23.0.60; Indic Script is not rendered properly

23.0.60; Indic Script is not rendered properly

[Ashish Shukla]

In typing the text like "शुक्ल", I get this bug.

Following is the URL of screenshot of what I'm getting currently:
http://flickr.com/photos/wahjava/2525377097/

An earlier version of Emacs CVS used to render that correctly, as
shown in the following URL of screenshot:
http://flickr.com/photos/wahjava/2488290981/

I'm currently running emacs-cvs (checked-out an hour ago) port, in
Gentoo GNU/Linux :) .

Emacs is not crashed.

In GNU Emacs 23.0.60.1 (x86_64-pc-linux-gnu, GTK+ Version 2.12.8)
 of 2008-05-27 on chateau
Windowing system distributor `The X.Org Foundation', version 11.0.10300000
configured using `configure  '--prefix=/usr' '--host=x86_64-pc-linux-gnu' '--mandir=/usr/share/man' '--infodir=/usr/share/info' '--datadir=/usr/share' '--sysconfdir=/etc' '--localstatedir=/var/lib' '--program-suffix=-emacs-23' '--infodir=/usr/share/info/emacs-23' '--without-carbon' '--with-sound' '--with-x' '--with-toolkit-scroll-bars' '--with-gif' '--with-jpeg' '--with-png' '--with-rsvg' '--with-tiff' '--with-xpm' '--enable-font-backend' '--with-freetype' '--with-xft' '--with-libotf' '--with-m17n-flt' '--with-x-toolkit=gtk' '--without-hesiod' '--without-kerberos' '--without-kerberos5' '--with-gpm' '--with-dbus' '--libdir=/usr/lib64' '--build=x86_64-pc-linux-gnu' 'build_alias=x86_64-pc-linux-gnu' 'host_alias=x86_64-pc-linux-gnu' 'CFLAGS=-O2 -pipe -mtune=nocona -march=nocona -ggdb' 'LDFLAGS
 =-Wl,--hash-style,both''

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: en_US.UTF-8
  value of $XMODIFIERS: nil
  locale-coding-system: nil
  default-enable-multibyte-characters: t

Major mode: Lisp Interaction

Minor modes in effect:
  jabber-activity-mode: t
  erc-track-mode: t
  erc-spelling-mode: t
  erc-services-mode: t
  erc-networks-mode: t
  erc-ring-mode: t
  erc-netsplit-mode: t
  erc-menu-mode: t
  erc-match-mode: t
  erc-pcomplete-mode: t
  erc-button-mode: t
  erc-fill-mode: t
  erc-stamp-mode: t
  erc-autojoin-mode: t
  erc-smiley-mode: t
  erc-irccontrols-mode: t
  erc-noncommands-mode: t
  erc-readonly-mode: t
  display-time-mode: t
  server-mode: t
  show-paren-mode: t
  xterm-mouse-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  global-auto-composition-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  column-number-mode: t
  line-number-mode: t
  transient-mark-mode: t

Recent input:
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <menu-bar> <help-menu> <send-emacs-bug
-report>

Recent messages:
line-number-mode mode already enabled.
column-number-mode enabled.
display-time-mode enabled.
Toggling blink-cursor-mode off; better pass an explicit argument.
Toggling tool-bar-mode off; better pass an explicit argument.
tool-bar-mode disabled.
Image conversion failed (code `1') [2 times]
Source file `/usr/share/emacs/site-lisp/cedet/semantic/bovine/bovine-debug.el' newer than byte-compiled file
Source file `/usr/share/emacs/site-lisp/cedet/semantic/semantic-debug.el' newer than byte-compiled file
For information about GNU Emacs and the GNU system, type C-h C-a.

Re: 23.0.60; Indic Script is not rendered properly

[Kenichi Handa]

In article <20080526232520.349FC1332D2@chateau.d.lf>, Ashish Shukla <wahjava@gmail.com> writes:

> In typing the text like "शुक्ल", I get this bug.
> Following is the URL of screenshot of what I'm getting currently:
> http://flickr.com/photos/wahjava/2525377097/

> An earlier version of Emacs CVS used to render that correctly, as
> shown in the following URL of screenshot:
> http://flickr.com/photos/wahjava/2488290981/

Please use C-u C-x = on a devanagari character to check
which font is used for it.

And also, if you know how to use gdb, please set break point
on Ffont_shape_text, and check why shaping doesn't work.

