Re: Try to fix remaining unicode CJK failures?

2019-01-10 Thread Kornel Benko
Am Donnerstag, 10. Januar 2019 22:39:11 CET schrieb Scott Kostyshak 
:
> On Thu, Jan 10, 2019 at 10:41:44PM +, Guenter Milde wrote:
> 
> > After LyX was rebuilt from git commit hash 48483017, all 20 utf8-cjk
> > tests fail!
> > 
> > The reason is:
> > 
> >   ! Undefined control sequence.
> >\inputencoding 
> > {latin9}~
> >   l.64 \item [{\inputencoding{latin9}~}]
> > 
> > which is also very strange, because the encoding should be utf8!
> > 
> > Günter  
> 
> All 20 pass for me with a fresh build of 48483017.

Here too.

> Scott
> 

Kornel


signature.asc
Description: This is a digitally signed message part.


Re: Try to fix remaining unicode CJK failures?

2019-01-10 Thread Scott Kostyshak
On Thu, Jan 10, 2019 at 10:41:44PM +, Guenter Milde wrote:

> After LyX was rebuilt from git commit hash 48483017, all 20 utf8-cjk
> tests fail!
> 
> The reason is:
> 
>   ! Undefined control sequence.
>\inputencoding 
> {latin9}~
>   l.64 \item [{\inputencoding{latin9}~}]  
> 
> which is also very strange, because the encoding should be utf8!
> 
> Günter

All 20 pass for me with a fresh build of 48483017.

Scott


signature.asc
Description: PGP signature


Selecting exponents

2019-01-10 Thread Andrew Parsloe
Recently I wished to select e^{2\pi it} from within a larger formula (in 
LyX 2.3.2). Using the mouse I ended with the selection shown in the 
attached .png. The highlighting suggests the exponent has been selected, 
but on pasting (also shown in the .png)  only the "e" was pasted. This 
caused a moment's confusion, including checking that the same thing 
happened in LyX 2.2.3.  If I had moved the mouse pointer another pixel 
to the right then the selection would have embraced the whole formula, 
as desired, but such minute adjustments are difficult with the mouse. In 
the maths editor, I can see a virtue in having two shades (or colours) 
of selection highlighting to indicate the different stages of selection.


Andrew



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


Re: Try to fix remaining unicode CJK failures?

2019-01-10 Thread Guenter Milde
On 2019-01-10, Scott Kostyshak wrote:

> [-- Type: text/plain, Encoding: quoted-printable --]

> On Thu, Jan 10, 2019 at 08:47:50PM +0100, Kornel Benko wrote:
>> Am Donnerstag, 10. Januar 2019 17:50:42 CET schrieb Guenter Milde 
>> :
>> > On 2019-01-09, Guenter Milde wrote:
>> > 
>> > > I assume, that we need to change *all* occurences of force=utf8 because
>> > > utf8-cjk add just CJK characters that are not handled in
>> > > lib/unicodesymbols.
>> > 
>> > Jürgen fixed most cases in lib/unicodesymbols. Thanks!
>> > 
>> > In TeXLive 17, now all non-forced characters (except ASCII) come out as
>> > two wrong characters in the PDF because the exported latex files misses the
>> > inputenc declaration line
>> > 
>> >   \usepackage[utf8]{inputenc}
>> > 
>> > (in TeXLive18 the new default kicks in). This also lead to some compilation
>> > errors due to missing chars.
>> > 
>> > In my "manual tests", adding
>> > 
>> >   \usepackage[utf8]{inputenc}
>> > 
>> > to the LaTeX preamble helped export with pdflatex in some cases and
>> > exposed some more missing force=utf8-cjk in lib/unicodesymbols.
>> > 
>> > 
>> > Furthermore, there are some cases of "force!=utf8" which should become
>> > "force!=utf8;utf8-cjk". However, this change led to forced conversion in
>> > all encodings! There seems to be a bug in the parsing of 
>> > lib/unicodesymbols.
>> > 
>> > 
>> > Günter
>> > 
>> > 
>> > 
>> > 

