[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-08-02 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

  Status|VERIFIED  |CLOSED





--- Additional comments from [EMAIL PROTECTED] Thu Aug  2 14:46:09 + 
2007 ---
Closing.


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-05 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

  Status|RESOLVED  |VERIFIED





--- Additional comments from [EMAIL PROTECTED] Fri Jan  5 06:34:16 -0800 
2007 ---
verified.

Patch is present in pj69.


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-04 Thread sb
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905





--- Additional comments from [EMAIL PROTECTED] Thu Jan  4 02:33:58 -0800 
2007 ---
"clytie: please file new issue for the new bug in encodings in Windows installer
strings."  Issue 73103.

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-03 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

  Status|REOPENED  |RESOLVED

  Resolution|  |FIXED





--- Additional comments from [EMAIL PROTECTED] Wed Jan  3 10:35:58 -0800 
2007 ---
sb: agreed.

clytie: please file new issue for the new bug in encodings in Windows installer 
strings.

I'll mark this issue as fixed, because the encoding is changed in pj69.


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-03 Thread sb
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905





--- Additional comments from [EMAIL PROTECTED] Wed Jan  3 05:52:25 -0800 
2007 ---
The conversion tables for MS1258 in sal/textenc/tcvteas1.tab:1.7 appear to be
somewhat naive in that they do not support composed Unicode characters like
U+1ECB LATIN SMALL LETTER I WITH DOT BELOW.  What works is to map from Unicode

  U+0069 LATIN SMALL LETTER I  U+0323 COMBINING DOT BELOW

to MS 1258

  0x69 0xF2

but what fails is to map from Unicode

  U+1ECB LATIN SMALL LETTER I WITH DOT BELOW

to MS 1258

  0x69 0xF2.

In the other direction, MS 1258

  0x69 0xF2

is translated to

  U+0069 LATIN SMALL LETTER I  U+0323 COMBINING DOT BELOW

instead of the (functionally equivalent)

  U+1ECB LATIN SMALL LETTER I WITH DOT BELOW

(and in light of that, the current mnMaxCharSize = mnAveCharSize = 1 does make
sense).

I doubt this problem has been addressed for any encoding in sal/textenc that has
combining characters like 0xF2 in MS 1258 (though MS 1258 might be the only such
encoding; I did not bother to check).  Also, this problem touches on the issue
of handling canonical-equivalent Unicode sequences (see
 and
).

The conversion tables for MS1258 in sal/textenc/tcvteas1.tab:1.7 are based on
,
which does not directly support composed Unicode characters, either.  Does
anybody have data files exactly which (composed) Unicode characters should be
mapped to which MS 1258 byte sequences?  If yes, I could adapt
sal/textenc/tcvteas1.tab:1.7 accordingly.

[Anyway, I think this problem should be addressed in a fresh issue, linking back
to this original issue for context.]

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-03 Thread timar
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905





--- Additional comments from [EMAIL PROTECTED] Wed Jan  3 04:29:33 -0800 
2007 ---
Eike, I think you are right. On the other hand "sal_uInt8 mnMaxCharSize" and
"sal_uInt8 mnAveCharSize" are both defined as 1 in al/textenc/tcvteas1.tab#331
and #332 which may be wrong. For example Vietnamese letter ị (UTF8 E1,BB,8B) is
encoded as i + combining dot below (69 + F2) in CP1258.

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-03 Thread er
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User er changed the following:

What|Old value |New value

  CC|'er,is,timar' |'er,is,sb,timar'





--- Additional comments from [EMAIL PROTECTED] Wed Jan  3 03:29:35 -0800 
2007 ---
Text encodings and conversions are maintained by 'sb' Stephan Bergmann (CC'ed).
However, I don't think the

sal/textenc/tcvteas1.tab#334
"iso8859-1", /* TODO! correct? */

is related to the Windows installer problem, as that field of
ImplTextEncodingData is declared as
char const * mpBestUnixCharset;

  Eike

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2007-01-02 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

  CC|'is,timar'|'er,is,timar'

  Status|RESOLVED  |REOPENED

  Resolution|FIXED |





--- Additional comments from [EMAIL PROTECTED] Tue Jan  2 23:55:54 -0800 
2007 ---
so: reopening this issue.

Eike: who takes care about this area?


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-28 Thread timar
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905





--- Additional comments from [EMAIL PROTECTED] Thu Dec 28 12:26:21 -0800 
2006 ---
ihi,
ulfconv is used here: http://go-oo.org/lxr/source/tools/solenv/inc/rules.mk#692

In ulfconv there is a call of OUStringToOString which actually converts the
strings from Unicode to RTL_TEXTENCODING_MS_1258, if codepage 1258 is selected
for "vi" in msi-encodinglist.txt. The character encoding for 
RTL_TEXTENCODING_MS_1258 is defined in sal/textenc/tcvteas1.tab - probably this
is not correct, but I do not know. The following line looks suspicious:
http://go-oo.org/lxr/source/porting/sal/textenc/tcvteas1.tab#334
"iso8859-1", /* TODO! correct? */
As clytie wrote, this is what we saw. Only the latin1 subset of cp1258 was
converted.

