aboutsummaryrefslogtreecommitdiff
path: root/doc/patchqueue/more_class__61____34____34___for_css.mdwn
blob: 49affd29bfd40315b85d814c0d74bee33e0cc35d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
I'm writing my own CSS for ikiwiki. During this effort I often found the need of adding more class="" attributes to the default ikiwiki templates. This way more presentational aspects of visual formatting can be delegated to CSS and removed from the HTML structure.

In this patch I plan to collect changes in this direction.

The first, one-liner, patch is to use a "div" element with a
class="actions" attribute for inline page as is done with non-inlined page.
This way the same CSS formatting can be applied to div.actions in the CSS,
while at the moment it must be duplicated for a span.actions (which I
believe is also incorrect, since it will contain a "ul" element, not sure
though). In case the markup should be differentiated it will still be
possible relying on the fact that a div.actions is contained or not in a
div.inlinepage.

Here's the one-liner:

> applied --[[Joey]]

The following adds a div element with class="trailer" around the meta-information
added after an inlined page (namely: the post date, the tags, and the actions):

    --- inlinepage.tmpl.orig        2006-12-28 16:56:49.000000000 +0100
    +++ inlinepage.tmpl     2006-12-28 17:02:06.000000000 +0100
    @@ -17,6 +17,8 @@
     </span>
     <TMPL_VAR CONTENT>

    +<div class="trailer">
    +
     <span class="pageinfo">
     Posted <TMPL_VAR CTIME>
     </span>
    @@ -44,3 +46,5 @@
     </TMPL_IF>

     </div>
    +
    +</div>

> Unfortunately, the inlinepage content passes through markdown, and markdown
> gets confused by these nested div's and puts p's around one of them, generating
> broken html. If you can come up with a way to put in the div that passes
> the test suite, or a fix to markdown, I will accept it, but the above patch
> fails the test suite. --[[Joey]] 

>> Just a note...  This discrepancy doesn't exist in [pandoc](http://code.google.com/p/pandoc/) as
>> demonstrated in the relevant [page](http://code.google.com/p/pandoc/wiki/PandocVsMarkdownPl).
>> Pandoc is a _real parser_ for markdown (contrasting the regexp based implementation of 
>> markdown.pl).  I've almost finished the Debian packaging.  John is working on a `--strict` mode
>> which will hopefully make pandoc a drop-in replacement for markdown.  I'll upload pandoc after 
>> his work has finished.  Whether it could be used in IkiWiki is an open question, but having
>> alternatives is always a good thing and perhaps, the fact that pandoc can make markdown->LaTeX
>> conversion may lead to new possibilities. --[[Roktas]]

>>> I confirm that this ([[debbug 405058]]) has just been fixed in markdown
>>> [`1.0.2b7`](http://packages.debian.org/experimental/web/markdown) (BTW, thanks to your bug
>>> report Joey).  FYI, I've observed some performance drop with `1.0.2b7` compared to `1.0.1`,
>>> especially noticable with big files.  This was also confirmed by someone else, for example,
>>> see this [thread](http://six.pairlist.net/pipermail/markdown-discuss/2006-August/000152.html)
>>> --[[Roktas]]