Issue 80299 - Add hungarian localized keywords to CELL and INFO
Summary: Add hungarian localized keywords to CELL and INFO
Status: CLOSED FIXED
Alias: None
Product: Calc
Classification: Application
Component: code (show other issues)
Version: 680m223
Hardware: All All
: P3 Trivial (vote)
Target Milestone: ---
Assignee: frank
QA Contact: issues@sc
URL:
Keywords: ms_interoperability
Depends on:
Blocks:
 
Reported: 2007-08-02 21:59 UTC by kyoshida
Modified: 2013-08-07 15:14 UTC (History)
4 users (show)

See Also:
Issue Type: ENHANCEMENT
Latest Confirmation in: ---
Developer Difficulty: ---


Attachments
keyword text file transferred from Issue 77996. (798 bytes, text/plain)
2007-08-02 22:00 UTC, kyoshida
no flags Details
test document (11.12 KB, application/vnd.sun.xml.calc)
2007-09-06 04:25 UTC, kyoshida
no flags Details
revised test document (11.33 KB, application/vnd.oasis.opendocument.spreadsheet)
2007-10-02 16:44 UTC, kyoshida
no flags Details
Are you sure? COORD, FORMAT, PARENTHESES, SHEET and TYPE seem to work... (7.94 KB, application/vnd.oasis.opendocument.spreadsheet)
2007-10-03 19:10 UTC, timar74
no flags Details
revised test document to incorporate the missing Hungarian keywords. (11.94 KB, application/vnd.sun.xml.calc)
2007-10-03 21:38 UTC, kyoshida
no flags Details
Testdocument (69.00 KB, application/vnd.ms-excel)
2007-10-18 11:37 UTC, frank
no flags Details

Note You need to log in before you can comment on or make changes to this issue.
Description kyoshida 2007-08-02 21:59:08 UTC
This is an extension of Issue 77996.  Add the hungarian keywords to CELL and
INFO functions for Excel compatibility.
Comment 1 kyoshida 2007-08-02 22:00:16 UTC
Created attachment 47276 [details]
keyword text file transferred from Issue 77996.
Comment 2 kyoshida 2007-08-02 22:00:45 UTC
adding timar to CC.
Comment 3 kyoshida 2007-08-02 22:03:01 UTC
adding keyword.
Comment 4 ooo 2007-08-05 12:35:15 UTC
Note that including non-ASCII characters needs some changes. A static const
sal_Char table, String::CreateFromAscii() to insert to the map, and
String::ToUpperAscii() before calls won't do anymore. CharClass::toUpper()
should be used instead (see unotools/charclass.hxx), accessible via
ScGlobal::pCharClass, and the table been kept as sal_Unicode in UCS2 encoding.
Comment 5 kyoshida 2007-09-06 04:25:52 UTC
Created attachment 47998 [details]
test document
Comment 6 kyoshida 2007-09-06 05:11:48 UTC
Work is ongoing in the celltrans02 CWS.
Comment 7 kyoshida 2007-09-06 15:18:12 UTC
@KAMI or timar

Could one of you take the test document I have attached, save it as xls, and
open it in the Hungarian version of Excel?

I want to make sure that those localized keywords *do* work in the Hungarian
version of Excel.

Thanks
Kohei
Comment 8 kyoshida 2007-09-06 18:27:23 UTC
The implementation is complete in celltrans02 CWS.  The patch has been comitted
to ooo-build trunk first.
Comment 9 kyoshida 2007-09-10 14:43:43 UTC
I'm going to resolve this issue as FIXED.

@kami_, timar

Please reopen this if there is any issues.
Comment 10 kyoshida 2007-09-10 16:14:56 UTC
reopening it for qa.
Comment 11 kyoshida 2007-09-10 16:15:27 UTC
re-assigning it to fst for qa.
Comment 12 frank 2007-10-02 09:57:33 UTC
Hi Kohei,

I need at least two installation sets to proceed with the QA work for this CWS.
Preferedare Windows and Solaris. At least Windows and Linux are a must have.

Please provide me such Installationsets.

Frank
Comment 13 timar74 2007-10-02 10:42:02 UTC
Hi Kohei,

Thanks for your work. I found one issue in the test document:
FILENAME parameter of CELL should be FILENÉV not FÃJLNÉV (btw. Hungarian Excel
Help is buggy in this regard). The rest of the translations work well - even
those that you left out from the test file (LAP, KOORD, TÃPUS, FORMA, ZÃRÓJELEK).

I volunteer to create Windows and Linux builds for this CWS.

Thanks,
Andras
Comment 14 kyoshida 2007-10-02 16:21:33 UTC
Andras,

Thanks for testing this, and thank you very much for volunteering to build the
Windows and Linux builds! :-)  You made my day.

Let me put the fix to address the FILENÉV keyword issue, and I'll let you know
when it's done.
Comment 15 kyoshida 2007-10-02 16:44:41 UTC
Created attachment 48654 [details]
revised test document
Comment 16 kyoshida 2007-10-02 17:49:19 UTC
Andras,

The CWS is ready.  So, please go ahead with the build.  Again, thank you very
much for your help!

