Probably not with me. I never really worked with Visual Basic, that stuff was done in C++. And for Word, we used the available docs and yes we knew that it was

Probably not with me. I never really worked with Visual Basic, that stuff was done in C++. And for Word, we used the available docs and yes we knew that it was a compound file as an outer shell (while we wrote our own fault-tolerant parser for it, which worked around damaged pieces whenever possible, while initial development was done by using intact files and using the Docfile API). The devil was deeper, the format of document stream and formatting stream was quite complicated, and documentation was very outdated and lacking detail. Extracting just the text was easy - simply go through plcfpcd and extract positions of pieces and their encoding (ASCII or Unicode), convert all to unicode and print to standard output. Extracting formatting (we used RTF for output) was crazy. PST was indeed much, much harder to do, it was a crazy effort which took about a man-year and it was all done by hand. And we never had web frontend for it, at least not on my watch (i left in 2007)