Chris Mohler wrote:
On Tue, Mar 18, 2008 at 12:02 PM, Tom Holroyd <tomh@xxxxxxxxxxxxxxxxxxx> wrote:
PDF is an output format. Much like the way that a C program is compiled
into an executable, a source .doc or .tex file is converted to PDF for
display. You need the source code to edit. Open Source!
I don't think C is a very good analogy. All Illustrator files are PDF
format these days and many programs can edit PDFs - I just haven't
come across anything very good that runs on linux yet. PDFEditor
looks promising, but I don't find it usable yet. If you wrote a
program that followed Adobe's specs, nothing would prevent you from
reading/editing/writing PDF files:
http://www.adobe.com/devnet/pdf/pdf_reference.html
I generally have to convert a PDF to various formats based on what I
want out of it, then recreate it in Scribus or OpenOffice if I want a
new PDF. I often use this script to pull out vector data:
<snip>
#!/bin/bash
# you need gs-common, pstoedit and skencil to
# get this script working
export BASENAME=$1
#convert to ps
pdf2ps ${BASENAME} ${BASENAME}.ps
# Outline fonts
eps2eps -dNOCACHE ${BASENAME}.ps ${BASENAME}-TEMP
# Fix bounding box
ps2epsi ${BASENAME}-TEMP ${BASENAME}.ps
rm ${BASENAME}-TEMP
# convert to svg
pstoedit -f plot-svg ${BASENAME}.ps ${BASENAME}.svg
rm ${BASENAME}.ps
<snip>
Or I might rip the PDF in GIMP to extract the images. Sometimes I
open it in Reader and copy/paste the text. Printing to file from
Reader or running pdf2ps will sometimes yield a postscript document
than can be imported into Inkscape - but that hasn't worked very
reliably in my experience.
In the end, you *can* edit a PDF in linux, it's just often very, very
painful - and I don't think that the pain is caused by anything
"closed source" - the PDF spec has been available for years.
Chris
I just did a quick test with KWord and it opens pdfs for editing. I
don't know how good it is. I understand OOo 3.0 will have pdf editing
abilities.
--
Robin Laing
