Re: Retrieve file BLOB from SQL Server 'Image' field type

[choroba]

See Handling BLOB data in ORACLE using perl script, another today's message. Are you colleagues?

[porg]

Hi, thanks for the reply - I read that post before posting this (we are NOT colleagues), and I attempted to integrate his retrieval method using STDOUT, but I still receive the same result. Also it seems his question had a different use-case... Should I always be using STDOUT/fetchrow() for this type of retrieval?

[porg]

This post http://stackoverflow.com/questions/8309181/perl-module-for-generating-doc-or-docx-document-files seems to imply that there is no way for Perl to write a DOCX, however, would this really matter if the data is stored as a blob? It seems that if all the binary data is already there, then all that should be done to reconstruct the DOCX file would be to dump the binary... right?

[Myrddin Wyllt]

Dealing with a general docx file, especially one produced by Word, is difficult because Microsoft don't follow their own specs in all cases. Usually, most specific cases are easier - it's perfectly possible to produce valid, well formatted docx documents using e.g. Template::Toolkit and Archive::Zip, since docx files are just zipped xml at heart.

I have no knowledge of SQL Server BLOB formats, but it might be worth trying to unzip the .docx file you produce to see if:-

1. It's a valid zip file.
2. It contains stuff that looks like xml

If it doesn't look anything like a zip file, you probably need to do some more processing on the BLOB before writing it to file; if it looks docx-ish, you just have to work out what changes you need to make Word eat it.

[porg]

Your recommendation led me to the answer. Thank you for your assistance.