Giuseppe Bilotta [Mon, 21 Oct 2019 15:18:35 +0000 (17:18 +0200)]
img: add support for objects too
The primary intended use is for interactive SVGs and SVGs that depend on
external data, which would otherwise not render correctly in most modern
browsers because security.
Giuseppe Bilotta [Thu, 29 Aug 2019 19:59:38 +0000 (21:59 +0200)]
Improved OpenGraph title
Giuseppe Bilotta [Thu, 29 Aug 2019 19:36:05 +0000 (21:36 +0200)]
Preliminary OpenGraph support
Joey Hess [Sat, 16 Feb 2013 16:20:27 +0000 (12:20 -0400)]
memoize check for up in trail (untested)
Giuseppe Bilotta [Fri, 15 Feb 2013 21:43:50 +0000 (22:43 +0100)]
trail: no 'up' page if it's a member of the trail
When a trail includes its index page among the members, the pages
before/after it will have “double links” to the index page (both as
next/prev and as 'up') and the index page will include an 'up' link to
itself.
Prevent this by not defining the appropriate template variables when the
index page is member of the trail itself. The default page template is
also adapted to skip the link at all in this case.
Giuseppe Bilotta [Sun, 26 Aug 2012 05:35:02 +0000 (07:35 +0200)]
inline: fix spurious feed removal from nested inlining
Situation:
* page A has an inline directive that (directly) generates a feed F
* page B inlines A, thus (indirectly) generating F again
* page B is rendered after page A
Without this patch, F is rendered by both A and B. However, if a change
in B causes it to _not_ inline A anymore, the following would happen:
* page A is built, rendering F
* page B is built, _not_ rendering F (which it used to render)
* F is removed because it's not rendered by B anymore
With this patch, rendering of F is always controlled by A, preventing
the spurious feed removals from happening.
A wiki rebuild is necessary after this patch is included.
Note that although this issue is triggered by last year's changes
bb8f76a4a04686def8cc6f21bcca80cb2cc3b2c9 and
72c8f01b36c841b0e83a2ad7ad1365b9116075c5 (that allowed feed generation
from nested inlines, coalescing it to be page-based instead of
destpage-based), there is potential for it popping up in other cases.
Specifically, the generated pages removal logic curently relies on the
assumption that a given output has only one generator. This assumption
was violated for feeds with the introduction of the mentioned commits,
and is now being re-instated (for feeds) with this patch, but other
similar cases may pop up for other plugins in the future.
Giuseppe Bilotta [Thu, 10 May 2012 05:52:49 +0000 (07:52 +0200)]
linkbase plugin
The linkbase plugin defines a directive to manage linkbases in the
document. These are typically defined in the linkbases file at the
document root, much like shortcuts.
Giuseppe Bilotta [Wed, 9 May 2012 07:08:16 +0000 (09:08 +0200)]
linkbase feature
The linkbase feature allows pages to specify one or more (other)
locations in which link targets should be searched for. This may be
defined either for pagespecs (all pages matching the spec will have
that linkbase) or for specific pagenames.
Giuseppe Bilotta [Thu, 19 Apr 2012 21:24:14 +0000 (23:24 +0200)]
IkiWiki: option to not inline images automatically
Giuseppe Bilotta [Mon, 20 Jun 2011 06:48:04 +0000 (08:48 +0200)]
inline: define a pageid
Giuseppe Bilotta [Thu, 5 Apr 2012 18:44:26 +0000 (20:44 +0200)]
inline: make baseurl available to the inlinepage template
Giuseppe Bilotta [Wed, 25 Apr 2012 18:13:28 +0000 (20:13 +0200)]
inline: propagate inline mtime/ctime too
If page A inlines page B that inlines page C, then we want the inline
mtime of page A to match that of C if it's more recent than the mtime of
B.
Giuseppe Bilotta [Wed, 21 Dec 2011 16:29:49 +0000 (17:29 +0100)]
inline: initialize mtime field with ctime
This ensures proper ordering when the mtime field is absent or bogus.