>> Checking
>>  $ ctest -R 'unicodesymbols.*cjk'
>> ...
>> 89% tests passed, 2 tests failed out of 19
>> ...
>> The following tests FAILED:
>>  813 - 
>> INVERTED.DEFAULTOUTPUT.TODO_export/export/latex/unicodesymbols/070-72-general-punctuation-sub-super-currency_utf8-cjk_pdf2
>>  (Failed)
>>  861 - 
>> INVERTED.DEFAULTOUTPUT.TODO_export/export/latex/unicodesymbols/074-76-letterlike-numberforms-arrows_utf8-cjk_pdf2
>>  (Failed)
>> Errors while running CTest

>> That means, these two may disappear from invertedTests too.

> +1 same results here. I think we just need to remove the only
> cjk-specific regex line in invertedTests. Your turn to commit :)

After LyX was rebuilt from git commit hash 48483017, all 20 utf8-cjk
tests fail!

The reason is:

  ! Undefined control sequence.
   \inputencoding 
{latin9}~
  l.64 \item [{\inputencoding{latin9}~}]

which is also very strange, because the encoding should be utf8!

Günter  



Re: Try to fix remaining unicode CJK failures?

2019-01-10 Thread Scott Kostyshak
On Thu, Jan 10, 2019 at 08:47:50PM +0100, Kornel Benko wrote:
> Am Donnerstag, 10. Januar 2019 17:50:42 CET schrieb Guenter Milde 
> :
> > On 2019-01-09, Guenter Milde wrote:
> > 
> > > I assume, that we need to change *all* occurences of force=utf8 because
> > > utf8-cjk add just CJK characters that are not handled in
> > > lib/unicodesymbols.
> > 
> > Jürgen fixed most cases in lib/unicodesymbols. Thanks!
> > 
> > In TeXLive 17, now all non-forced characters (except ASCII) come out as
> > two wrong characters in the PDF because the exported latex files misses the
> > inputenc declaration line
> > 
> >   \usepackage[utf8]{inputenc}
> > 
> > (in TeXLive18 the new default kicks in). This also lead to some compilation
> > errors due to missing chars.
> > 
> > In my "manual tests", adding
> > 
> >   \usepackage[utf8]{inputenc}
> > 
> > to the LaTeX preamble helped export with pdflatex in some cases and
> > exposed some more missing force=utf8-cjk in lib/unicodesymbols.
> > 
> > 
> > Furthermore, there are some cases of "force!=utf8" which should become
> > "force!=utf8;utf8-cjk". However, this change led to forced conversion in
> > all encodings! There seems to be a bug in the parsing of lib/unicodesymbols.
> > 
> > 
> > Günter
> > 
> > 
> > 
> > 
> 
> Checking
>   $ ctest -R 'unicodesymbols.*cjk'
> ...
> 89% tests passed, 2 tests failed out of 19
> ...
> The following tests FAILED:
>   813 - 
> INVERTED.DEFAULTOUTPUT.TODO_export/export/latex/unicodesymbols/070-72-general-punctuation-sub-super-currency_utf8-cjk_pdf2
>  (Failed)
>   861 - 
> INVERTED.DEFAULTOUTPUT.TODO_export/export/latex/unicodesymbols/074-76-letterlike-numberforms-arrows_utf8-cjk_pdf2
>  (Failed)
> Errors while running CTest
> 
> That means, these two may disappear from invertedTests too.

+1 same results here. I think we just need to remove the only
cjk-specific regex line in invertedTests. Your turn to commit :)

Scott


signature.asc
Description: PGP signature


Re: Try to fix remaining unicode CJK failures?

