: Ia7c5692e1da4d24d985f0813e07631898d7100ce
Reviewed-on: https://gerrit.libreoffice.org/2459
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>
|
|
While 'Kazakh' can be written as 'Kazak' or 'Qazaq', the 'Kazakh' form
is the most widespread according to online sources (Wikipedia,
LCID list from Microsoft). Also the 'Kazakh' form is preferred by a
Kazakh contributor, see http://listarchives.libreoffice.org/global/l10n/msg05444.html
In LibreOffice both 'Kazak' and 'Kazakh' was used at different places.
Change-Id: Ief83eb05e92af16cb5b795f2bbb1e3d6f7c9659a
Reviewed-on: https://gerrit.libreoffice.org/1876
Tested-by: Eike Rathke <erack@redhat.com>
Reviewed-by: Eike Rathke <erack@redhat.com>
|
|
Change-Id: I5293fea9b5404b82e72761407d325c408a2e45ca
(cherry picked from commit bbf2f413958e7be2bef34c62932fc76f83e7ed18)
Signed-off-by: David Tardon <dtardon@redhat.com>
|
|
Change-Id: I8c0da25a000e10d65479665a61be054fdb2486a5
Reviewed-on: https://gerrit.libreoffice.org/1453
Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: Tomáš Chvátal <tchvatal@suse.cz>
|
|
Change-Id: I7313cca3335e1b41df2a70fe968bc1e6386a1e36
(cherry picked from commit 896e326ace53bedf9b17fa624d260032b91bcfd1)
Signed-off-by: Eike Rathke <erack@redhat.com>
|
|
Doing word-count by switching per-word between two different
kinds of word iterator was insanely slow. This preserves an
ICU break-iterator for each type of word-breaking.
|
|
Change-Id: Ie212f2e91b3ef262997bcd125817769ae24c16c3
(cherry picked from commit fff4d120866a0be3cd8185f2c67bb9f59b1a6a3f)
|
|
Various places use an empty locale or string to indicate different
meanings and/or variants of "this is not a real locale but I didn't know
what to use else" or abuse LANGUAGE_DONTKNOW in one or the other way. In
preparation of changing that awkward situation now offer the ISO 639
codes for multiple languages [mul] and language undetermined [und],
mapping to LANGUAGE_MULTIPLE and LANGUAGE_UNDETERMINED.
Change-Id: I687de23ffc00a9a056f2837b024e0a62658e3df2
(cherry picked from commit 09bc1464ec9dde61e69cca393e65e72143bdd383)
|
|
For "simple" known (to us) locales avoid the overhead of liblangtag,
also in preparation of not having to initialize its database during
startup.
This also enables recognition of the 'qtz' pseudolocale for key ID
resources again, see fdo#57413
Change-Id: Id52a1ddc37b42063e3bf68a6dbeaeacfbfa704ef
(cherry picked from commit 4d24507451fbc2b7d72cf5f7048a424d52b63850)
|
|
As per https://bugs.freedesktop.org/show_bug.cgi?id=52240#c59
Change-Id: I9f5ba81c5fd6732ca7e90e0480fcc02ed2709674
(cherry picked from commit b1184f8a4e4a3949d38407c5f1d5155c2c2b42d5)
|
|
Also referenced by nb-NO and ny-NO
Changes as per http://erack.org/blog/archives/18-comments.html#c351
Change-Id: I9e07f4307800e99fdff21bccde91af4178b244f3
(cherry picked from commit 638993d9b80b3793dfd31db0dacba3746ceb98fd)
|
|
Change-Id: I80af0399037e4f68113338139e7f2ad2400e65ab
|
|
Change-Id: Ic4794d9a908b60220a4a849ff263eaa08776550c
|
|
Change-Id: Icacd44abfa5bf5f75f0ee719b39fba6821976a8a
|
|
Change-Id: I683c54ed278d5a6ba12eaefb8040c3753a24f119
Signed-off-by: Paula Mannes <paulamannes@gmail.com>
Reviewed-on: https://gerrit.libreoffice.org/1204
Reviewed-by: Olivier Hallot <olivier.hallot@alta.org.br>
Tested-by: Olivier Hallot <olivier.hallot@alta.org.br>
|
|
Change-Id: I195506b32a5cbf751ac90ad1ab13680097c8c428
|
|
Change-Id: I7c7fb38d7df5aecda8e81443aa7045479e0efceb
|
|
Change-Id: I0736670512dd54e67fc08b65b9fad56ab904c92d
|
|
Changes to commit a81a9109ed43e0b3880c9bd88e58f2b21065226b
The ..._CONGO identifier and "... (Congo)" resource string changes are
only to match the current convention that if a locale is not of the
default country of a language we mention the country. This helps in case
we later add the other locale of a language. Having two Lang-IDs then
with different sublanguage and adding the ISO locales is good practice.