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-27 Thread ihi
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User ihi changed the following:

What|Old value |New value

  CC|'timar'   |'is,timar'





--- Additional comments from [EMAIL PROTECTED] Wed Dec 27 10:45:44 -0800 
2006 ---
ingo, please have a look. As I remember right the ulfconv tool is not used for
the GUI installer strings. I found only a reference to the licensefiles ... but
I am not a installer expert.

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-25 Thread clytie
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905





--- Additional comments from [EMAIL PROTECTED] Mon Dec 25 23:54:35 -0800 
2006 ---
Note that the Latin-1 (ISO 8859-1) character set is a minor subset of the 
Vietnamese character range. 
Most of our accented characters lie outside Latin-1, but some (single grave, 
circumflex or acute accent 
on vowel a, e, i, o or u) are inside it. The conversion reported by timar has 
only worked on a vowel which 
is part of the Latin-1 set, and the ones which did not convert properly are 
outside the Latin-1 set, so I 
would speculate that the conversion supported Latin-1, but nothing else.

Pavel, your Windows builds for SRC680_m198 aren't up yet, only Linux, but when 
the Windows builds do 
come up, do you still want me to test the build for this bug? I am happy to do 
so. It's excellent PR for 
OpenOffice.org if we can fix this long-standing problem with localized Windows 
installers. :))

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-25 Thread timar
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User timar changed the following:

What|Old value |New value

  CC|''|'timar'





--- Additional comments from [EMAIL PROTECTED] Mon Dec 25 10:28:20 -0800 
2006 ---
I have not had the chance to try Pavel's build but in my build environment I
found that this patch was not enough to resolve the issue.

The ulfconv tool which converts UTF8 encoded .ulf files to ANSI encoded .mlf
files could not convert from UTF8 to CP1258. For example the line

vi = "Chuẩn bị cài đặt %PRODUCTNAME %PRODUCTVERSION" (in UTF8)

became

vi = "Chu?n b? cài đ?t %PRODUCTNAME %PRODUCTVERSION" (in CP1258)

Some characters were converted correctly but some became ? marks. When I used
iconv command (iconv -f utf-8 -t cp1258) all characters were converted
correctly. Therefore I suspect there was a problem with ulfconv
(setup_native\source\ulfconv\ulfconv.cxx) or rather in the underlying libraries
which ulfconv used.



-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-23 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905





--- Additional comments from [EMAIL PROTECTED] Sat Dec 23 09:54:25 -0800 
2006 ---
... I manually patche current m198 build, so this patch will be in my Windows 
build of SRC680_m198.

Please verify that this patch fixes your problem.


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-23 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

  Status|NEW   |RESOLVED

  Resolution|  |FIXED





--- Additional comments from [EMAIL PROTECTED] Sat Dec 23 09:45:44 -0800 
2006 ---
applied to pj69.


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-23 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

 Assigned to|ihi   |pjanik

  Status|STARTED   |NEW





--- Additional comments from [EMAIL PROTECTED] Sat Dec 23 09:16:09 -0800 
2006 ---
.

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-23 Thread pjanik
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905


User pjanik changed the following:

What|Old value |New value

  Ever confirmed|  |1

  Status|UNCONFIRMED   |STARTED

Target milestone|---   |OOo 2.2





--- Additional comments from [EMAIL PROTECTED] Sat Dec 23 09:15:36 -0800 
2006 ---
Ivo: I'll take it because I do have another issue in this file (#i72425#).

Clytie: please test it in my build that will be uploaded *after* my SRC680_m198 
which doesn't contain it 
yet.


-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[l10n-issues] [Issue 72905] Patch for Unicode bug in N SIS

2006-12-23 Thread clytie
To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=72905
 Issue #|72905
 Summary|Patch for Unicode bug in NSIS
   Component|l10n
 Version|OOo 2.1
Platform|All
 URL|
  OS/Version|Windows, all
  Status|UNCONFIRMED
   Status whiteboard|
Keywords|
  Resolution|
  Issue type|PATCH
Priority|P3
Subcomponent|code
 Assigned to|ihi
 Reported by|clytie





--- Additional comments from [EMAIL PROTECTED] Sat Dec 23 03:19:32 -0800 
2006 ---
Contributed by Andras Timar:
___
--- setup_native/source/win32/msi-encodinglist.orig 2006-12-01
18:56:55.0 +0100
+++ setup_native/source/win32/msi-encodinglist.txt  2006-12-20
18:45:46.0 +0100
@@ -79,7 +79,7 @@
 ts   0  1073   # Tsonga
 uk1251  1058   # Ukrainian
 ur1256  1056   # Urdu
-vi   0  1066   # Vietnamese
+vi1258  1066   # Vietnamese
 xh   0  1076   # Xhosa
 ns   0  1132   # Northern Sotho (Sepedi)
 yi   0  1085   # Yiddish
___

Andras says this patch will convert the UTF8 strings to single-byte encodings 
for the different 
languages. This should fix the munged characters in install dialogues. See 
issue 48741.

-
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]