2019-01-10 Thread Kornel Benko
Am Donnerstag, 10. Januar 2019 17:50:42 CET schrieb Guenter Milde 
:
> On 2019-01-09, Guenter Milde wrote:
> 
> > I assume, that we need to change *all* occurences of force=utf8 because
> > utf8-cjk add just CJK characters that are not handled in
> > lib/unicodesymbols.
> 
> Jürgen fixed most cases in lib/unicodesymbols. Thanks!
> 
> In TeXLive 17, now all non-forced characters (except ASCII) come out as
> two wrong characters in the PDF because the exported latex files misses the
> inputenc declaration line
> 
>   \usepackage[utf8]{inputenc}
> 
> (in TeXLive18 the new default kicks in). This also lead to some compilation
> errors due to missing chars.
> 
> In my "manual tests", adding
> 
>   \usepackage[utf8]{inputenc}
> 
> to the LaTeX preamble helped export with pdflatex in some cases and
> exposed some more missing force=utf8-cjk in lib/unicodesymbols.
> 
> 
> Furthermore, there are some cases of "force!=utf8" which should become
> "force!=utf8;utf8-cjk". However, this change led to forced conversion in
> all encodings! There seems to be a bug in the parsing of lib/unicodesymbols.
> 
> 
> Günter
> 
> 
> 
> 

Checking
$ ctest -R 'unicodesymbols.*cjk'
...
89% tests passed, 2 tests failed out of 19
...
The following tests FAILED:
813 - 
INVERTED.DEFAULTOUTPUT.TODO_export/export/latex/unicodesymbols/070-72-general-punctuation-sub-super-currency_utf8-cjk_pdf2
 (Failed)
861 - 
INVERTED.DEFAULTOUTPUT.TODO_export/export/latex/unicodesymbols/074-76-letterlike-numberforms-arrows_utf8-cjk_pdf2
 (Failed)
Errors while running CTest

That means, these two may disappear from invertedTests too.

Kornel

signature.asc
Description: This is a digitally signed message part.


Re: [LyX/master] CJKutf8 is only to be used with the dedicated utf8-cjk encoding

2019-01-10 Thread Scott Kostyshak
On Thu, Jan 10, 2019 at 10:38:26AM +0100, Kornel Benko wrote:
> Am Donnerstag, 10. Januar 2019 09:58:13 CET schrieb Jürgen Spitzmüller 
> :
> > Am Do., 10. Jan. 2019 um 01:33 Uhr schrieb Scott Kostyshak  > >:
> > 
> > > There are some tests newly failing on master. I bisected based on one of
> > > those tests and ended up here. Here is the one I bisected on:
> > >
> > >   005-7-ipa-modifiers-combining_utf8-cjk_pdf2
> > >
> > > If I open the document
> > >
> > >   autotests/export/latex/unicodesymbols/005-7-ipa-modifiers-combining.lyx
> > >
> > > and change
> > >
> > >   Document > Settings > Language > Encoding
> > >
> > > to "Unicode (CJK) (utf8)", when I export I get the following error:
> > >
> > >   ! Package inputenc Error: Unicode character ʰ (U+02B0)
> > >
> > 
> > Should be fixed with 4786c59a8c and 81af82d1e8. Two errors are still
> > remaining in 084-misc-symbols_utf8-cjk, but these occur also with normal
> > utf8 encoding.
> > 
> > Jürgen
> > 
> 
> All test mentioned by Scott pass here now. Also 084-misc-symbols_utf8-cjk.

With Jürgen's fix, all unicodesymbols tests pass for me now. Thanks,
Jürgen.

Scott


signature.asc
Description: PGP signature


Re: Zero size file in repo

2019-01-10 Thread Richard Kimberly Heck
On 1/10/19 12:38 PM, Kornel Benko wrote:
> Size of src/frontends/qt4/GuiGraphicsUi.h is 0, is that intended?

Looks like an accidental commit.

> git log GuiGraphicsUi.h
commit a1cec91afaca91968b46e695533c10ad2a3f73d3
Author: André Pönitz 
Date:   Fri Aug 31 05:53:55 2007 +

    move our stuff off the Q* namespace
   
    git-svn-id: svn://svn.lyx.org/lyx/lyx-devel/trunk@19935
a592a061-630c-0410-9148-cb99ea01b6c8

That's all a rename, basically. The original file, QGraphicsUi.h, was
commited at 12e5a52b92, and it was empty then, too.

I've deleted it (and verified that compilation is fine).

Riki




Re: Try to fix remaining unicode CJK failures?