---
Kenichi Handa
handa@ni.aist.go.jp

Re: 23.0.60; Indic Script is not rendered properly

[Manoj Srivastava]

[-- Attachment #1: Type: text/plain, Size: 911 bytes --]

On Tue, 27 May 2008 13:19:31 +0900, Kenichi Handa <handa@m17n.org> said: 

> In article <20080526232520.349FC1332D2@chateau.d.lf>, Ashish Shukla
> <wahjava@gmail.com> writes:
>> In typing the text like "शुक्ल", I get this bug.  Following is the
>> URL of screenshot of what I'm getting currently:
>> http://flickr.com/photos/wahjava/2525377097/

>> An earlier version of Emacs CVS used to render that correctly, as
>> shown in the following URL of screenshot:
>> http://flickr.com/photos/wahjava/2488290981/

> Please use C-u C-x = on a devanagari character to check which font is
> used for it.

> And also, if you know how to use gdb, please set break point on
> Ffont_shape_text, and check why shaping doesn't work.

        OK. Here is my take on it.
 “मनोज श्रिवास्तव” is how to write my name; this is how an older version
 of emacs used to render it (correctly)


[-- Attachment #2: Manoj Srivastava, rendered correctly --]
[-- Type: image/png, Size: 26295 bytes --]

[-- Attachment #3: Type: text/plain, Size: 42 bytes --]


        This is what it looks like now:


[-- Attachment #4: incorrect rendering --]
[-- Type: image/png, Size: 17394 bytes --]

[-- Attachment #5: Type: text/plain, Size: 934 bytes --]


        C-u C-x = on the first character of the second word says:
--8<---------------cut here---------------start------------->8---
        character: श (2358, #o4466, #x936)
preferred charset: unicode (Unicode (ISO10646))
       code point: 0x0936
           syntax: w 	which means: word
         category: i:Indian
      buffer code: #xE0 #xA4 #xB6
        file code: #xE0 #xA4 #xB6 (encoded by coding system utf-8-emacs)
          display: by this font (glyph code)
     -unknown-FreeSans-normal-normal-normal-*-11-*-*-*-*-0-iso10646-1 (#x4DB)
--8<---------------cut here---------------end--------------->8---

        I ran gdb, but I did not know what I was looking for, and I got
 lost.

        manoj
-- 
To find a friend one must close one eye; to keep him -- two. Norman
Douglas
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Re: 23.0.60; Indic Script is not rendered properly

[Ashish Shukla आशीष शुक्ल]

[-- Attachment #1: Type: text/plain, Size: 3965 bytes --]

>>>>> Kenichi Handa writes:
    Kenichi> In article <20080526232520.349FC1332D2@chateau.d.lf>, Ashish Shukla <wahjava@gmail.com> writes:

    >> In typing the text like "शुक्ल", I get this bug.
    >> Following is the URL of screenshot of what I'm getting currently:
    >> http://flickr.com/photos/wahjava/2525377097/

    >> An earlier version of Emacs CVS used to render that correctly, as
    >> shown in the following URL of screenshot:
    >> http://flickr.com/photos/wahjava/2488290981/

    Kenichi> Please use C-u C-x = on a devanagari character to check
    Kenichi> which font is used for it.

I typed characters क्ल and placed cursor on them, and then did C-u C-x =,
and following is the output.

---->8---->8----
        character: क (2325, #o4425, #x915)
preferred charset: unicode (Unicode (ISO10646))
       code point: 0x0915
           syntax: w 	which means: word
         category: i:Indian
         to input: type "ka" with devanagari-itrans
      buffer code: #xE0 #xA4 #x95
        file code: #xE0 #xA4 #x95
		   (encoded by coding system utf-8-unix)
          display:
		   composed to form "क्ल" (see below)

Composed with the following character(s) "्ल" using this font:
  -unknown-FreeSans-normal-normal-normal-*-12-*-*-*-*-0-iso10646-1
by these glyphs:
  [#<font-object "-unknown-FreeSans-normal-normal-normal-*-12-*-*-*-*-0-iso10646-1"> 17 -2 9 8 4]
  [0 2 2325 1210 9 9 0 8 1 nil]
  [0 2 2381 1265 0 -8 -11 0 3 [-3 -1 0]]
  [0 2 2354 1239 8 8 0 8 1 nil]

Character code properties are not shown: customize what to show

There are text properties here:
  auto-composed        t
  composition          [Show]
  fontified            t
---->8---->8----

Now I got that Emacs is using FreeSans font to render Devanagari
characters which is not expected, so I modified fontset-default to use
"Lohit Hindi-10" Xft font using following Elisp code:

---->8---->8----
(set-fontset-font "fontset-default"
(cons (decode-char 'ucs ?\x0900)
(decode-char 'ucs ?\x097f))
"Lohit Hindi-10")
---->8---->8----

And typed characters शुक्ल, and this time text got rendered perfectly and
following is the output of C-u C-x = on the text.

