regexp to strip off LaTeX command

regexp to strip off LaTeX command

[henry atting]

Hi,

I am searching a regexp for `replace-regexp' to remove only one LaTeX
command. Let's say I want to remove all \textit{} commands but not the
text within the braces, how can I do this?

henry

Re: regexp to strip off LaTeX command

[Andreas Politz]

henry atting wrote:
> Hi,
> 
> I am searching a regexp for `replace-regexp' to remove only one LaTeX
> command. Let's say I want to remove all \textit{} commands but not the
> text within the braces, how can I do this?
> 
> henry
> 

\\textit{\([^}]*\)} -> \1

If you need to handle escaped brackets the regexp gets a little longer.

-ap

Re: regexp to strip off LaTeX command

[henry atting]

Zitat - Andreas Politz * Do Dez 18 2008 um 10:19 -

> henry atting wrote:
>> Hi,
>>
>> I am searching a regexp for `replace-regexp' to remove only one LaTeX
>> command. Let's say I want to remove all \textit{} commands but not the
>> text within the braces, how can I do this?
>>
>> henry
>>
>
> \\textit{\([^}]*\)} -> \1
>
> If you need to handle escaped brackets the regexp gets a little longer.
>
> -ap

Thanks, works fine. :)

I understand the first part but can you please explain what `\1' stands
for?

henry

Re: regexp to strip off LaTeX command

[Andreas Politz]

henry atting wrote:
> Zitat - Andreas Politz * Do Dez 18 2008 um 10:19 -
> 
>> henry atting wrote:
>>> Hi,
>>>
>>> I am searching a regexp for `replace-regexp' to remove only one LaTeX
>>> command. Let's say I want to remove all \textit{} commands but not the
>>> text within the braces, how can I do this?
>>>
>>> henry
>>>
>> \\textit{\([^}]*\)} -> \1
>>
>> If you need to handle escaped brackets the regexp gets a little longer.
>>
>> -ap
> 
> Thanks, works fine. :)
> 
> I understand the first part but can you please explain what `\1' stands
> for?
> 
> henry

\1 is a backreference and refers to the 1st parenthese group, much like
a variable.

Here is a better explanation, somewhere in the middle.

(info "(elisp)Regexp Backslash")

-ap

Re: regexp to strip off LaTeX command

[Arnaldo Mandel]

[-- Attachment #1: Type: text/plain, Size: 701 bytes --]

On Thu, Dec 18, 2008 at 7:19 AM, Andreas Politz <politza@fh-trier.de> wrote:

> henry atting wrote:
>
>> Hi,
>>
>> I am searching a regexp for `replace-regexp' to remove only one LaTeX
>> command. Let's say I want to remove all \textit{} commands but not the
>> text within the braces, how can I do this?
>>
>> henry
>>
>>
> \\textit{\([^}]*\)} -> \1
>
> If you need to handle escaped brackets the regexp gets a little longer.
>

They can get impossibly longer.  There is no way to get regexps to do this
job in full.
Your best bet is to use an utility called detex together with
shell-command-on-region.
detex is in the Debian package texlive-extra-utils, but can certainly be
obtained directly.

am

[-- Attachment #2: Type: text/html, Size: 1182 bytes --]

Re: regexp to strip off LaTeX command

[Plamen Tanovski]

henry atting <nspm_01@literaturlatenight.de> writes:

> I am searching a regexp for `replace-regexp' to remove only one LaTeX
> command. Let's say I want to remove all \textit{} commands but not the
> text within the braces, how can I do this?

This is what I use:

(defun remove-LaTeX-command  (befehl)
"removes simple LaTeX command (not starred and without opt. args)
  and the surrounding braces"
(interactive "scommand (without the backslash): ")
(save-excursion
  (goto-char (point-min))
  (while
      (re-search-forward (concat "\\\\" befehl "{") nil t)
    (replace-match "{")
    (save-excursion
      (goto-char (TeX-find-balanced-brace))
      (delete-char -1))
    (delete-char -1))))

best regards

RE: regexp to strip off LaTeX command

[Bourgneuf Francois]

> -----Message d'origine-----
> De : 
> help-gnu-emacs-bounces+francois.bourgneuf=groupe-mma.fr@gnu.or
> g 
> [mailto:help-gnu-emacs-bounces+francois.bourgneuf=groupe-mma.f
> r@gnu.org] De la part de henry atting
> Envoyé : jeudi 18 décembre 2008 10:48
> À : help-gnu-emacs@gnu.org
> Objet : Re: regexp to strip off LaTeX command
> 
> Zitat - Andreas Politz * Do Dez 18 2008 um 10:19 -
> 
> > henry atting wrote:
> >> Hi,
> >>
> >> I am searching a regexp for `replace-regexp' to remove 
> only one LaTeX
> >> command. Let's say I want to remove all \textit{} commands 
> but not the
> >> text within the braces, how can I do this?
> >>
> >> henry
> >>
> >
> > \\textit{\([^}]*\)} -> \1
> >
> > If you need to handle escaped brackets the regexp gets a 
> little longer.
> >
> > -ap
> 
> Thanks, works fine. :)
> 
> I understand the first part but can you please explain what 
> `\1' stands
> for?
> 
> henry
> 

\1 stands for "what is found between the first parenthesis \( [^}]* \)"
Example :
\(foo\).*\(bar\) ->\1 \2 would return foo bar (if foo and bar were found in the text, of course).
Bour9

Re: regexp to strip off LaTeX command

[henry atting]

Zitat - Bourgneuf Francois * Fr Dez 19 2008 um 10:47 -

>> -----Message d'origine-----
>> De : 
>> help-gnu-emacs-bounces+francois.bourgneuf=groupe-mma.fr@gnu.or
>> g 
>> [mailto:help-gnu-emacs-bounces+francois.bourgneuf=groupe-mma.f
>> r@gnu.org] De la part de henry atting
>> Envoyé : jeudi 18 décembre 2008 10:48
>> À : help-gnu-emacs@gnu.org
>> Objet : Re: regexp to strip off LaTeX command
>> 
>> Zitat - Andreas Politz * Do Dez 18 2008 um 10:19 -
>> 
>> > henry atting wrote:
>> >> Hi,
>> >>
>> >> I am searching a regexp for `replace-regexp' to remove 
>> only one LaTeX
>> >> command. Let's say I want to remove all \textit{} commands 
>> but not the
>> >> text within the braces, how can I do this?
>> >>
>> >> henry
>> >>
>> >
>> > \\textit{\([^}]*\)} -> \1
>> >
>> > If you need to handle escaped brackets the regexp gets a 
>> little longer.
>> >
>> > -ap
>> 
>> Thanks, works fine. :)
>> 
>> I understand the first part but can you please explain what 
>> `\1' stands
>> for?
>> 
>> henry
>> 
>
> \1 stands for "what is found between the first parenthesis \( [^}]* \)"
> Example :
> \(foo\).*\(bar\) ->\1 \2 would return foo bar (if foo and bar were
> found  in the text, of course).
> Bour9

Copy that ;) 
Great thanks to all! 

Detex is good for stripping off *all* commands of a region but for
a single command I find `query-replace-regexp' more convenient.

henry