2019-01-10 Thread Guenter Milde
On 2019-01-09, Guenter Milde wrote:

> I assume, that we need to change *all* occurences of force=utf8 because
> utf8-cjk add just CJK characters that are not handled in
> lib/unicodesymbols.

Jürgen fixed most cases in lib/unicodesymbols. Thanks!

In TeXLive 17, now all non-forced characters (except ASCII) come out as
two wrong characters in the PDF because the exported latex files misses the
inputenc declaration line

  \usepackage[utf8]{inputenc}

(in TeXLive18 the new default kicks in). This also lead to some compilation
errors due to missing chars.

In my "manual tests", adding

  \usepackage[utf8]{inputenc}

to the LaTeX preamble helped export with pdflatex in some cases and
exposed some more missing force=utf8-cjk in lib/unicodesymbols.


Furthermore, there are some cases of "force!=utf8" which should become
"force!=utf8;utf8-cjk". However, this change led to forced conversion in
all encodings! There seems to be a bug in the parsing of lib/unicodesymbols.


Günter





Zero size file in repo

2019-01-10 Thread Kornel Benko
Size of src/frontends/qt4/GuiGraphicsUi.h is 0, is that intended?

Kornel

signature.asc
Description: This is a digitally signed message part.


Re: [LyX/master] CJKutf8 is only to be used with the dedicated utf8-cjk encoding

2019-01-10 Thread Jürgen Spitzmüller
Am Do., 10. Jan. 2019 um 10:38 Uhr schrieb Kornel Benko :

> All test mentioned by Scott pass here now. Also 084-misc-symbols_utf8-cjk.
>

Also candidate for stable, then.

Jürgen


>
> Kornel
>


Re: [LyX/master] CJKutf8 is only to be used with the dedicated utf8-cjk encoding

2019-01-10 Thread Kornel Benko
Am Donnerstag, 10. Januar 2019 09:58:13 CET schrieb Jürgen Spitzmüller 
:
> Am Do., 10. Jan. 2019 um 01:33 Uhr schrieb Scott Kostyshak  >:
> 
> > There are some tests newly failing on master. I bisected based on one of
> > those tests and ended up here. Here is the one I bisected on:
> >
> >   005-7-ipa-modifiers-combining_utf8-cjk_pdf2
> >
> > If I open the document
> >
> >   autotests/export/latex/unicodesymbols/005-7-ipa-modifiers-combining.lyx
> >
> > and change
> >
> >   Document > Settings > Language > Encoding
> >
> > to "Unicode (CJK) (utf8)", when I export I get the following error:
> >
> >   ! Package inputenc Error: Unicode character ʰ (U+02B0)
> >
> 
> Should be fixed with 4786c59a8c and 81af82d1e8. Two errors are still
> remaining in 084-misc-symbols_utf8-cjk, but these occur also with normal
> utf8 encoding.
> 
> Jürgen
> 

All test mentioned by Scott pass here now. Also 084-misc-symbols_utf8-cjk.

Kornel


signature.asc
Description: This is a digitally signed message part.


Re: [LyX/master] CJKutf8 is only to be used with the dedicated utf8-cjk encoding

2019-01-10 Thread Jürgen Spitzmüller
Am Do., 10. Jan. 2019 um 01:33 Uhr schrieb Scott Kostyshak :

> There are some tests newly failing on master. I bisected based on one of
> those tests and ended up here. Here is the one I bisected on:
>
>   005-7-ipa-modifiers-combining_utf8-cjk_pdf2
>
> If I open the document
>
>   autotests/export/latex/unicodesymbols/005-7-ipa-modifiers-combining.lyx
>
> and change
>
>   Document > Settings > Language > Encoding
>
> to "Unicode (CJK) (utf8)", when I export I get the following error:
>
>   ! Package inputenc Error: Unicode character ʰ (U+02B0)
>

Should be fixed with 4786c59a8c and 81af82d1e8. Two errors are still
remaining in 084-misc-symbols_utf8-cjk, but these occur also with normal
utf8 encoding.

Jürgen