bug#2307: Bootstrap hangs in leim

bug#2307: Bootstrap hangs in leim

[Juanma Barranquero]

Package: emacs
X-Debbugs-CC: handa@m17n.org
X-Debbugs-No-Ack: yes

(After circumventing bug#2304) bootstrapping Emacs hangs at this point:

Converting ziranma.cin to ZIRANMA.el...
Loading vc-cvs...
Converting ziranma.cin to ZIRANMA.el...done
Converting pinyin.map to PY.el...
Converting pinyin.map to PY.el...done
Converting CTLau.html to CTLau.el...
Converting CTLau.html to CTLau.el...done
Converting CTLau-b5.html to CTLau-b5.el...
Converting CTLau-b5.html to CTLau-b5.el...done
Converting cangjie-table.cns to tsang-cns.el...

[and it loops forever using a lot of CPU]

    Juanma

bug#2307: Bootstrap hangs in leim

[martin rudalics]

 > Converting cangjie-table.cns to tsang-cns.el...
 >
 > [and it loops forever using a lot of CPU]

Was looping for some ten minutes here on Monday - but got out just when
I was about to kill it.

martin

bug#2307: Bootstrap hangs in leim

[Juanma Barranquero]

On Thu, Feb 12, 2009 at 16:43, martin rudalics <rudalics@gmx.at> wrote:

> Was looping for some ten minutes here on Monday - but got out just when
> I was about to kill it.

Mine looped for ten/fifteen minutes and I just ran out of patience.

Looping forever or not, it is a regression, though.

    Juanma

bug#2307: Bootstrap hangs in leim

[martin rudalics]

Both

   /leim/MISC-DIC/cangjie-table.b5
   /leim/MISC-DIC/cangjie-table.cns

have CRLF line-endings here.

Maybe they should be checked out with -kb?

martin

bug#2307: Bootstrap hangs in leim

[Juanma Barranquero]

> Both
>
>  /leim/MISC-DIC/cangjie-table.b5
>  /leim/MISC-DIC/cangjie-table.cns
>
> have CRLF line-endings here.

Now that you mention it,

  leim/MISC-DIC/cangjie-table.b5
  leim/MISC-DIC/cangjie-table.cns
  leim/MISC-DIC/pinyin.map

have CRCRLF here, so they're CRLF in the repository.

> Maybe they should be checked out with -kb?

Doing that fixes the problem, though I have three questions now:

 1.- Why are these files CRLF in the first place?
 2.- Why the sudden trouble, when the files haven't been modified for years?
 3.- Should the -kb workaround be documented in nt/INSTALL, or the
files changed to LF?

    Juanma

bug#2307: Bootstrap hangs in leim

[Eli Zaretskii]

> Date: Fri, 13 Feb 2009 10:30:14 +0100
> From: Juanma Barranquero <lekktu@gmail.com>
> Cc: 2307@emacsbugs.donarmstrong.com, Eli Zaretskii <eliz@gnu.org>, 
> 	Kenichi Handa <handa@m17n.org>
> 
>   leim/MISC-DIC/cangjie-table.b5
>   leim/MISC-DIC/cangjie-table.cns
>   leim/MISC-DIC/pinyin.map
> 
> have CRCRLF here, so they're CRLF in the repository.
> 
> > Maybe they should be checked out with -kb?
> 
> Doing that fixes the problem, though I have three questions now:
> 
>  1.- Why are these files CRLF in the first place?

I believe that's because these files are distributed in this form by
their original authors, see leim/MISC-DIC/README.

>  3.- Should the -kb workaround be documented in nt/INSTALL, or the
> files changed to LF?

It already is documented, although we may wish to mention these Leim
files explicitly.

bug#2307: Bootstrap hangs in leim

[Juanma Barranquero]

On Fri, Feb 13, 2009 at 11:54, Eli Zaretskii <eliz@gnu.org> wrote:

> I believe that's because these files are distributed in this form by
> their original authors, see leim/MISC-DIC/README.

Shouldn't they be -kb in the repository, then?

> It already is documented, although we may wish to mention these Leim
> files explicitly.

nt/INSTALL talks about these files:

  - nt/*
  - lisp/international/uni-*
  - lisp/ldefs-boot.el

so yes, mentioning it explicitly would be in order.

But I'm still puzzled why the sudden trouble.

    Juanma

bug#2307: Bootstrap hangs in leim

[Eli Zaretskii]

> Date: Fri, 13 Feb 2009 12:04:43 +0100
> From: Juanma Barranquero <lekktu@gmail.com>
> Cc: rudalics@gmx.at, 2307@emacsbugs.donarmstrong.com, handa@m17n.org
> 
> On Fri, Feb 13, 2009 at 11:54, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > I believe that's because these files are distributed in this form by
> > their original authors, see leim/MISC-DIC/README.
> 
> Shouldn't they be -kb in the repository, then?

I don't think so: people who use CVS on Unix will then lose the
ability to update them by patching, as opposed to by sending the whole
file downstream, which is much slower with large files.

> But I'm still puzzled why the sudden trouble.

Me too.

bug#2307: Bootstrap hangs in leim

[Juanma Barranquero]

On Fri, Feb 13, 2009 at 12:18, Eli Zaretskii <eliz@gnu.org> wrote:

> I don't think so: people who use CVS on Unix will then lose the
> ability to update them by patching, as opposed to by sending the whole
> file downstream, which is much slower with large files.