Giuseppe Bilotta [Fri, 26 Aug 2011 16:10:16 +0000 (18:10 +0200)]
conditional: also trigger preprocessing during scan
This is necessary to correctly support directives that have a preprocess
hook with scan=>1 and which are nested inside an if directive.
While we're at it, we also properly propagate the preview state.
Giuseppe Bilotta [Sun, 10 Apr 2011 17:35:52 +0000 (19:35 +0200)]
meta: ignore_mtime option
This option makes 'updated' fall back to the ctime instead of the mtime.
Giuseppe Bilotta [Fri, 28 Jan 2011 16:42:39 +0000 (17:42 +0100)]
inline: add a couple of TODOs
Giuseppe Bilotta [Fri, 28 Jan 2011 16:36:20 +0000 (17:36 +0100)]
inline: cope with addition/removal of file
Since the inline times must be updated during the scan of the inlining
(rather than of the inlined) document(s), we must find the inlining
pages before the scan phase (specifically, during the needsbuild
phase) and mark them as changed too.
Additionally, we must also consider if there are pages being removed,
and in the affirmative case we must know which ones they are so that
they be ignored during the scan phase.
Giuseppe Bilotta [Tue, 25 Jan 2011 18:34:18 +0000 (19:34 +0100)]
inline: ctime and mtime sorting
Two new sort methods, inline([cm]time), are defined to allow sorting
pages by the mtime/ctime of their inlined pages rather than by their
own.
Giuseppe Bilotta [Tue, 25 Jan 2011 18:31:23 +0000 (19:31 +0100)]
post-scan hooks
During the scan phase plugins can push anonymous functions into the
@post_scan array, to be called after all pages have been scanned.
This allows plugin to run actions between the scanning and building
phase, which is important for pre-render actions that e.g. need to
ensure that pagespec patterns will match correctly.
Giuseppe Bilotta [Sat, 1 Jan 2011 21:39:36 +0000 (22:39 +0100)]
inline: improve feed title and description management
Move feed title generation outside of genfeed(), and allow it to be
customized by an appropriate parameter to the inline directive.
The feed description is also extended to base it on the page description
(if present), and defaulting to a configurable union of wiki name and
page title otherwise (rather than hard-coding it to the wiki name).
Giuseppe Bilotta [Mon, 28 Feb 2011 16:22:11 +0000 (17:22 +0100)]
Specify that the URL should not have a trailing /
Giuseppe Bilotta [Tue, 4 Jan 2011 12:19:35 +0000 (13:19 +0100)]
google: add google analytics
The code is based primarily on the flattr plugin.
Giuseppe Bilotta [Tue, 4 Jan 2011 12:11:37 +0000 (13:11 +0100)]
google: add config to enable/disable sitesearch
In preparation for the inclusion of other Google services, make the
Google sitesearch inclusion configurable, to allow loading the google
plugin to use the other services without having to edit the default
templates to remove the search form.
A similar effect can be achieved by creating an empty
templates/googleform.tmpl, but we don't want to force people to hack
around automatic stuff this way.
dave@2ab82f485adf7e2ce787066e35f5f9789bff430b [Thu, 10 Oct 2019 18:40:58 +0000 (14:40 -0400)]
I had the same problem somebody on the discussion page had so I changed the instructions so nobody else will.
jmtd [Thu, 10 Oct 2019 15:40:08 +0000 (11:40 -0400)]
libtext-csv-xs-perl not relevant
jmtd [Thu, 10 Oct 2019 15:36:29 +0000 (11:36 -0400)]
sigh, still broken, different error message.
