Re: Portuguese spell files update

2010-08-13 Fir de Conversatie Bram Moolenaar

Jakson Aquino wrote:

   On Tue, Aug 10, 2010 at 6:15 PM, Bram Moolenaar b...@moolenaar.net 
   wrote:
Unfortunately, when I run Aap, applying the diff fails. Â It looks 
like
the diff was made for another pt_PT.aff file. Â The one that was
downloaded looks like this:
   
   
SET UTF-8
LANG pt_PT
TRY 
aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÃ
 Yà Éà ÓèÂÚ
KEY qwertyuiop|asdfghjkl|zxcvbnm
WORDCHARS -
   
   
The diff expects:
   
 SET ISO8859-1
- TRY 
aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÃ
 Yà Éà ÓèÂÚ
   
   
Also, the .aff file has lots of trailing text. Â Are we using the 
same
download?
I have made an attempt in fixing this. Â Please check the files on 
the
ftp server: ftp://ftp.vim.org/pub/vim/runtime/spell/pt/
   
Also check spell suggestions with the generated .spl file.
   
  
   It seems that everything is ok now. Thank you very much for fixing it!
  
   I deleted all files from my local spell/pt directory and restored them
   with hg updated -C. I compared the files generated by hg update -C
   with the ones in the ftp server and they have the same md5sums. Then,
   I generated and copied the pt.utf-8.spl and pt.latin1.spl files to
   ~/.vim/spell. European and Brazilian spell check seems to be correct
   and up to date and the spell suggestions are good.
  
   However, the pt.utf-8.spl from the ftp server seems to be outdated...
   Did you do aap clean before generating it?
  
   No, I didn't update the pt_BR files. Â I'll do that now. Â Hmm, I see a
   few warnings...
 
  I see some duplicate warnings...
 
  fixdup.vim can delete 122 duplicates, but about 1600 would remain.
 
  The lines weren't duplicates in the original pt_PT.dic because each
  root entry---even though carrying the same suffixes---had a different
  grammatical category. They became duplicates after the deletion of the
  grammatical information. Most of the remaining duplicates can be
  deleted with:
 
  :2,$ sort u
 
  But about 200 duplicated roots would still remain because they have
  different suffixes. Example:
 
  advindo/L
  advindo/fp
 
  I guess that, if there is no problem in the .dic being sorted, we can
  use :2,$ sort u to delete the real duplicates, but I guess that it's
  better to let the remaining 200 duplicates in the file.
 
  So, what to do?
 
  The duplicates are not a real problem. Â It's useful if you are writing a
  dictionary.
 
  I think using that sort command is good. Â You can verify that with and
  without the sort command the spell file is the same.
 
 I've made a copy of pt_BR.aff, pt_BR.dic, pt_PT.aff and pt_PT.dic into
 four directories. In one directory I ran :so fixdup.vim on
 pt_PT.dic; in other I ran 2,$ sort u; and in the third one I did
 both. Then I did :mkspell. The result:
 
 md5sum pt*/pt.utf*
 9aee104f71db890d10f6cb49525abc0b  ptFixdup/pt.utf-8.spl
 9aee104f71db890d10f6cb49525abc0b  ptFixdupSorted/pt.utf-8.spl
 99bdb0f4a6eaa09503206f604022d2a1  ptOrig/pt.utf-8.spl
 99bdb0f4a6eaa09503206f604022d2a1  ptSorted/pt.utf-8.spl
 
 It's confusing to work with so many files with the same name... If I
 didn't do any mistake the conclusion is: the sort command doesn't
 change the result, but the generated spell file is not the same if
 fixdup.vim is sourced. I'm sending a patch to add the sort command to
 the preparation of pt_PT.dic.

Thanks, this looks safe to include.

-- 
hundred-and-one symptoms of being an internet addict:
52. You ask a plumber how much it would cost to replace the chair in front of
your computer with a toilet.

 /// Bram Moolenaar -- b...@moolenaar.net -- http://www.Moolenaar.net   \\\
///sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\download, build and distribute -- http://www.A-A-P.org///
 \\\help me help AIDS victims -- http://ICCF-Holland.org///

-- 
You received this message from the vim_dev maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php


Re: Portuguese spell files update

2010-08-12 Fir de Conversatie Bram Moolenaar

Jakson Aquino wrote:

  On Tue, Aug 10, 2010 at 6:15 PM, Bram Moolenaar b...@moolenaar.net wrote:
   Unfortunately, when I run Aap, applying the diff fails. Â It looks like
   the diff was made for another pt_PT.aff file. Â The one that was
   downloaded looks like this:
  
  
   SET UTF-8
   LANG pt_PT
   TRY 
   aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÃ
Yà Éà ÓèÂÚ
   KEY qwertyuiop|asdfghjkl|zxcvbnm
   WORDCHARS -
  
  
   The diff expects:
  
   Â SET ISO8859-1
   - TRY 
   aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÃ
Yà Éà ÓèÂÚ
  
  
   Also, the .aff file has lots of trailing text. Â Are we using the same
   download?
   I have made an attempt in fixing this. Â Please check the files on the
   ftp server: ftp://ftp.vim.org/pub/vim/runtime/spell/pt/
  
   Also check spell suggestions with the generated .spl file.
  
