En réponse à NOKUBI Takatsugu <knok(_at_)daionet(_dot_)gr(_dot_)jp>:
HEllo
It works a bit better, but not very good. I've take the version 1.30 of
msword.pl
I have the errors, when creating index with mknmz:
** WARNING **: warning: file '/xxx/NMZ.word.tmp': 245721 bytes, non-integer
number of blocks
** WARNING **: Tried to mark as unused the block 471 which has -1 (several
times this errors, with blocks 471 to 474)
** WARNING **: Invalid root chain
** WARNING **: '/xxx/NMZ.word.tmp' : duff file!
I have these warnings with all .doc files (I have only doc files to be
indexed). Are these warnings of mknmz or of wvHtml?
My 2 doc files have been added to the index, but with only 13 keywords (there's
must be at least 100 different keywords!). I think these keywords are not words
in the text, but tje name of the document, name of auhor maybe,...
Thank you for your help.
I don't understand wha's wrong, now, because it used to work on my computer!
Regards
Julien Sobrier
In article <1026224144(_dot_)3d2af010533cb(_at_)imp(_dot_)free(_dot_)fr>
jusob(_at_)free(_dot_)fr writes:
But now, I have always a problem when trying to create a new
index: /home/julien/test/index/NMZ.word2.tmp: No such file or
directory
I do mknmz --output-dir=/home/julien/test/index
/home/julien/test/data
No file ist created in /home/julien/tet/index
I've trie to create a blak file NMZ.word2.tmp in
/home/julien/test/index, but I
have always the same problem.
I think you used Namazu with lastest wvWare. Latest released Namazu
supports a little bit old one. If you want to use it, get latest
msword filter from viewCVS
<http://cvs.namazu.org/namazu/filter/msword.pl> and replace it.
The problem will be fixed in the next release.
--
NOKUBI Takatsugu
E-mail: knok(_at_)daionet(_dot_)gr(_dot_)jp
knok(_at_)namazu(_dot_)org / knok(_at_)debian(_dot_)org