jmtd [Mon, 7 Oct 2019 18:15:10 +0000 (14:15 -0400)]
libpel5
jmtd [Mon, 7 Oct 2019 15:39:05 +0000 (11:39 -0400)]
more info: libperl5, buster versus stretch
jmtd [Mon, 7 Oct 2019 13:57:00 +0000 (09:57 -0400)]
follow-up: sorry, I appreciate that wasn't as useful as it could have been
anarcat [Wed, 2 Oct 2019 14:27:16 +0000 (10:27 -0400)]
split done and not done patch list, remove patches from the other lists
anarcat [Wed, 2 Oct 2019 14:25:39 +0000 (10:25 -0400)]
explicitely tag this as a patch
intrigeri [Mon, 30 Sep 2019 14:59:33 +0000 (10:59 -0400)]
Gentle ping on proposed branch
Simon McVittie [Sun, 29 Sep 2019 17:42:40 +0000 (18:42 +0100)]
cannot reproduce new bug
Simon McVittie [Sun, 29 Sep 2019 16:48:25 +0000 (17:48 +0100)]
Update changelog with schmonz's changes
poobah@5486f0334d386be645fa4deeed4bf2d8593152c5 [Sat, 28 Sep 2019 17:52:05 +0000 (13:52 -0400)]
jmtd [Mon, 23 Sep 2019 10:06:59 +0000 (06:06 -0400)]
bug remains for inline table case
Joey Hess [Fri, 20 Sep 2019 16:10:18 +0000 (12:10 -0400)]
bug report
Amitai Schleier [Tue, 10 Sep 2019 13:59:26 +0000 (09:59 -0400)]
Fix build with po4a 0.56 via consistent "\n" use.
Errors look like so:
PERL5LIB=.. ./po2wiki underlay.setup
Invalid po file
/var/folders/y5/k8bh4xsj75g4j1n631f2gqz80000gp/T/ikiwiki-po-filter-in.OPXJxZ0juW:
/var/folders/y5/k8bh4xsj75g4j1n631f2gqz80000gp/T/ikiwiki-po-filter-in.OPXJxZ0juW:293:
'msgid' and 'msgstr' entries do not both end with '\n'
/var/folders/y5/k8bh4xsj75g4j1n631f2gqz80000gp/T/ikiwiki-po-filter-in.OPXJxZ0juW:589:
'msgid' and 'msgstr' entries do not both end with '\n'
msgfmt: found 2 fatal errors
make[1]: *** [po2wiki_stamp] Error 255
anarcat [Fri, 6 Sep 2019 14:16:05 +0000 (10:16 -0400)]
not sure this is necessary
add link to auto-numbering on CSS
Amitai Schleier [Wed, 4 Sep 2019 23:07:46 +0000 (19:07 -0400)]
MultiMarkdown footnote un-inversion: done
Amitai Schleier [Wed, 4 Sep 2019 23:01:28 +0000 (19:01 -0400)]
Fix inverted footnote config with MultiMarkdown.
Bug spotted and fix from Giuseppe Bilotta <giuseppe.bilotta@gmail.com>.
Extend mdwn tests to cover MultiMarkdown, where applicable.
Amitai Schleier [Wed, 4 Sep 2019 18:38:43 +0000 (14:38 -0400)]
Please test sudo-mass-rebuild branch on Debian
xloem [Tue, 3 Sep 2019 14:04:47 +0000 (10:04 -0400)]
Amitai Schleier [Sun, 1 Sep 2019 14:39:48 +0000 (10:39 -0400)]
Response: am trying sudo
rename bugs/MultiMarkdown_footnot_support_seems_to_be_broken.mdwn to bugs/MultiMarkdown_footnote_support_seems_to_be_broken.mdwn
Directive fix
jmtd [Sun, 25 Aug 2019 21:04:57 +0000 (17:04 -0400)]
I've not looked at Shake yet, thanks, I'll give it a look! — [[Jon]]
simon@4e26c0bdeb57cbb3cc79e234379178f37fa71d06 [Thu, 22 Aug 2019 11:39:54 +0000 (07:39 -0400)]
simon@4e26c0bdeb57cbb3cc79e234379178f37fa71d06 [Thu, 22 Aug 2019 11:36:03 +0000 (07:36 -0400)]
Suggest shake
jmtd [Wed, 21 Aug 2019 15:38:08 +0000 (11:38 -0400)]
some beginning code to migrate an ikiwiki to hakyll
jmtd [Fri, 16 Aug 2019 13:21:33 +0000 (09:21 -0400)]
lol
jmtd [Fri, 16 Aug 2019 13:18:35 +0000 (09:18 -0400)]
saw typesetter-css and was reminded of my appeal against the anti-theme as default
lkcl@d317c3c240102454f3d5ffa9210ae2178eb5af3a [Tue, 25 Jun 2019 15:48:24 +0000 (11:48 -0400)]
https://launchpad.net/~skellat [Mon, 17 Jun 2019 01:32:02 +0000 (21:32 -0400)]
Adding my work-in-progress blog to the list
Amitai Schleier [Fri, 7 Jun 2019 21:12:46 +0000 (17:12 -0400)]
No longer hosting these on Branchable.
