The focus of this page ('anyone think I should re-do it as a Wiki?) is on the products available to convert to and from PDF images. IDR Solutions explains the challenge.
Ghostscript/Ghostview answers many questions, at least partially.
David Boddie's pdftools and David Leonard's PDFFile provide interesting Python-coded raw materials for those unafraid of dirtying their hands with programming. Early in 2005, one appreciated correspondent wrote me that the latter "handles things like decryption better."
.pdf and/or .ps source. Here
are a few of the alternatives:
gs -dBATCH -dNOPAUSE -q -sDEVICE=pdfwrite \
-sOutputFile=result.pdf input1.pdf input2.pdf
pdftk A=in1.pdf B=in2.pdf cat B1 A1-7even A1-7odd \
output out1.pdf
I have yet to test it.
At least, that's my usual first response, although, as 2004 begins, a couple of products are making me soften that stance. I understand all the situations that make text-extraction appear to be desirable; I've lived through most of them myself. As several sages have counseled, however, from a programmatic standpoint, "think of PDF as paper", by which they mean you could use scissors and glue on it, but there's almost certainly a better way. Almost always, you're--we're--better off going upstream to the data where the PDFs originated.
If you insist on extracting text from PDF, get help, probably from the following list. This list remains partial; you're welcome to write me to ask that I unpack more of my notes.
pdftotext.exe.
[doesn't handle compression?]
Yes, HTMLDOC is both commercial and free.
In 2007, I began to use iText also for PDF transformations.
All other HTML->PS or HTML->PDF products apparently don't automate well and/or are available only for Windows.
While I still have no experience with Win*-based activePDF WebGrabber, its function apparently is exactly to convert HTML to PDF.
The only independent converter I've found so far is PStill.
Adobe certainly wants people to think of it as the vendor-of-preference for all such needs. I respect Adobe for their business success and technical achievements. My experience as a front-line customer of theirs is ... mixed. My first instinct is to look for alternatives.
The dominant producers of PDF documents in the current market are Acrobat and Word. I suspect someone has reasonably accurate measurements of the share each holds; my rough impression is that the latter dominates. It certainly is feasible to automate Word in principle. While most Word scripters use VBA, I rely most on Tcl or Python ... There should be no effective barriers to full automation using Word's built-in facilities.
Word, however, emits bad PDF, and is often slow and unreliable, at least for the tasks that matter to me. Adobe frustrates me; I have a terrible history at trying to find out the simplest product information from the company. When I want "industrial-strength" automation, I turn to Antiword or OpenOffice. The latter produces higher-quality PDF than Word, and is more open about its scripting capabilities, at least on an ideologic level.
For special purposes, I've built even more involved "production lines" involving intermediate steps with PS, TeX, and other formats and technologies.
Enfocus Pitstop is a PDF preflight and editing package for the print industry.
PDF Crystal ...
[Explain capabilities and applicability of pdflatex, pdfpages ...]
Storypad ...
Cameron
Laird's personal notes on PDF conversion
utilities/claird@phaseit.net