On 2005-08-23 01:01:09, Daniel Kasak wrote:
Riccardo Bertonati wrote:
Hello,
I'd like you to add a couple of things to the suite, in particular:
(SNIP)
2 - the possibility to edit PDF files. You could start with minimal
editing functions like underlining text, changing text background
colour and adding freehand drawing capabilities. Of course it must
be possible to save the modified pdf into pdf format once again.
This one gets discussed regularly. PDF is not a document type that
you can edit. Think of it as an image type. When you save something
to PDF, everything gets converted to printer-friendly instructions on
how to *print* the document. You can't take those instructions and
turn them back into a highly-structured format. It's like cracking an
egg, frying it, and asking someone to reconstruct the egg for you.
The best you can do is treat the PDF file like an image file, and do
Optical Character Recognition on it, and this is also far from
perfect.
(SNIP)
Actually this turns out to be inaccurate. I was very surprised
recently to learn that kword CAN open and edit pdf files to a great
extent. [koffice-1.4 - I'm using kword 1.4.1
(http://www.koffice.org/kword) and the kde desktop
(kdebase-3.4.2-0.fc3.2) on linux (fedora fc3)]
While the text conversion is usually very good and a lot of formatting
is actually preserved (pagination, font faces, sizes, weights, etc.),
ad while images are so far generally preserved (not sure of the format
they are saved in), there are occasional minor glitches... but so far
nothing really serious has shown up after about 50 pdf docs brought up
in kword. For instance while page headers and footers are preserved,
they show up as text, not as a footer or header - and while they are in
the correct places, it is usually better to edit the doc and turn them
back into headers/footer items as appropriate. Similarly footnotes are
there, but not as OOo footnotes - again you have to convert back. And
while it might be possible to do this with a macro, I haven't tried yet
so I don't know.
Tables are a little strange - they show up as tab-delimited text, so
the conversion back to table format can be easy, but a complex table
formtting is lost and must be re-added.
Actually the biggest problem appears with the strategy (PDF->kword-
OOo) to be that while kword can save the docs on oasis format file
(".odt" file which can be used by writer in OOo-1.9.122), the kde folks
have a slightly different interpretation of that format from that used
by OOo. (BTW, kword can also save as an OOo-1.1 [".sxw"] format file,
but I have yet to create one which was openable by OOo or star
office...)
The only other "gotcha" here is that kword uses a *lot* of frames in
the conversion and sometimes it is easier (for me, anyway) to move
everything inside the frame back to the body of the doc and blow the
frame away. (maybe my fault: I tend not to use frames that much and am
not fully comfortable fixing them.)
So there are between "some" and "a lot" of formatting issues to fix
when the kword doc is saved as an ".odt" file and brought up in OOo
writer.
Overall however, this is light-years better than the scanning solution
and is about 90% effective for small to medium-sized docs (the largest
I have done this with was a 104-page tech report with about 20 figures,
~ 25 tables, and 15(+/-) images. I didn't have to retype anything in
most of them (thank goodness). I will say that this is also the only
use I have for the koffice suite (so far, anyway) and that otherwise
OOo is far superior technically.
--
william w. austin [EMAIL PROTECTED]
"life is just another phase i'm going through. this time, anyway ..."
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