For axk-CG it should be LANGUAGE_USER_YAKA_CONGO instead of
LANGUAGE_USER_AKA and the resource string should be "Aka (Congo)",
because it is not the default axk-CF (Central African Republic), see
http://www.ethnologue.com/show_language.asp?code=axk
For kng-CG it should be LANGUAGE_USER_KIKONGO_CONGO instead of
LANGUAGE_USER_KIKONGO and the resource string should be "Kikongo
(Congo)", because it is not the default kng-CD (Democratic Republic of
the Congo), see http://www.ethnologue.com/show_language.asp?code=kng
However, changed name to Koongo instead of Kikongo because that is the
default name even for Congo, see
http://www.ethnologue.com/show_language.asp?code=kng, and used
LANGUAGE_USER_KOONGO_CONGO instead.
For njy-CG it should be LANGUAGE_USER_NJYEM_CONGO instead of
LANGUAGE_USER_NJYEM and the resource string should be "Njyem (Congo)",
because it is not the default njy-CM (Cameroon), see
http://www.ethnologue.com/show_language.asp?code=njy
For puu-CG it should be LANGUAGE_USER_PUNU_CONGO instead of
LANGUAGE_USER_PUNU and the resource string should be "Punu (Congo)",
because it is not the default puu-GA (Gabon), see
http://www.ethnologue.com/show_language.asp?code=puu
For yom-CG it should be LANGUAGE_USER_YOMBE_CONGO instead of
LANGUAGE_USER_YOMBE and the resource string should be "Yombe (Congo)",
because it is not the default yom-CD (Democratic Republic of the Congo),
see http://www.ethnologue.com/show_language.asp?code=yom
For each of these ..._CONGO lang-IDs introduced the default ID and made
the ..._CONGO ID with sublanguage 0x20, adapted corresponding IDs in
*.xml files.
All IDs in those *_CG.xml files needed to be corrected anyway as they
all had a replaceTo="[$CFA-63A]" copied from sg_CF.xml, the ID needs to
match that of the LangID instead.
As all LC_FORMAT element section were identical, made them refer to
kng_CG instead of having duplicated copies, same for LC_CURRENCY.
Removed the executable permission bits from *_CG.xml files.
Change-Id: Iae06af5fbf19e3a3c83a4f4b4dcb857213df629e
|
|
Change-Id: I9f65022cee5dc0201ab45192b3731115ba50f27d
|
|
Change-Id: If6ad17fa9e274beff7ba872a095ced65438962af
|
|
LANGUAGE_NONE was resolved to UI language, don't do that, if desired use
LANGUAGE_HID_HUMAN_INTERFACE_DEVICE instead.
Change-Id: I4a45653c9dfef7cc2cddb941ea1174189708cea2
|
|
Change-Id: I2e6800d0f5a6346fc743d5a57bab34d1e36b4ee4
|
|
Change-Id: I092ca8f912e26f0743909920c6e740d648b8677e
|
|
...let BRAND_BASE_DIR point to solver's unittest/installation during CppUnit
tests, and get rid of LIBLANGTAG_SHARE special case. (Intending to add further
code to CppUnit tests soon that requires a fake BRAND_BASE_DIR, so clean this
up now.)
