Support for downloading dynamic images through the webserver

modified version to download all images from the webserver

Hi Nickolas,

Please, I need a patch file with your code. Which trac version are you using?

Regards

Diorgenes F. Grzesiuk

Hi Diorgenes

I'll re-do the code for that tomorrow, I was using Trac 0.11. I'll see if I can do the code a bit cleaner. (and optional)

I also have some other changes where I work around some HTMLDoc limitations, and if using HTMLDoc 1.9 (unstable) it even does code highlighting.

Patch to add Downloading Images & enable Code highlighting wht using HTMLDoc 1.9

Sample input

Sample Output - HTMLDoc 1.9

Hi Diorgenes

Added trac.ini options:

[wikitopdf] tmp_dir = /tmp/wikitopdf trac_uri = http://trac.blah.is.co.za css_file = /var/lib/trac/IS/conf/htmldoc.css

both trac_uri and tmp_dir has to be defined for the dynamic image downloading to work. (Otherwise it uses the existing method, to not break upgrades) the css_file points to a css file (included in patch) for HTMLDoc 1.9, so that code highlighting can work.

The patch also adds tables around div frames, so that the output looks closer to the wiki pages.

I know that there is a lot of changes, but you can take your time evaluating them.

Thanks, Nickolas

[wikitopdf]
tmp_dir = /tmp/wikitopdf
trac_uri = http://trac.blah.is.co.za
css_file = /var/lib/trac/IS/conf/htmldoc.css

Replying to grigi:

{{{ [wikitopdf] tmp_dir = /tmp/wikitopdf trac_uri = http://trac.blah.is.co.za css_file = /var/lib/trac/IS/conf/htmldoc.css }}}

Hi

I need more time for study your contribution. Just for now thank you.

Regards

Diorgenes F. Grzesiuk

Replying to nagrigoriadis@gmail.com:

Since a significant number of images in our wiki is dynamicly generated images, such as emoticons and latex-math or similar, and there is no way to configure the current wikitopdf plugon to do it, I did a quick hack to wikitopdf.py to download the images to a temporary location and then use those.

Hi everyone,

i had similar problems. In wikitopdf.py (45-50) the base_dir is set as prefix for image source.

45 	    imgpos = page.find('<img')
46 	
47 	    while imgpos != -1:
48 	        addrpos = page.find('src=',imgpos)
49 	        page = page[:addrpos+5] + base_dir + page[addrpos+5:]
50 	        imgpos = page.find('<img', addrpos)

But the image source is like https://localhost:8443/trac/sandbox/attachment/wiki/WikiStart/image.jpg. After adding base_dir, which points to the trac project, i had base_dirhttps://localhost:8443/trac/sandbox/attachment/wiki/WikiStart/071small.jpg which isn't the right image source. So i substitute the base trac uri by base_dir

abs_ref = (req or env).abs_href.base
href = (req or env).href.base

As result now i get !base_dir/attachment/wiki/WikiStart/image.jpg.

By the way i also added two options for supporting TracMathPlugin and GraphvizPlugin without downloading via wget. The options are

[wikitopdf]
graphviz_dir = 
tracmath_dir =

These are the direcories where the both plugins save there images. A patch file is added as attachment.

Best regards,
Andreas Gebhardt

support for TracMathPlugin and Graphviz Plugin, base_dir substituion

Update of patch for WikiToPdf? 2.2.2

diorgenes, Do you need me to split the patch into several small parts? I would like to get as much of this into the mainline version.

Patch to fix downloading images that have params in their name like "?width=" for example like the ones generated by google charts plugin.

Just a note: htmldoc version htmldoc-1.9.x-r1586 has broken headers/footers, version htmldoc-1.9.x-r1563 works fine.

You only need the unstable 1.9 version if you want syntax highlighting and/or css support.

This ticket's patches has been added to the svn repos. Leaving ticket open until updating documentation.