Identifying context files

Identifying context files

[Aditya Mahajan]

I use (g)vim to edit both context and latex files. Unfortunately, both 
of them usually have *.tex extension. This mean that detecting 
filetype from extension is not possible, so one should look into the 
contents of the file to see if it a context file or not.

I am planning to submit a ftdetect for context to vim. Right now, I 
check if the first six lines of the file contain any of 
'\\start\|\\enablemode\|\\unprotect\|\\setvariables\|\\module\|\\usemodule' 
and if so, set the filetype to context, otherwise it is set to tex 
(that loads latex plugins).

This works for my context writing style. I would like to know about 
other people's preference.

1. Do you write some keyword unique to context in the first few lines 
of the file. Should I also check the last few line lines.

2. Are there any other keywords that you will like to include.

3. Is it enough to check the first 6 line or should I check more. I do 
not want to check more lines as this will make the detection slower 
(by a few mili secs).

Thanks,
Aditya

-- 
Aditya Mahajan, EECS Systems, University of Michigan
http://www.eecs.umich.edu/~adityam || Ph: 7342624008

Re: Identifying context files

[Thomas A. Schmitz]

That sounds like an excellent idea, and I'd be very grateful to have  
such a detector. As to keywords: most of my ConTeXt files start with  
\enableregime; you may want to add this to your list.

Best

Thomas


On Feb 3, 2006, at 3:17 AM, Aditya Mahajan wrote:

> I use (g)vim to edit both context and latex files. Unfortunately, both
> of them usually have *.tex extension. This mean that detecting
> filetype from extension is not possible, so one should look into the
> contents of the file to see if it a context file or not.
>
> I am planning to submit a ftdetect for context to vim. Right now, I
> check if the first six lines of the file contain any of
> '\\start\|\\enablemode\|\\unprotect\|\\setvariables\|\\module\|\ 
> \usemodule'
> and if so, set the filetype to context, otherwise it is set to tex
> (that loads latex plugins).
>
> This works for my context writing style. I would like to know about
> other people's preference.
>
> 1. Do you write some keyword unique to context in the first few lines
> of the file. Should I also check the last few line lines.
>
> 2. Are there any other keywords that you will like to include.
>
> 3. Is it enough to check the first 6 line or should I check more. I do
> not want to check more lines as this will make the detection slower
> (by a few mili secs).
>
> Thanks,
> Aditya
>
> -- 
> Aditya Mahajan, EECS Systems, University of Michigan
> http://www.eecs.umich.edu/~adityam || Ph: 7342624008
> _______________________________________________
> ntg-context mailing list
> ntg-context@ntg.nl
> http://www.ntg.nl/mailman/listinfo/ntg-context

Re: Identifying context files

[Hans Hagen]

Thomas A. Schmitz wrote:
> That sounds like an excellent idea, and I'd be very grateful to have  
> such a detector. As to keywords: most of my ConTeXt files start with  
> \enableregime; you may want to add this to your list.
>
> Best
>
> Thomas
>
>
> On Feb 3, 2006, at 3:17 AM, Aditya Mahajan wrote:
>
>   
>> I use (g)vim to edit both context and latex files. Unfortunately, both
>> of them usually have *.tex extension. This mean that detecting
>> filetype from extension is not possible, so one should look into the
>> contents of the file to see if it a context file or not.
>>
>> I am planning to submit a ftdetect for context to vim. Right now, I
>> check if the first six lines of the file contain any of
>> '\\start\|\\enablemode\|\\unprotect\|\\setvariables\|\\module\|\ 
>> \usemodule'
>> and if so, set the filetype to context, otherwise it is set to tex
>> (that loads latex plugins).
>>
>> This works for my context writing style. I would like to know about
>> other people's preference.
>>
>> 1. Do you write some keyword unique to context in the first few lines
>> of the file. Should I also check the last few line lines.
>>     
convention:

% interface=en language=nl program=pdfetex

etc, the following are understood by texexec

        ['tex','texengine'],
        ['program','texengine'],
        ['translate','tcxfilter'],
        ['tcx','tcxfilter'],
        ['output','backend'],
        ['mode','mode'],
        ['ctx','ctxfile'],
        ['version','contextversion'],
        ['format','texformats'],
        ['interface','texformats']

