Can openoffice count words from console?

Question

i have a small problem i need to count words inside the console to read doc, docx, pptx, ppt, xls, xlsx, odt, pdf ... so don't suggest me | wc -w or grep because they work only with text or console output and they count only spaces and in japanese, chinese, arabic , hindu , hebrew they use diferent delimiter so the word count is wrong and i tried to count with this

pdftotext file.pdf -| wc -w
/usr/local/bin/docx2txt.pl < file.docx | wc -w
/usr/local/bin/pptx2txt.pl < file.pptx | wc -w
antiword file.doc -| wc -w 
antiword file.word -| wc -w

in some cases microsoft word , openoffice sad 1000 words and the counters return 10 or 300 words if the language is ( japanese , chinese, hindu ect... ) , but if i use normal characters then i have no issue the biggest mistake is in some case 3 chars less witch is "OK"

i tried to convert with soffice , openoffice and then try WC -w but i can't even convert ,

soffice --headless --nofirststartwizard --accept=socket,host=127.0.0.1,port=8100; --convert-to pdf some.pdf /var/www/domains/vocabridge.com/devel/temp_files/23/0/东京_1000_words_Docx.docx 

OR

 openoffice.org  --headless  --convert-to  ........

OR

openoffice.org3 --invisible 

so if someone know any way to count correctly or display document statistic with openoffice or anything else or linux with the console please share it

thanks.

Answer 1

If you have Microsoft Word (and Windows, obviously) you can write a VBA macro or if you want to run straight from the command line you can write a VBScript script with something like the following:

wordApp = CreateObject("Word.Application")
doc = ... ' open up a Word document using wordApp
docWordCount = doc.Words.Count
' Rinse and repeat...

If you have OpenOffice.org/LibreOffice you have similar (but more) options. If you want to stay in the office app and run a macro you can probably do that. I don't know the StarBasic API well enough to tell you how but I can give you the alternative: creating a Python script to get the word count from the command line. Roughly speaking, you do the following:

• Start up your copy of OOo/LibO from the command line with the appropriate parameters to accept incoming socket connections. http://www.openoffice.org/udk/python/python-bridge.html has instructions on how to do that. Go there and use the browser's in-page find feature to search for `accept=socket'

• Write a Python script to use the OOo/LibO UNO bridge (basically equivalent to the VBScript example above) to open up your Word/ODT documents one at a time and get the word count from each. The above page should give you a good start to doing that.

• You get the word count from a document model object's WordCount property: http://www.openoffice.org/api/docs/common/ref/com/sun/star/text/GenericTextDocument.html#WordCount

Answer 2

I found the answer create one service

#!/bin/sh
#
# chkconfig: 345 99 01
#
# description: your script is a test service
#

(while sleep 1; do
  ls pathwithfiles/in | while read file; do
    libreoffice --headless -convert-to pdf "pathwithfiles/in/$file" --outdir pathwithfiles/out
    rm "pathwithfiles/in/$file"
  done
done) &

then the php script that i needed counted everything

 $ext = pathinfo($absolute_file_path, PATHINFO_EXTENSION);
        if ($ext !== 'txt' && $ext !== 'pdf') {
            // Convert to pdf
            $tb = mktime() . mt_rand();
            $tempfile = 'locationofpdfs/in/' . $tb . '.' . $ext;
            copy($absolute_file_path, $tempfile);
            $absolute_file_path = 'locationofpdfs/out/' . $tb . '.pdf';
            $ext = 'pdf';
            while (!is_file($absolute_file_path)) sleep(1);
        }
        if ($ext !== 'txt') {
            // Convert to txt
            $tempfile = tempnam(sys_get_temp_dir(), '');
            shell_exec('pdftotext "' . $absolute_file_path . '" ' . $tempfile);
            $absolute_file_path = $tempfile;
            $ext = 'txt';
        }
        if ($ext === 'txt') {
            $seq = '/[\s\.,;:!\? ]+/mu';
            $plain = file_get_contents($absolute_file_path);
            $plain = preg_replace('#\{{{.*?\}}}#su', "", $plain);
            $str = preg_replace($seq, '', $plain);
            $chars = count(preg_split('//u', $str, -1, PREG_SPLIT_NO_EMPTY));
            $words = count(preg_split($seq, $plain, -1, PREG_SPLIT_NO_EMPTY));
            if ($words === 0) return $chars;
            if ($chars / $words > 10) $words = $chars;
            return $words;
        }