---->8---->8----
        character: क (2325, #o4425, #x915)
preferred charset: unicode (Unicode (ISO10646))
       code point: 0x0915
           syntax: w 	which means: word
         category: i:Indian
      buffer code: #xE0 #xA4 #x95
        file code: #xE0 #xA4 #x95 (encoded by coding system utf-8-emacs)
          display: composed to form "क्ल" (see below)

Composed with the following character(s) "्ल" using this font:
  -unknown-Lohit Hindi-normal-normal-normal-*-12-*-*-*-*-0-*-*
by these glyphs:
  [#<font-object "-unknown-Lohit Hindi-normal-normal-normal-*-12-*-*-*-*-0-*-*"> 16 8 7 8 0]
  [0 2 0 179 7 8 0 8 0 nil]
  [0 2 2354 65 9 9 0 8 0 nil]

Character code properties are not shown: customize what to show

There are text properties here:
  auto-composed        t
  composition          [Show]
  fontified            t
---->8---->8----

So I guess this is the font issue, that Emacs is not able to figure out
which font to use. For the moment, I added that LISP code in my .emacs,
but is this the proper fix for this issue ? Or, I've to change something
in fonts.conf or any such font-related configuration, hmmm..?

    Kenichi> And also, if you know how to use gdb, please set break point
    Kenichi> on Ffont_shape_text, and check why shaping doesn't work.

Okay, I know how to use gdb, but how do I know if shaping is working or
not, hmm...?

    Kenichi> ---
    Kenichi> Kenichi Handa
    Kenichi> handa@ni.aist.go.jp

Thanks for the help :) .
-- 
Ashish Shukla आशीष शुक्ल                      http://wahjava.wordpress.com/
·-- ·- ···· ·--- ·- ···- ·- ·--·-· --· -- ·- ·· ·-·· ·-·-·- -·-· --- --

[-- Attachment #2: Type: application/pgp-signature, Size: 196 bytes --]

Re: 23.0.60; Indic Script is not rendered properly

[Kenichi Handa]

In article <87lk1wuqke.fsf@chateau.d.lf>, wahjava@gmail.com (Ashish Shukla आशीष शुक्ल) writes:

> I typed characters क्ल and placed cursor on them, and then did C-u C-x =,
> and following is the output.

> ---->8---->8----
>         character: क (2325, #o4425, #x915)
> preferred charset: unicode (Unicode (ISO10646))
>        code point: 0x0915
>            syntax: w 	which means: word
>          category: i:Indian
>          to input: type "ka" with devanagari-itrans
>       buffer code: #xE0 #xA4 #x95
>         file code: #xE0 #xA4 #x95
> 		   (encoded by coding system utf-8-unix)
>           display:
> 		   composed to form "क्ल" (see below)

> Composed with the following character(s) "्ल" using this font:
>   -unknown-FreeSans-normal-normal-normal-*-12-*-*-*-*-0-iso10646-1
> by these glyphs:
>   [#<font-object "-unknown-FreeSans-normal-normal-normal-*-12-*-*-*-*-0-iso10646-1"> 17 -2 9 8 4]
>   [0 2 2325 1210 9 9 0 8 1 nil]
>   [0 2 2381 1265 0 -8 -11 0 3 [-3 -1 0]]
>   [0 2 2354 1239 8 8 0 8 1 nil]

Hmmm, it seems that your FreeSans.ttf is old.  At least, my
FreeSans.ttf (Version 1.79 of date 16-5-2007) displays that
text correctly.

[...]
> So I guess this is the font issue, that Emacs is not able to figure out
> which font to use. For the moment, I added that LISP code in my .emacs,
> but is this the proper fix for this issue ? Or, I've to change something
> in fonts.conf or any such font-related configuration, hmmm..?

For Devanagari, by default, Emacs uses an OpenType font that
supports the script `deva' and has feature `rphf'.  If
multiple fonts match that spec, Emacs selects one
arbitrarily.  Perhaps, we must implement a way to reflect
the preference of underlying font-backend.