>> 2. Are there any other keywords that you will like to include.
>>
>> 3. Is it enough to check the first 6 line or should I check more. I do
>> not want to check more lines as this will make the detection slower
>> (by a few mili secs).
>>
>>     
checking does not take much time, for instance in scite, i check till i 
know; also (probably goes unnoticed), texexec does soem checking: it 
needs to figure out the interface:


Hans

Re: Identifying context files

[Taco Hoekwater]

Hans Hagen wrote:
> 
> checking does not take much time, for instance in scite, i check till i 
> know; also (probably goes unnoticed), texexec does soem checking: it 
> needs to figure out the interface:

While on that subject, can we make texexec honour

   % format={latex,pdflatex,plain,pdfetex}

etc.? No need to do smart things, just run the command
that results from engine+format, once.

Taco

Re: Identifying context files

[Hans Hagen]

Taco Hoekwater wrote:
> Hans Hagen wrote:
>   
>> checking does not take much time, for instance in scite, i check till i 
>> know; also (probably goes unnoticed), texexec does soem checking: it 
>> needs to figure out the interface:
>>     
>
> While on that subject, can we make texexec honour
>
>    % format={latex,pdflatex,plain,pdfetex}
>
> etc.? No need to do smart things, just run the command
> that results from engine+format, once.
>   
the ruby version of texexec can handle that (if i did it right); 

Hans 

-----------------------------------------------------------------
                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
-----------------------------------------------------------------

Re: Identifying context files

[Aditya Mahajan]

>> On Feb 3, 2006, at 3:17 AM, Aditya Mahajan wrote:
>>
>>> 3. Is it enough to check the first 6 line or should I check more. I do
>>> not want to check more lines as this will make the detection slower
>>> (by a few mili secs).

On Fri, 3 Feb 2006, Hans Hagen wrote:
>
> checking does not take much time, for instance in scite, i check till i
> know; also (probably goes unnoticed), texexec does soem checking: it
> needs to figure out the interface:
>

Can you send me the scite code which does the checking. I do not know lua, 
but it should not be too hard to covnert it to VimL. I do not want to 
reinvent the wheel.

Aditya

Re: Identifying context files

[Hans Hagen]

Aditya Mahajan wrote:
>>> On Feb 3, 2006, at 3:17 AM, Aditya Mahajan wrote:
>>>
>>>       
>>>> 3. Is it enough to check the first 6 line or should I check more. I do
>>>> not want to check more lines as this will make the detection slower
>>>> (by a few mili secs).
>>>>         
>
> On Fri, 3 Feb 2006, Hans Hagen wrote:
>   
>> checking does not take much time, for instance in scite, i check till i
>> know; also (probably goes unnoticed), texexec does soem checking: it
>> needs to figure out the interface:
>>
>>     
>
> Can you send me the scite code which does the checking. I do not know lua, 
> but it should not be too hard to covnert it to VimL. I do not want to 
> reinvent the wheel.
>   

her eis the ruby method used in newtexexec:

    def scantexcontent(filename)
        if FileTest.file?(filename) and tex = File.open(filename) then
            while str = tex.gets do
                case str.chomp
                    when /^\%/o then
                        # next
                    when /\\(starttekst|stoptekst|startonderdeel|startdocument|startoverzicht)/o then
                        setvariable('texformats','nl') ; break
                    when /\\(stelle|verwende|umgebung|benutze)/o then
                        setvariable('texformats','de') ; break
                    when /\\(stel|gebruik|omgeving)/o then
                        setvariable('texformats','nl') ; break
                    when /\\(use|setup|environment)/o then
                        setvariable('texformats','en') ; break
                    when /\\(usa|imposta|ambiente)/o then
                        setvariable('texformats','it') ; break
                    when /(height|width|style)=/o then
                        setvariable('texformats','en') ; break
                    when /(hoehe|breite|schrift)=/o then
                        setvariable('texformats','de') ; break
                    when /(hoogte|breedte|letter)=/o then
                        setvariable('texformats','nl') ; break
                    when /(altezza|ampiezza|stile)=/o then
                        setvariable('texformats','it') ; break
                    when /externfiguur/o then
                        setvariable('texformats','nl') ; break
                    when /externalfigure/o then
                        setvariable('texformats','en') ; break
                    when /externeabbildung/o then
                        setvariable('texformats','de') ; break
                    when /figuraesterna/o then
                        setvariable('texformats','it') ; break
                end
            end
            tex.close
        end

    end


-----------------------------------------------------------------
                                          Hans Hagen | PRAGMA ADE
              Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
     tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com
                                             | www.pragma-pod.nl
-----------------------------------------------------------------

