How to Parse Microsoft Word Document

rubyonrails-talk
1

Hi People,

I just joined the group and I want to ask something about my problem. I'm still learning Ruby on Rails and now I have a task to parse Microsoft Word and store the content into database.

Do you have any suggestion how to do it?

FYI, I develop it under Unix Environment. So, I don't have a chance to use win32ole on it, CMIIW.

I also have searched the internet about this. But all I found that I need to use JRuby and combine it with Apache POI or else I need to use win32ole. As far as I know, to use JRuby I need to create the rails project also with JRuby but unfortunately I already created the project with plain Ruby.

So, I don't know what to do anymore. Does anybody have clue?

Regards,

Hafiz Badrie Lubis

2

You can run poi as a separate process and then grab its output.

3

Hi People,

I just joined the group and I want to ask something about my problem. I'm still learning Ruby on Rails and now I have a task to parse Microsoft Word and store the content into database.

Do you have any suggestion how to do it?

FYI, I develop it under Unix Environment. So, I don't have a chance to use win32ole on it, CMIIW.

I also have searched the internet about this. But all I found that I need to use JRuby and combine it with Apache POI or else I need to use win32ole. As far as I know, to use JRuby I need to create the rails project also with JRuby but unfortunately I already created the project with plain Ruby.

So, I don't know what to do anymore. Does anybody have clue?

I did a project in PHP quite a few years ago, and I used some venerable unix cli converters to do this. I stored the files as is, and then used these converters to rip out their text and stored that in the database for searching. They aren't perfect, but they do a good enough job for search results.

$translators = array(   'pdf' => '/usr/local/bin/pdftotext ./pdf/%s.pdf -',   'ppt' => '/usr/local/bin/catppt -d ascii ./ppt/%s.ppt',   'xls' => '/usr/local/bin/xls2csv -d ascii ./xls/%s.xls',   'doc' => '/usr/local/bin/catdoc -d ascii ./doc/%s.doc' ); //these translators all pipe to stdout, which means that shell_exec will return their text value

Walter

4

1. Convert .doc to .pdf with PyODConverter

2. Convert .pdf to .tiff with ImageMagick

3. Process .tiff through Tesseract OCR and get .txt

5

Can you show it to me how to do it? Do you have a reference? To make a collaboration between a rails project with JRuby codes.

6

It has nothing whatsoever to do with JRuby. You can run Java apps from Ruby exactly like any other command-line process. I don't know if POI is just a library, or has a full app utility as well. If it's just a lib, you'd have to write the program, probably a half-dozen lines of Java.

7

Wow, talk about a long slow way to potentially lose text flow and introduce errors...

8

Ok thank you, Scott. I'll try your advice.

9

I'm coming to this late and I've partially deleted the thread, so I may be way off base...

An old plugin might be of help:

https://github.com/kete/convert_attachment_to

It makes use of existing command line converter utility programs.

Cheers, Walter