I know, but as you said we're using the files as distributed by their
original authors, and don't seem to change that often:

/cangjie-table.b5/1.1/Sat Mar 31 02:19:51 2001/-ko/
/cangjie-table.cns/1.1/Sat Mar 31 02:19:51 2001/-ko/
/pinyin.map/1.3/Fri Feb  1 16:01:30 2008/-ko/

pinyin.map has had two changes since it was added in 2001-03-31, and
cangjie-table.* none. Perhaps -kb would do more good than bad in this
particular case.

    Juanma

bug#2307: Bootstrap hangs in leim

[Kenichi Handa]

In article <uvdrewqt1.fsf@gnu.org>, Eli Zaretskii <eliz@gnu.org> writes:

> > Date: Fri, 13 Feb 2009 10:30:14 +0100
> > From: Juanma Barranquero <lekktu@gmail.com>
> > Cc: 2307@emacsbugs.donarmstrong.com, Eli Zaretskii <eliz@gnu.org>, 
> > 	Kenichi Handa <handa@m17n.org>
> > 
> >   leim/MISC-DIC/cangjie-table.b5
> >   leim/MISC-DIC/cangjie-table.cns
> >   leim/MISC-DIC/pinyin.map
> > 
> > have CRCRLF here, so they're CRLF in the repository.
> > 
> > > Maybe they should be checked out with -kb?
> > 
> > Doing that fixes the problem, though I have three questions now:
> > 
> >  1.- Why are these files CRLF in the first place?

> I believe that's because these files are distributed in this form by
> their original authors, see leim/MISC-DIC/README.

Yes.

> >  3.- Should the -kb workaround be documented in nt/INSTALL, or the
> > files changed to LF?

> It already is documented, although we may wish to mention these Leim
> files explicitly.

That will be good.  Anyway, I'm now changing
titdic-conert.el (which generates leim/*.el from those
files) to handle all cases (LF, CRLF, CRCRLF).

---
Kenichi Handa
handa@m17n.org

bug#2307: Bootstrap hangs in leim

[Jason Rumney]

Eli Zaretskii wrote:

>>  3.- Should the -kb workaround be documented in nt/INSTALL, or the
>> files changed to LF?
>>     
>
> It already is documented, although we may wish to mention these Leim
> files explicitly.
>   

The leim files should not require it. Some effort was spent a couple of 
years ago tracking down the problems in the lisp functions that process 
the leim files and fixing them to handle the various line ends that CVS 
clients create. There must have been a recent change that broke this.

bug#2307: marked as done (Bootstrap hangs in leim)

[Emacs bug Tracking System]

[-- Attachment #1: Type: text/plain, Size: 875 bytes --]


Your message dated Fri, 13 Feb 2009 13:56:27 +0100
with message-id <f7ccd24b0902130456v1d874c39pd900a005bcec4a11@mail.gmail.com>
and subject line Re: bug#2307: Bootstrap hangs in leim
has caused the Emacs bug report #2307,
regarding Bootstrap hangs in leim
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@emacsbugs.donarmstrong.com
immediately.)


-- 
2307: http://emacsbugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=2307
Emacs Bug Tracking System
Contact owner@emacsbugs.donarmstrong.com with problems

[-- Attachment #2: Type: message/rfc822, Size: 2677 bytes --]

From: Juanma Barranquero <lekktu@gmail.com>
To: Emacs Bug Tracker <submit@emacsbugs.donarmstrong.com>
Subject: Bootstrap hangs in leim
Date: Thu, 12 Feb 2009 12:01:14 +0100
Message-ID: <f7ccd24b0902120301v5c00807x7bb663cef5e07d3a@mail.gmail.com>

Package: emacs
X-Debbugs-CC: handa@m17n.org
X-Debbugs-No-Ack: yes

(After circumventing bug#2304) bootstrapping Emacs hangs at this point:

Converting ziranma.cin to ZIRANMA.el...
Loading vc-cvs...
Converting ziranma.cin to ZIRANMA.el...done
Converting pinyin.map to PY.el...
Converting pinyin.map to PY.el...done
Converting CTLau.html to CTLau.el...
Converting CTLau.html to CTLau.el...done
Converting CTLau-b5.html to CTLau-b5.el...
Converting CTLau-b5.html to CTLau-b5.el...done
Converting cangjie-table.cns to tsang-cns.el...

[and it loops forever using a lot of CPU]

    Juanma



[-- Attachment #3: Type: message/rfc822, Size: 3023 bytes --]

From: Juanma Barranquero <lekktu@gmail.com>
To: 2307-done@emacsbugs.donarmstrong.com
Subject: Re: bug#2307: Bootstrap hangs in leim
Date: Fri, 13 Feb 2009 13:56:27 +0100
Message-ID: <f7ccd24b0902130456v1d874c39pd900a005bcec4a11@mail.gmail.com>

Fixed by this change:

2009-02-13  Kenichi Handa  <handa@m17n.org>

	* international/titdic-cnv.el (tit-process-header): Use
	buffer-substring-no-properties instead of buffer-substring.
	(tit-process-body): Likewise.
	(tsang-quick-converter): Handle superfluous CR at eol.
	(py-converter): Use buffer-substring-no-properties instead of
	buffer-substring.  Handle superfluous CR at eol.
	(ziranma-converter): Likewise
	(ctlau-converter): Likewise.
	(miscdic-convert): Read a file without eol conversion.