 
  It seems that everything is ok now. Thank you very much for fixing it!
 
  I deleted all files from my local spell/pt directory and restored them
  with hg updated -C. I compared the files generated by hg update -C
  with the ones in the ftp server and they have the same md5sums. Then,
  I generated and copied the pt.utf-8.spl and pt.latin1.spl files to
  ~/.vim/spell. European and Brazilian spell check seems to be correct
  and up to date and the spell suggestions are good.
 
  However, the pt.utf-8.spl from the ftp server seems to be outdated...
  Did you do aap clean before generating it?
 
  No, I didn't update the pt_BR files. Â I'll do that now. Â Hmm, I see a
  few warnings...
 
 I see some duplicate warnings...
 
 fixdup.vim can delete 122 duplicates, but about 1600 would remain.
 
 The lines weren't duplicates in the original pt_PT.dic because each
 root entry---even though carrying the same suffixes---had a different
 grammatical category. They became duplicates after the deletion of the
 grammatical information. Most of the remaining duplicates can be
 deleted with:
 
 :2,$ sort u
 
 But about 200 duplicated roots would still remain because they have
 different suffixes. Example:
 
 advindo/L
 advindo/fp
 
 I guess that, if there is no problem in the .dic being sorted, we can
 use :2,$ sort u to delete the real duplicates, but I guess that it's
 better to let the remaining 200 duplicates in the file.
 
 So, what to do?

The duplicates are not a real problem.  It's useful if you are writing a
dictionary.

I think using that sort command is good.  You can verify that with and
without the sort command the spell file is the same.

-- 
hundred-and-one symptoms of being an internet addict:
39. You move into a new house and decide to Netscape before you landscape.

 /// Bram Moolenaar -- b...@moolenaar.net -- http://www.Moolenaar.net   \\\
///sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\download, build and distribute -- http://www.A-A-P.org///
 \\\help me help AIDS victims -- http://ICCF-Holland.org///

-- 
You received this message from the vim_dev maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php


Re: Portuguese spell files update

2010-08-11 Fir de Conversatie Bram Moolenaar

Jakson Aquino wrote:

 On Tue, Aug 10, 2010 at 6:15 PM, Bram Moolenaar b...@moolenaar.net wrote:
  Unfortunately, when I run Aap, applying the diff fails.  It looks like
  the diff was made for another pt_PT.aff file.  The one that was
  downloaded looks like this:
 
 
  SET UTF-8
  LANG pt_PT
  TRY aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÁYÍÉàÓèÂÚ
  KEY qwertyuiop|asdfghjkl|zxcvbnm
  WORDCHARS -
 
 
  The diff expects:
 
   SET ISO8859-1
  - TRY 
  aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÁYÍÉàÓèÂÚ
 
 
  Also, the .aff file has lots of trailing text.  Are we using the same
  download?
  I have made an attempt in fixing this.  Please check the files on the
  ftp server: ftp://ftp.vim.org/pub/vim/runtime/spell/pt/
 
  Also check spell suggestions with the generated .spl file.
 
 
 It seems that everything is ok now. Thank you very much for fixing it!
 
 I deleted all files from my local spell/pt directory and restored them
 with hg updated -C. I compared the files generated by hg update -C
 with the ones in the ftp server and they have the same md5sums. Then,
 I generated and copied the pt.utf-8.spl and pt.latin1.spl files to
 ~/.vim/spell. European and Brazilian spell check seems to be correct
 and up to date and the spell suggestions are good.
 
 However, the pt.utf-8.spl from the ftp server seems to be outdated...
 Did you do aap clean before generating it?

No, I didn't update the pt_BR files.  I'll do that now.  Hmm, I see a
few warnings...

 NOTE: the broffice site was offline. So I used the
 Vero_pt_BR_V207AOC.oxt that I have downloaded to my home directory
 yesterday and did some of the commands manually.
 
 I'm sending a small patch to delete unnecessary files extracted from
 the Brazilian oxt file, as you have done with the files extracted from
 the European oxt.

Thanks.

-- 
A fine is a tax for doing wrong.  A tax is a fine for doing well.

 /// Bram Moolenaar -- b...@moolenaar.net -- http://www.Moolenaar.net   \\\
///sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\download, build and distribute -- http://www.A-A-P.org///
 \\\help me help AIDS victims -- http://ICCF-Holland.org///

-- 
You received this message from the vim_dev maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php


Re: Portuguese spell files update

2010-08-10 Fir de Conversatie Bram Moolenaar

Jakson Aquino wrote:

 The Portuguese dictionaries downloaded by runtime/spell/pt/main.aap
 are outdated. The attached patch fix this and I hope the patch is
 ok...
 