---
Kenichi Handa
handa@ni.aist.go.jp

Re: 23.0.60; Indic Script is not rendered properly

[आशीष शुक्ल Ashish Shukla]

[-- Attachment #1: Type: text/plain, Size: 816 bytes --]

,--[ On Tue, May 27, 2008 at 08:48:40PM +0900, Kenichi Handa wrote:

[snip]

| Hmmm, it seems that your FreeSans.ttf is old.  At least, my
| FreeSans.ttf (Version 1.79 of date 16-5-2007) displays that
| text correctly.

Are you using FreeSans.ttf from repositories, hmm..? The version of
GNU freefont[1], I'm using is 20080323, released on 24-Mar-2008. BtW, 
I'm using freefont-ttf. And they've made no releases in 2007.

And on hexdump -C FreeSans.ttf, I got following version, I hope thats
the correct version:

Version: $Revision: 1.136$

So, are there any issues with new version of freefont ?
-- 
Ashish Shukla आशीष शुक्ल                      http://wahjava.wordpress.com/
·-- ·- ···· ·--- ·- ···- ·- ·--·-· --· -- ·- ·· ·-·· ·-·-·- -·-· --- --

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: 23.0.60; Indic Script is not rendered properly

[Manoj Srivastava]

On Tue, 27 May 2008 21:45:23 +0530, आशीष शुक्ल Ashish Shukla <wahjava@gmail.com> said: 

> ,--[ On Tue, May 27, 2008 at 08:48:40PM +0900, Kenichi Handa wrote:
> [snip]

>> Hmmm, it seems that your FreeSans.ttf is old.  At least, my
>> FreeSans.ttf (Version 1.79 of date 16-5-2007) displays that text
>> correctly.

> Are you using FreeSans.ttf from repositories, hmm..? The version of
> GNU freefont[1], I'm using is 20080323, released on 24-Mar-2008. BtW,
> I'm using freefont-ttf. And they've made no releases in 2007.

> And on hexdump -C FreeSans.ttf, I got following version, I hope thats
> the correct version:
> Version: $Revision: 1.136$

> So, are there any issues with new version of freefont ?

        To chime in, I have the same version.  It comes from:
--8<---------------cut here---------------start------------->8---
__> dpkg -l ttf-freefont
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
ii  ttf-freefont   20080323-3     Freefont Serif, Sans and Mono Truetype fonts
--8<---------------cut here---------------end--------------->8---

        manoj
-- 
Schizophrenia beats being alone.
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Re: 23.0.60; Indic Script is not rendered properly

[Kenichi Handa]

In article <20080527161523.GA7942@chateau.d.lf>, आशीष शुक्ल Ashish Shukla <wahjava@gmail.com> writes:

> | Hmmm, it seems that your FreeSans.ttf is old.  At least, my
> | FreeSans.ttf (Version 1.79 of date 16-5-2007) displays that
> | text correctly.

> Are you using FreeSans.ttf from repositories, hmm..?

I don't remember from where I got it.

> The version of GNU freefont[1], I'm using is 20080323,
> released on 24-Mar-2008. BtW, I'm using freefont-ttf. And
> they've made no releases in 2007.

> And on hexdump -C FreeSans.ttf, I got following version, I hope thats
> the correct version:

> Version: $Revision: 1.136$

Ummm, then perhaps, some version after 1.79 instroduced a
bug, or my version is the one hacked by some other person.

> So, are there any issues with new version of freefont ?

I don't know.

---
Kenichi Handa
handa@ni.aist.go.jp

Re: 23.0.60; Indic Script is not rendered properly

[Manoj Srivastava]

On Tue, 27 May 2008 13:27:53 +0530, Ashish Shukla आशीष शुक्ल
<wahjava@gmail.com> said:  

----> 8---->8----
> (set-fontset-font "fontset-default" (cons (decode-char 'ucs ?\x0900)
> (decode-char 'ucs ?\x097f)) "Lohit Hindi-10")
----> 8---->8----

        This still does not work for me. When I type in "श्रिवास्तव",
 the composition is not happening (the 'sri' should compose into a
 single character, and the 'st' should compose together as well -- what
 I get is the 'स्' character next to 'त', and they should be joined).

        I think back when using the non-free CDAC fonts I could get it
 to work, but I no longer have those fonts on my machine.


