Jean Hollis Weber wrote:
Yes, I think it would be useful.
--Jean
PicWiki work flow
Here is the proposed work flow for the new code (not written yet) to
simplify picture handling during wiki conversion. This assumes that one
chapter of a Guide is to be converted.
1) User saves a copy of the .odt file in the directory where the
pictures are to be stored.
2) User runs the (new) PicWiki code on this copy. The code saves the
pictures, assigns them coded names, and writes some trash into the .odt
file. (Good thing it’s a copy, right?) This file can be saved if you want.
3) User exports this file using the MediaWiki Export Filter (MWEF). The
new wiki-text file can be saved anywhere convenient.
4) User loads this wiki-text file, and runs the Wiki Cleanup (formerly,
TNC Cleanup) code on it. This code cleans up the Tip / Note / Caution
references, formats the picture references (hopefully, exactly as
needed), and cleans up any other trash it can find.
5) The real user work starts here:
a) check the file for any other trash
b) break file into pages
c) upload pictures and pages
d) create TOC stuff
e) and probably more . . .
To help with uploading the pictures, I can generate a page with table
lines like:
[copy] local file URL | [copy] wiki URL | [copy] description (caption)
where [copy] is a button, click to copy the associated text to the
clipboard.
If the picture doesn't have a caption, I can show the picture itself, so
you can make up the description. The work flow would be: click on a
[copy] button, click on the browser field, Ctrl+V, click on the next
[copy] button ...
Something like this will probably be necessary, because the generated
picture names will likely be ugly to type. Either that, or the wiki
directory name will be ugly, and it’s in every URL for every picture.
Actually, I favor the ugly directory name, so that the URL might look
like .../0104WG3/P2, in the loading and in the actual image references.
All negotiable.
In the longer term, it should be possible to set up a script, or a bot,
to accept a file as input, and do the picture uploading automagically.
That’s beyond me, for now.
Comments and suggestions welcome. I still have a lot of research to do,
before I start coding. This is a really good time for, “It would be nice
if ...”, or “I wish I could ...”
--
/tj/