Joey Hess [Thu, 6 Jun 2019 21:18:22 +0000 (17:18 -0400)]
Merge branch 'master' of ssh://git.ikiwiki.info
Joey Hess [Thu, 6 Jun 2019 21:17:58 +0000 (17:17 -0400)]
Updated German program translation from Sebastian Kuhnert.
Amitai Schleier [Wed, 5 Jun 2019 17:52:41 +0000 (13:52 -0400)]
Fix link in previous.
Amitai Schleier [Wed, 5 Jun 2019 17:50:22 +0000 (13:50 -0400)]
Reply to fancypodcast question.
Joey Hess [Wed, 5 Jun 2019 16:09:08 +0000 (12:09 -0400)]
Updated German basewiki and directives translation from Sebastian Kuhnert.
https://launchpad.net/~skellat [Sun, 2 Jun 2019 04:56:41 +0000 (00:56 -0400)]
Open up the discussion
Amitai Schleier [Mon, 27 May 2019 03:45:10 +0000 (23:45 -0400)]
Catch up to highlight 3.51 API change.
As of 3.51, searchFile() is no longer provided in highlight's Perl
bindings (at least on NetBSD and OS X, as built from pkgsrc). This
leaves us falling through to getConfDir(), which has been gone
rather longer.
From highlight git, it appears searchFile() and getFiletypesConfPath()
both originated in the 3.14 release. The latter is still available in
3.51, and returns the same result searchFile() used to. Switch to it.
canvon [Tue, 21 May 2019 13:19:34 +0000 (09:19 -0400)]
Add my ikiwiki/blog
jsewell@560d759be38e126eba7f1c40503adf786fa28229 [Fri, 10 May 2019 17:26:11 +0000 (13:26 -0400)]
Amitai Schleier [Mon, 6 May 2019 11:43:17 +0000 (07:43 -0400)]
Response to jsewell's bug report.
jsewell@560d759be38e126eba7f1c40503adf786fa28229 [Mon, 29 Apr 2019 14:19:12 +0000 (10:19 -0400)]
jmtd [Wed, 24 Apr 2019 19:19:32 +0000 (15:19 -0400)]
add gitbranch template, prompted by Integeri. Thanks
intrigeri [Wed, 10 Apr 2019 21:02:16 +0000 (17:02 -0400)]
Fix syntax.
intrigeri [Wed, 10 Apr 2019 21:01:07 +0000 (17:01 -0400)]
Submit branch for review.
jmtd [Sat, 30 Mar 2019 21:34:11 +0000 (17:34 -0400)]
update my comment to reflect new commits on that branch
jmtd [Fri, 29 Mar 2019 11:17:45 +0000 (07:17 -0400)]
initial implementation (I was unable to use untrusted git push to add this comment)
mike@8d1a742254a41aaff8dd19404183dce96fac24ba [Thu, 7 Mar 2019 00:05:56 +0000 (20:05 -0400)]
Simon McVittie [Thu, 28 Feb 2019 18:03:18 +0000 (18:03 +0000)]
Merge remote-tracking branch 'origin/master'
Simon McVittie [Thu, 28 Feb 2019 14:14:12 +0000 (14:14 +0000)]
Announce 3.
20190228 and 3.
20170111.1
Simon McVittie [Thu, 28 Feb 2019 14:11:20 +0000 (14:11 +0000)]
Add an anchor for /security/#cve-2019-9187
Simon McVittie [Tue, 26 Feb 2019 21:06:45 +0000 (21:06 +0000)]
Prepare 3.
20190228 for future release
Simon McVittie [Sun, 10 Feb 2019 16:56:41 +0000 (16:56 +0000)]
doc: Document security issues involving LWP::UserAgent
Recommend the LWPx::ParanoidAgent module where appropriate.
