Hi, it depends on the image format and how it is stored. We use a 3rd
party program called "wvWare" to convert MS Word files into .html. It
can extract some images but not others.

> 2) if in source doc. there was inserted picture from a file, then during
> importing in Log file was observed double paths like C:Program
> FilesgsdlcollectdocfilestmpC:Program
> Filesgsdlcollectdocfiles mpD10.jpg and as result images were not
> accessible from collection.
>
> When I installed Greenstone into directory like C:gsdl , then the
> second problem mentioned above
> disappeared (the 1st one still stays).
>
> It looks the problem is about Paths with space inside...(C:Program
> Files) !!

Thanks for reporting this... looking at the code it looks like it
checks that links consist of numbers, letters, "."s and "/".
Could you look in your collection's etc/fail.log - I think it prints
out "HTMLPlug: ERROR - badly formatted tag ignored" for any link that
doesn't match that rule.

According to the standards (eg http://www.ietf.org/rfc/rfc2396.txt),
spaces aren't allowed in URIs/URLs (they should be escaped as %20), so
we need to track down whether greenstone or wvWare is at fault here, but
either way this will definitely be fixed for the next release of greenstone.

Thanks
John McPherson

greenstone or wvWare