Recently, I asked if anybody on the list had some ideas on how to convert
pdf-files to sgml.
I want to thank those five of you on the list who answered my query (in
alphabetical order): Frederik Fouvry, Lee Gillam, Sofie Johansson Kokkinakis,
Daniel Ridings and Larry Spitz.
As it seems, there is no fool-proof way of extracting running text from a pdf
file. (And, of course, that's not very surprising.) However, the easiest way to
do this (and hopefully extract som of the text structure, as well) is to convert
the pdf file to a postscript file first and then try to convert this ps-file to
either text (ps2ascii, pstotext) or html (ps2html). After that you need to write
your own program to tag the text according to whatever DTD you're using. Still,
you are bound to face problems with dehyphenation, absent word boundaries and
peculiar structure.
There are also som web sites that offer pdf to html or pdf to text email
service. This might be useful for converting smaller pdf files. But, who wants
to send sensitive material over the Internet?
Here are some more or less useful web sites (some mentioned by those who
answered the query):
http://www.research.digital.com/SRC/virtualpaper/pstotext.html
http://www.nzdl.org/cgi-bin/gw?a=page&p=Prescript
http://www.pdfzone.com/products/software/tool_pdf2txt.html
http://access.adobe.com/access_tools.html
http://www.bcl-computers.com/
http://www.csd.uch.gr/~nikop/thesis.html
Mail-in-services:
pdf2txt@sun.trace.wisc.edu
pdf2html@sun.trace.wisc.edu
pdf2txt@adobe.com
pdf2html@adobe.com
Thanks for your time!
Mickel Grönroos
University of Helsinki
www.ling.helsinki.fi/~mcgronro/ | Mickel.Gronroos@helsinki.fi
---------------------------------|----------------------------
Inst. för allmän språkvetenskap | Dep. of General Linguistics
PB 4 (Fabiansgatan 28) | tfn/phone +358-9-191 22707
FI-00014 Helsingfors universitet | fax +358-9-191 23598