Frank, please expect the builds from Andras.
Comment 17 timar74 2007-10-03 11:33:03 UTC
en-US builds are ready to download from here:
ftp://ftp.fsf.hu/OpenOffice.org_hu/devel/SRC680_m231+cws_celltrans02/

Kohei, may I ask you a question. There can be 14 parameters of CELL function.
Why did you implement only 9? Translation of COORD, FORMAT, PARENTHESES, SHEET
and TYPE are missing. 
Comment 18 kyoshida 2007-10-03 14:12:14 UTC
Hi Andras,

>There can be 14 parameters of CELL function.  Why did you implement only 9?
Translation of COORD, FORMAT, PARENTHESES, SHEET and TYPE are missing.

The reason is that Calc doesn't offer functionality for those keywords even in
English, so there is no point in translating them (you'll get Error result anyway).

In future if Calc decides to implement these missing parameters, then it'll make
sense to translate them at that point.
Comment 19 timar74 2007-10-03 19:10:36 UTC
Created attachment 48679 [details]
Are you sure? COORD, FORMAT, PARENTHESES, SHEET and TYPE seem to work...
Comment 20 kyoshida 2007-10-03 19:41:42 UTC
@timar

Yes, you're right!  How did I miss that!

Let me add those translations that you provided.  Sorry about this....

Kohei
Comment 21 kyoshida 2007-10-03 21:38:10 UTC
Created attachment 48686 [details]
revised test document to incorporate the missing Hungarian keywords.
Comment 22 kyoshida 2007-10-03 21:48:45 UTC
Ok.  I've just committed into celltrans02 CWS the additional Hungarian keywords
that Andras pointed out as missing.  Thanks for pointing it out!
Comment 23 kyoshida 2007-10-03 21:52:16 UTC
When rebuilding, you need to touch sc/source/core/tool/cellkeytranslator.cxx for
the keyword change to take effect.  This is because the keywords are encoded in
sc/source/core/tool/cellkeywords.inl which is autogenerated by a script. 
Because this file is not in makefile.mk, a change in this file will not be
noticed by the build system.  So you need to touch cellkeytranslator.cxx to
force it to be re-compiled.

Kohei
Comment 24 timar74 2007-10-04 13:11:47 UTC
Looks good now, thanks. Builds have been updated at
ftp://ftp.fsf.hu/OpenOffice.org_hu/devel/SRC680_m231+cws_celltrans02/
Comment 25 frank 2007-10-09 11:10:39 UTC
Hi all,

sorry for the delay, had a tight schedule.

@timar

Thanks for building the packages. Will test them soon.

Frank
Comment 26 ooo 2007-10-09 11:48:13 UTC
@kohei:
> When rebuilding, you need to touch sc/source/core/tool/cellkeytranslator.cxx

That's awkward. The #include "cellkeywords.inl" isn't caught by the dependency
generator? e.g. sc/unxlngi6.pro/misc/all_tool.dpslo doesn't have a
../../../unxlngi6.pro/slo/cellkeytranslator.obj entry for it? Not even after a
dmake depend=true
in sc/source/core/tool and another subsequent dmake?

  Eike
Comment 27 kyoshida 2007-10-09 12:19:24 UTC
@er: yes, it's awkward.  Do you have any idea how to make it less awkward? :-)
Comment 28 ooo 2007-10-09 15:56:30 UTC
If the dependency isn't generated from the include statement it sounds like a
bug of makedepend to me. Cc'ing Ause 'hjs' for this.

Having fiddled around with it a bit it seems that makedepend ignores includes
that don't start at the very first column, e.g.
#include "lala.inl"
works fine and generates a dependency, while
  #include "lala.inl"
doesn't.
Comment 29 ooo 2007-10-09 16:56:05 UTC
For the records: issue 82442.
Comment 30 frank 2007-10-10 15:21:19 UTC
Hi kohei, timar,

a lot of the required autotests have been failed due to the fact that the binary
file filters are broken or not present. Please check that the fix itself does
not corrupt the binary file filters. I don't expect a problem here but to be
sure, you know.

Timar could you please check this and build new installation sets ? I've used
Windows and Linux build. Both do show this problem. Without a proper fixing I
can't pass this CWS over to RE for integration as we could not be sure what
caused this problem.

Thanks.

Frank
Comment 31 kyoshida 2007-10-10 15:26:35 UTC
The code doesn't even remotely touch the filter code.  The major portion of the
code change is well isolated, and it only affects two built-in functions.

So, corrupting the binary filter code is not likely caused by this change.  At
least I don't think so.

Kohei
Comment 32 timar74 2007-10-10 16:50:40 UTC
I compiled with --disable-binfilter to save compilation time. Please re-confirm
if you still need new builds.
Comment 33 frank 2007-10-11 08:22:48 UTC
Hi Timar,

yes, I need new Installationssets with binfilter compiled because if we do not
run the autotests completly, we can not be sure that nothing is broken. This is
a must for cws handling and the cws gatekeeper will reject this cws if we do not
proceed as described.

Thanks for building the inst sets.

Frank
Comment 34 timar74 2007-10-14 07:55:32 UTC
I uploaded builds with binfilter to
ftp://ftp.fsf.hu/OpenOffice.org_hu/devel/SRC680_m231+cws_celltrans02
Comment 35 frank 2007-10-15 10:37:27 UTC
thanks a lot
Comment 36 frank 2007-10-18 11:37:38 UTC
Created attachment 48998 [details]
Testdocument
Comment 37 frank 2007-10-18 12:01:05 UTC
found fixed on cws celltrans02
Comment 38 frank 2007-12-14 11:49:52 UTC
found integrated on master m239 using Linux, Solaris and Windows build