It is particularly important for openid, since unauthenticated users
can control which URLs that plugin will contact. Conversely, it is
non-critical for blogspam, since the URL to be contacted is under
the wiki administrator's control.
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Sun, 10 Feb 2019 17:22:06 +0000 (17:22 +0000)]
useragent: Automatically choose whether to use LWPx::ParanoidAgent
The simple implementation of this, which I'd prefer to use, would be:
if we can import LWPx::ParanoidAgent, use it; otherwise, use
LWP::UserAgent.
However, aggregate has historically worked with proxies, and
LWPx::ParanoidAgent quite reasonably refuses to work with proxies
(because it can't know whether those proxies are going to do the same
filtering that LWPx::ParanoidAgent would).
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Sun, 10 Feb 2019 16:30:07 +0000 (16:30 +0000)]
useragent: Don't allow non-HTTP protocols to be used
This prevents the aggregate plugin from being used to read the contents
of local files via file:/// URLs.
Signed-off-by: Simon McVittie <smcv@debian.org>
anarcat [Tue, 26 Feb 2019 05:35:32 +0000 (01:35 -0400)]
machine_brain [Tue, 26 Feb 2019 00:01:58 +0000 (20:01 -0400)]
Simon McVittie [Sun, 10 Feb 2019 16:29:19 +0000 (16:29 +0000)]
useragent: Raise an exception if the LWP module can't be loaded
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Sat, 1 Dec 2018 21:18:23 +0000 (21:18 +0000)]
po: Always filter .po files
The input to filter hooks is meant to be the content of a source file
on disk. If we only filter once per (page, destpage) pair, and a page
is inlined into the same destpage more than once, then the second
occurrence will render as the result of htmlizing .po source as if
it was Markdown (or whatever the type of the corresponding master page
is), which is never going to end well.
The alreadyfiltered mechanism was added in commit
1e874b3f to avoid
preprocessing loops, but I'm not sure where it could lead to a loop:
filter hooks are only called from IkiWiki::filter, which is only called
on page content from disk or on proposed content being previewed.
According to <https://bugs.debian.org/911356#41>, deleting the
alreadyfiltered mechanism resolves the problem, as well as simplifying
the code.
Closes: #911356
Tested-by: intrigeri
Amitai Schleier [Thu, 14 Feb 2019 04:59:32 +0000 (23:59 -0500)]
Recommend against cvsps3 (haven't tried it).
Simon McVittie [Thu, 7 Feb 2019 11:32:21 +0000 (11:32 +0000)]
Announce v3.
20190207
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Thu, 7 Feb 2019 11:08:02 +0000 (11:08 +0000)]
Prepare new release
Signed-off-by: Simon McVittie <smcv@debian.org>
Simon McVittie [Sun, 3 Feb 2019 19:22:07 +0000 (19:22 +0000)]
review
Simon McVittie [Sun, 3 Feb 2019 18:53:23 +0000 (18:53 +0000)]
comment
smcv [Sun, 3 Feb 2019 18:40:29 +0000 (14:40 -0400)]
tag as reviewed
smcv [Sun, 3 Feb 2019 18:39:51 +0000 (14:39 -0400)]
respond
smcv [Sun, 3 Feb 2019 18:28:21 +0000 (14:28 -0400)]
Exclude reviewed patches from this list
Simon McVittie [Sun, 3 Feb 2019 17:10:45 +0000 (17:10 +0000)]
close
Antoine Beaupré [Mon, 6 Nov 2017 19:15:22 +0000 (14:15 -0500)]
append javascript after CSS
Javascript resources should be presented to browsers after CSS, and
"after the fold" (ATF) according to the best practices:
https://developers.google.com/speed/docs/insights/mobile#PutStylesBeforeScripts
This change allows the browser to download Javascript files in
parallel, by including Javascript on the *closing* </body> tag instead
of the opening tag.
We also improve the regex to tolerate spaces before the body tag, as
some templates have (proper) indentation for the tag.
Simon McVittie [Sun, 2 Sep 2018 11:49:16 +0000 (12:49 +0100)]
Remove unreachable git repositories
Signed-off-by: Simon McVittie <smcv@debian.org>