Re: Identifying context files

[Mojca Miklavec]

On 2/2/06, Aditya Mahajan wrote:
> I use (g)vim to edit both context and latex files. Unfortunately, both
> of them usually have *.tex extension. This mean that detecting
> filetype from extension is not possible, so one should look into the
> contents of the file to see if it a context file or not.
>
> I am planning to submit a ftdetect for context to vim. Right now, I
> check if the first six lines of the file contain any of
> '\\start\|\\enablemode\|\\unprotect\|\\setvariables\|\\module\|\\usemodule'
> and if so, set the filetype to context, otherwise it is set to tex
> (that loads latex plugins).
>
> This works for my context writing style. I would like to know about
> other people's preference.
>
> 1. Do you write some keyword unique to context in the first few lines
> of the file. Should I also check the last few line lines.

\enableregime, \setupoutput

Is \setup too general (will it recognize any LaTeX document)?

I don't know how slow/fast the detection is since I only use short
files, but if you check the last few lines, \stoptext (together with
its international alternatives) isn't a bad idea.

Also, if you check the last few lines: emacs users usually put some
lines of code at the end to mark the document as being written in
ConTeXt.

> 2. Are there any other keywords that you will like to include.

A keyword like
set tex_preferred_dialect=latex/context :)

and then

if tex_preferred_dialect == context
    if the first few lines contain \documentclass or \documentstyle,
    declare it latex, otherwise ConTeXt
else
   [your script]:
   if the first line contains % tex= or any keywords that Hans sent
   or if the first six lines contain the keywords you suggested
(together with international alternatives)

(don't take that suggestion too seriously)

Mojca

Re: Identifying context files

[Aditya Mahajan]

<--- On Feb 4, Mojca Miklavec wrote --->

> On 2/2/06, Aditya Mahajan wrote:
>> I use (g)vim to edit both context and latex files. Unfortunately, both
>> of them usually have *.tex extension. This mean that detecting
>> filetype from extension is not possible, so one should look into the
>> contents of the file to see if it a context file or not.
>>
>> I am planning to submit a ftdetect for context to vim. Right now, I
>> check if the first six lines of the file contain any of
>> '\\start\|\\enablemode\|\\unprotect\|\\setvariables\|\\module\|\\usemodule'
>> and if so, set the filetype to context, otherwise it is set to tex
>> (that loads latex plugins).
>>
>> This works for my context writing style. I would like to know about
>> other people's preference.
>>
>> 1. Do you write some keyword unique to context in the first few lines
>> of the file. Should I also check the last few line lines.
>
> \enableregime, \setupoutput
>
> Is \setup too general (will it recognize any LaTeX document)?

That is the main trouble. In principle, you can have latex document 
with \starttext as the first line. But most usual tex files do not 
have it. I have not seen a latex macro with \setup. So, I will include 
that also.

> I don't know how slow/fast the detection is since I only use short
> files, but if you check the last few lines, \stoptext (together with
> its international alternatives) isn't a bad idea.

The file may not necessarily end with \stoptext. It can be 
\stopcomponent, \stopproduct, \stopproject, \stopenvironment. Anything 
else?

> Also, if you check the last few lines: emacs users usually put some
> lines of code at the end to mark the document as being written in
> ConTeXt.

Ah, compatibility with Emacs. I will look at that later ;)

>> 2. Are there any other keywords that you will like to include.
>
> A keyword like
> set tex_preferred_dialect=latex/context :)

I looked it up and it seems that g:Tex_Flavor is meant for this kind 
of thing.

> and then
>
> if tex_preferred_dialect == context
>    if the first few lines contain \documentclass or \documentstyle,
>    declare it latex, otherwise ConTeXt
> else
>   [your script]:
>   if the first line contains % tex= or any keywords that Hans sent
>   or if the first six lines contain the keywords you suggested
> (together with international alternatives)

Actually, this may not be too hard to implement. Let me have a look at 
this, and I will get back to you when there is something working.