Change-Id: Ia5f0f247e54952f901c37ca31eda53f9c6591458
|
|
Change-Id: Ie5423578e34a295aad3a03d2b29f3a6b1a6572ce
|
|
Change-Id: I4f339ecaeb147a05f8882065e859902f8c14ac47
|
|
Change-Id: Iec2845cb3717520e1126ac4e9743413d5f3a5119
|
|
Change-Id: I0134593f8ba664b7a8e8a14f54485b96b749418d
|
|
Change-Id: I4dad05dbbe668ac7d7058cb43989f6a9f8e488a6
Reviewed-on: https://gerrit.libreoffice.org/1118
Reviewed-by: Luboš Luňák <l.lunak@suse.cz>
Tested-by: Luboš Luňák <l.lunak@suse.cz>
|
|
Change-Id: I6a89dd2c5e34035ac1e6c9b7e4723d881c5ceaa9
|
|
Change-Id: I8917c2958f021f11933d9da7fec8ef01609387c8
|
|
Change-Id: Id78a989ab981d658dd8f331b030e00ce201c8bc9
|
|
And use the rtl::Static pattern for statics.
Change-Id: I80d8cc0a4a35771f86bec27edf41224d71cdea14
|
|
Added bResolveSystem=true parameter to getBcp47(), getLocale() and
getLanguageType(). Other get...() and is...() methods now always resolve
to system locale.
Change-Id: I2d9718b8bd36aac5e047afd6921d462e52c6a235
|
|
|
|
|
|
|
|
|
|
introducing new Currency element and marking previous as legacyOnly
Change-Id: I40a762bf03a15f241fc80279bb61be9a2710d2f0
Reviewed-on: https://gerrit.libreoffice.org/967
Reviewed-by: Lior Kaplan <kaplanlior@gmail.com>
Tested-by: Lior Kaplan <kaplanlior@gmail.com>
Reviewed-by: Eike Rathke <erack@redhat.com>
|
|
|
|
Change-Id: I2fe03d438f52c31d0615f124f4d0a7ced2136928
|
|
ported from the broken java ver.
adds a native ICU regex sanity test.
|
|
|
|
Patches from Herbert Duerr:
#i118925# fix old word-boundary emulation
http://svn.apache.org/viewvc?view=revision&revision=1301596
#i119031# preserving ASCII in ignore-* transliterations allows full SIMPLE_TRANS_MASK
http://svn.apache.org/viewvc?view=revision&revision=1301104
#i118887# ignore zero-length matches in regexp search
http://svn.apache.org/viewvc?view=revision&revision=1242235
fix transliteration flags for search pattern preprocessing
http://svn.apache.org/viewvc?view=revision&revision=1236199
fix minor typo in comment
http://svn.apache.org/viewvc?view=revision&revision=1234799
improve standard conformance regarding unicode UAX#29 word boundaries
http://svn.apache.org/viewvc?view=revision&revision=1234786
emulate word boundary matching of old regex engine
The new ICU regex engine has much improved unicode capabilities.
The old regex engine had the extensions \< and \> for matching word boundaries.
For the convenience of a smooth upgrade experience these artifacts now get mapped
to \b which is supported by almost all regex engines.
http://svn.apache.org/viewvc?view=revision&revision=1234777
#i118723# fill in regexp matcher group details
http://svn.apache.org/viewvc?view=revision&revision=1228084
use ICU regexp instead of LGPL i18nregexp
http://svn.apache.org/viewvc?view=revision&revision=1177610
#i119031# default to preserve ASCII in ignore-width transliteration
http://svn.apache.org/viewvc?view=revision&revision=1301101
|