--8<---------------cut here---------------start------------->8---
        character: श (2358, #o4466, #x936)
preferred charset: unicode (Unicode (ISO10646))
       code point: 0x0936
           syntax: w 	which means: word
         category: i:Indian
      buffer code: #xE0 #xA4 #xB6
        file code: #xE0 #xA4 #xB6 (encoded by coding system utf-8-unix)
          display: by this font (glyph code)
     -unknown-Lohit Hindi-normal-normal-normal-*-11-*-*-*-*-0-*-* (#x45)
--8<---------------cut here---------------end--------------->8---

        manoj
-- 
The worst is not so long as we can say "This is the worst." King Lear
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Re: 23.0.60; Indic Script is not rendered properly

[Kenichi Handa]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=UTF-8, Size: 841 bytes --]

In article <87zlqbtx4g.fsf@anzu.internal.golden-gryphon.com>, Manoj Srivastava <srivasta@ieee.org> writes:

> On Tue, 27 May 2008 13:27:53 +0530, Ashish Shukla आशीष शुक्à¤> <wahjava@gmail.com> said:  

> ----> 8---->8----
> > (set-fontset-font "fontset-default" (cons (decode-char 'ucs ?\x0900)
> > (decode-char 'ucs ?\x097f)) "Lohit Hindi-10")
> ----> 8---->8----

>         This still does not work for me. When I type in "श्रिवास्तव",
>  the composition is not happening (the 'sri' should compose into a
>  single character, and the 'st' should compose together as well -- what
>  I get is the 'स्' character next to 'त', and they should be joined).

Strange.  It works for me.  Attached is the image of Emacs
display.  Is it correct?

---
Kenichi Handa
handa@ni.aist.go.jp


[-- Attachment #2: hindi.png --]
[-- Type: image/png, Size: 2996 bytes --]

Re: 23.0.60; Indic Script is not rendered properly

[Manoj Srivastava]

On Wed, 28 May 2008 11:19:15 +0900, Kenichi Handa <handa@m17n.org> said: 

> Strange.  It works for me.  Attached is the image of Emacs display.
> Is it correct?

        Yes, the figure shows a correctly composed version of my
 name. Well, this is good to know -- this means that the display issue
 is a matter of correctly selecting a font, and that Emacs itself is
 able to do the right thing. Now, if only I could replicate your set up
 ...


        manoj

-- 
I have made this letter longer than usual because I lack the time to
make it shorter. -- Blaise Pascal
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Re: 23.0.60; Indic Script is not rendered properly

[Kenichi Handa]

In article <87ej7mn8yu.fsf@anzu.internal.golden-gryphon.com>, Manoj Srivastava <srivasta@ieee.org> writes:

>         Yes, the figure shows a correctly composed version of my
>  name. Well, this is good to know -- this means that the display issue
>  is a matter of correctly selecting a font, and that Emacs itself is
>  able to do the right thing. Now, if only I could replicate your set up
>  ...

Please show me the results of:

% m17n-config --version
% libotf-config --version

Or, if you don't have those commands,

% pkg-config --modversion m17n-flt
% pkg-config --modversion libotf

---
Kenichi Handa
handa@ni.aist.go.jp

Re: 23.0.60; Indic Script is not rendered properly

[Manoj Srivastava]

On Thu, 29 May 2008 11:12:58 +0900, Kenichi Handa <handa@m17n.org> said: 

> In article <87ej7mn8yu.fsf@anzu.internal.golden-gryphon.com>, Manoj
> Srivastava <srivasta@ieee.org> writes:
>> Yes, the figure shows a correctly composed version of my name. Well,
>> this is good to know -- this means that the display issue is a matter
>> of correctly selecting a font, and that Emacs itself is able to do
>> the right thing. Now, if only I could replicate your set up ...

> Please show me the results of:

> % m17n-config --version 
> % libotf-config --version
--8<---------------cut here---------------start------------->8---
__> m17n-config --version
1.5.1
__> libotf-config --version
0.9.7
--8<---------------cut here---------------end--------------->8---

        manoj
-- 
If you want your program to be readable, consider supplying the
argument. --Larry Wall in the perl man page
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Re: 23.0.60; Indic Script is not rendered properly

[Miles Bader]

Manoj Srivastava <srivasta@ieee.org> writes:
>> Please show me the results of:
>
>> % m17n-config --version 
>> % libotf-config --version
> __> m17n-config --version
> 1.5.1
> __> libotf-config --version
> 0.9.7

Incidentally, the version of libm17n in debian is only 1.3.4 (I assume
you've compiled a newer version from source), which seems kind of old...