Aditya
-- 
Aditya Mahajan, EECS Systems, University of Michigan
http://www.eecs.umich.edu/~adityam || Ph: 7342624008

Re: Identifying context files

[Aditya Mahajan]

<--- On Feb 10, Aditya Mahajan wrote --->

> <--- On Feb 4, Mojca Miklavec wrote --->
>
>> if tex_preferred_dialect == context
>>    if the first few lines contain \documentclass or \documentstyle,
>>    declare it latex, otherwise ConTeXt
>> else
>>   [your script]:
>>   if the first line contains % tex= or any keywords that Hans sent
>>   or if the first six lines contain the keywords you suggested
>> (together with international alternatives)
>
> Actually, this may not be too hard to implement. Let me have a look at
> this, and I will get back to you when there is something working.

This is the else part of the script (modified from Hans ruby code to 
vimL). I tested it on the sources of the manuals, and the only thing 
that it could not detect correctly was metafun/mfun-ef.tex. Looking at 
that file, I do not think that it contains any context specific 
keyword.

Can others test on their context projects and see if this works 
correctly. I do not know the itnernational alternatives of all the 
commands. If your document uses any of those, let me know. Also check 
your latex documents to see that the script does not recognise it as 
context.

To run the script, save it as identify.vim . Inside vim,
:source identify.vim
:cd directory containing tex files
:argadd *.tex
:redir @a
:argdo call Identify_ConTeXt()
:redir END
:e temp_file
:put a

Now you have all the vim messages in temp_file

:g!/ConTeXt/d

Get rid of lines not containing context.

The message format from the script is
filename ConTeXt not detected/ detected at line ... format (en|nl|de|it)

Here is the script
"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""

" A bit more detailed evaluation for english. \\set and \\use may be too
" ambiguous. Checking for setvariables and use encoding

let g:ConTeXt_en =
       \ '\\\(starttext\|startproject\|startproduct\|startcomponent\|startenvironment'.
       \ '\|enablemode\|enableregime\|setvariables\|useencoding\|setup' .
       \ '\|starttypescript' .
       \ '\|usemodule\|externalfigure\)'

let g:ConTeXt_nl =
       \ '\\\(starttekst\|startprodukt\|startprojekt\|startonderdeel\|startdocument\|startoverzicht'.
       \ '\|stel\|gebruik\|omgeving' .
       \ '\|externfiguur\)'

let g:ConTeXt_de =
       \ '\\\(stelle\|verwende\|umgebung\|benutze' .
       \ '\|externeabbildung\)'

let g:ConTeXt_it =
       \  '\\\(usa\|imposta\|ambiente' .
       \  '\|figuraesterna\)'

let g:ConTeXt_generic =
       \  '\\\(protect\|unprotect\|startmode\|mainlanguage\)'


function! Identify_ConTeXt()
   let g:ConTeXt_texformat = ""
   let l:last_line_num = line('$')
   let l:curr_line_num = 1
   while l:curr_line_num <= l:last_line_num
     let l:curr_line = getline(l:curr_line_num)
     if l:curr_line !~ '^%'
       if l:curr_line =~ g:ConTeXt_en
         let g:ConTeXt_texformat = 'en'
         break
       endif
       if l:curr_line =~ g:ConTeXt_generic
         let g:ConTeXt_texformat = 'generic'
       endif
       if l:curr_line =~ g:ConTeXt_nl
         let g:ConTeXt_texformat = 'nl'
         break
       endif
       if l:curr_line =~ g:ConTeXt_de
         let g:ConTeXt_texformat = 'de'
         break
       endif
       if l:curr_line =~ g:ConTeXt_it
         let g:ConTeXt_texformat = 'it'
         break
       endif
     endif
     let l:curr_line_num = l:curr_line_num + 1
   endwhile
   if g:ConTeXt_texformat == ""
     echo bufname("%") . ' ConTeXt not detected!'
   else
     echo bufname("%") . ' ConTeXt detected at line ' . l:curr_line_num .
    \  ' format: ' . g:ConTeXt_texformat
   endif
endf

"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""

  -- 
Aditya Mahajan, EECS Systems, University of Michigan
http://www.eecs.umich.edu/~adityam || Ph: 7342624008