 The European Portuguese list of words has additional information about
 the grammatical category of the words, as in the lines below:
 
 abarrotar/XYPLM [CAT=3Dv,T=3Dinf,TR=3Dt]
 abastan=E7a   [CAT=3Dnc,G=3Df,N=3Ds]
 abastardar/XYPLM[CAT=3Dv,T=3Dinf,TR=3Dt]
 
 I had to delete this information with %s/\t.*// because :mkspell
 wasn't catching words which didn't have suffix rules. Example: among
 the lines above, abastan=E7a wasn't added to Vim's word list. To avoid
 a huge pt_PT.diff file, I deleted the information before generating
 the pt_BR.orig.dic file. Is that ok?

Thanks.

Yes, deleting that extra information is good.  I didn't want to support
this in Vim, because it may hide mistakes.

Unfortunately, when I run Aap, applying the diff fails.  It looks like
the diff was made for another pt_PT.aff file.  The one that was
downloaded looks like this:


SET UTF-8
LANG pt_PT
TRY aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÁYÍÉàÓèÂÚ
KEY qwertyuiop|asdfghjkl|zxcvbnm
WORDCHARS -


The diff expects:

  SET ISO8859-1
- TRY aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÁYÍÉàÓèÂÚ
  
  
Also, the .aff file has lots of trailing text.  Are we using the same
download?

I have made an attempt in fixing this.  Please check the files on the
ftp server: ftp://ftp.vim.org/pub/vim/runtime/spell/pt/

Also check spell suggestions with the generated .spl file.


-- 
A day without sunshine is like, well, night.

 /// Bram Moolenaar -- b...@moolenaar.net -- http://www.Moolenaar.net   \\\
///sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\
\\\download, build and distribute -- http://www.A-A-P.org///
 \\\help me help AIDS victims -- http://ICCF-Holland.org///

-- 
You received this message from the vim_dev maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php


Re: Portuguese spell files update

2010-08-10 Fir de Conversatie Jakson A. Aquino
On Tue, Aug 10, 2010 at 6:15 PM, Bram Moolenaar b...@moolenaar.net wrote:
 Unfortunately, when I run Aap, applying the diff fails.  It looks like
 the diff was made for another pt_PT.aff file.  The one that was
 downloaded looks like this:


 SET UTF-8
 LANG pt_PT
 TRY aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÁYÍÉàÓèÂÚ
 KEY qwertyuiop|asdfghjkl|zxcvbnm
 WORDCHARS -


 The diff expects:

  SET ISO8859-1
 - TRY aerisontcdmlupvgbfzáhçqjíxãóéêâúõACMPSBTELGRIFVDkHJONôywUKXZWQÁYÍÉàÓèÂÚ


 Also, the .aff file has lots of trailing text.  Are we using the same
 download?
 I have made an attempt in fixing this.  Please check the files on the
 ftp server: ftp://ftp.vim.org/pub/vim/runtime/spell/pt/

 Also check spell suggestions with the generated .spl file.


It seems that everything is ok now. Thank you very much for fixing it!

I deleted all files from my local spell/pt directory and restored them
with hg updated -C. I compared the files generated by hg update -C
with the ones in the ftp server and they have the same md5sums. Then,
I generated and copied the pt.utf-8.spl and pt.latin1.spl files to
~/.vim/spell. European and Brazilian spell check seems to be correct
and up to date and the spell suggestions are good.

However, the pt.utf-8.spl from the ftp server seems to be outdated...
Did you do aap clean before generating it?

NOTE: the broffice site was offline. So I used the
Vero_pt_BR_V207AOC.oxt that I have downloaded to my home directory
yesterday and did some of the commands manually.

I'm sending a small patch to delete unnecessary files extracted from
the Brazilian oxt file, as you have done with the files extracted from
the European oxt.

Thank you!

Jakson

-- 
You received this message from the vim_dev maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
diff -r 6efdc2e61533 runtime/spell/pt/main.aap
--- a/runtime/spell/pt/main.aap	Tue Aug 10 22:53:02 2010 +0200
+++ b/runtime/spell/pt/main.aap	Tue Aug 10 23:13:54 2010 -0300
@@ -73,6 +73,12 @@
 :fetch $BR_FNAME
 :sys $UNZIP $BR_FNAME
 :delete $BR_FNAME
+:delete {f} description.xml
+:delete {f} dictionaries.xcu
+:delete {f} hyph_pt_BR.dic
+:delete {r}{f} META-INF
+:delete {f} README_en.TXT
+:delete {f} README_hyph_pt_BR.TXT
 :sys $VIM README_pt_BR.TXT -u NONE -N -e -c set ff=unix -c update -c q
 :move README_pt_BR.TXT README_pt_BR.txt