-Miles

-- 
A zen-buddhist walked into a pizza shop and
said, "Make me one with everything."

Re: 23.0.60; Indic Script is not rendered properly

[Kenichi Handa]

In article <87ve0xlq31.fsf@anzu.internal.golden-gryphon.com>, Manoj Srivastava <srivasta@ieee.org> writes:

> On Thu, 29 May 2008 11:12:58 +0900, Kenichi Handa <handa@m17n.org> said: 
> > In article <87ej7mn8yu.fsf@anzu.internal.golden-gryphon.com>, Manoj
> > Srivastava <srivasta@ieee.org> writes:
>>> Yes, the figure shows a correctly composed version of my name. Well,
>>> this is good to know -- this means that the display issue is a matter
>>> of correctly selecting a font, and that Emacs itself is able to do
>>> the right thing. Now, if only I could replicate your set up ...

> > Please show me the results of:

> > % m17n-config --version 
> > % libotf-config --version
> --8<---------------cut here---------------start------------->8---
__> m17n-config --version
> 1.5.1
__> libotf-config --version
> 0.9.7
> --8<---------------cut here---------------end--------------->8---

They are the latest version.  Ummm, I don't understand why
Emacs doesn't render Devanagari correctly in your
environment.

> > And also, if you know how to use gdb, please set break point on
> > Ffont_shape_text, and check why shaping doesn't work.
[...]
>         I ran gdb, but I did not know what I was looking for, and I got
>  lost.

When Emacs stops at Ffont_shape_text, please use n(ext)
command to step one line by one, and check where the
function return.

---
Kenichi Handa
handa@ni.aist.go.jp

Re: 23.0.60; Indic Script is not rendered properly

[Manoj Srivastava]

On Thu, 29 May 2008 13:53:06 +0900, Kenichi Handa <handa@m17n.org> said: 

> When Emacs stops at Ffont_shape_text, please use n(ext) command to
> step one line by one, and check where the function return.

        It seems to return at the end. Edited transcript follows.

        manoj
======================================================================
(gdb) break Ffont_shape_text
Breakpoint 1 at 0x552dd0: file /usr/local/git/emacs/src/font.c, line 3866.
(gdb) run
[Thread debugging using libthread_db enabled]
[New Thread 0x7fc73c6c4780 (LWP 5419)]
[Switching to Thread 0x7fc73c6c4780 (LWP 5419)]

Breakpoint 1, Ffont_shape_text (from=200, to=240, font_object=75047108,
string=10797201) at /usr/local/git/emacs/src/font.c:3866 



  <---- End of function break ---->
(gdb) break /usr/local/git/emacs/src/font.c:3989
Breakpoint 2 at 0x5533a5: file /usr/local/git/emacs/src/font.c, line 3989.

  <---- Intermediate returns ---->
(gdb) break /usr/local/git/emacs/src/font.c:3905
Breakpoint 3 at 0x552e0f: file /usr/local/git/emacs/src/font.c, line 3905.


3902       gstring = Fvconcat (2, args);
3903     }
3904   if (! INTEGERP (n) || XINT (n) == 0)
3905     return Qnil;
3906   len = XINT (n);
3907 

  <---- Intermediate returns ---->
(gdb) break /usr/local/git/emacs/src/font.c:3870
Breakpoint 4 at 0x552e8c: file /usr/local/git/emacs/src/font.c, line 3870.
(gdb) break /usr/local/git/emacs/src/font.c:3867
Breakpoint 6 at 0x552e7b: file /usr/local/git/emacs/src/font.c, line 3867.


3861   struct font_metrics metrics;
3862   EMACS_INT start, end;
3863   Lisp_Object gstring, n;
3864   int len, i;
3865 
3866   if (! FONT_OBJECT_P (font_object))
3867     return Qnil;
3868   font = XFONT_OBJECT (font_object);
3869   if (! font->driver->shape)
3870     return Qnil;
3871 


(gdb) cont
Breakpoint 2, Ffont_shape_text (from=232, to=240, font_object=75047108,
string=10797201) at /usr/local/git/emacs/src/font.c:3989 
(gdb) 

3985       else
3986         i = j;
3987     }
3988 
3989   return to;
3990 }
3991 

-- 
"Every institution I've ever been associated with has tried to screw
me." Stephen Wolfram
Manoj Srivastava <srivasta@acm.org> <http://www.golden-gryphon.com/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C