On Friday 11 February 2011 7:42:53 am Guillaume Nodet wrote: > On Fri, Feb 11, 2011 at 13:15, Benson Margulies <[email protected]> wrote: > > Currently, the deployed site is a 'forrest' antique. > > > > We have some pretty reasonable content sitting in Confluence. As you > > may know, in general, intrastructure@apache is pushing projects from > > Confluence to their new Markdown+svn CMS. The reason is their problems > > with the generic export technology used historically with Confluence. > > > > Some of us are not really thrilled at editing Markdown in a text > > editor or web browser. > > > > Dan Kulp has gone and created an alternative Confluence export device. > > It meets the strictures of infrastructure@ and it allows site > > maintenance in Confluence. > > I wasn't aware of this work. Is there any info available on that ?
Well, I wasn't actually planning to make it public yet as I don't have the time right now to have battles with Joe..... :-) Seriously, for 2 weeks now, CXF has not used the AutoExport plugin of confluence. The fact that no-one has noticed any functional difference is a good thing. It means it's working fairly well. Basically, in my cron, instead of using rsync to copy the exported content, I have a little java program that uses the Confluence SOAP API (the confluence REST api doesn't provide enough functionality, I tried) to grab the content and then run velocity locally to generate the site. In the process, it runs the HTML through tagsoup and does a bunch of attribute filtering and such to fix the HTML (and all the links and such as well). Thus, the resulting HTML is much better than the auto-export version. The CXF home page, for example, is HTML4/Transitional compliant now whereas it had 40+ errors (using the w3c validator) with the autoexport version. It does cache the mod times (file on disk) of the pages and such so only changed pages get regenerated. However, it's also smart enough to detect when one of the key pages (like Navigation and Banner) get updated to force a full rebuild. (autoexport doesn't do that) It also checks the RSS feed before any of the SOAP calls just to save time/resources. If no RSS changes, it exits fast. Another note: it doesn't do this yet, but it can easily detect page deletions and delete the files as well. I just need to add that. autoexport cannot do that at all right now. Right now, it just writes the files out to a directory. At some point, I want to update it to support subversion so it can auto checkin the files into subversion. That would allow svnpubsub usage for publishing the site live. I could then also move it from a cron to a build in buildbot or hudson. That would allow "instant" site updated by just launching a build as needed. If anyone wants to play with it: http://svn.apache.org/repos/asf/cxf/web/ Dan > > > I'd like to make ws.apache.org use Confluence (with some maven site > > output as appropriate), using Dan's technology. Anybody object? > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [email protected] > > For additional commands, e-mail: [email protected] -- Daniel Kulp [email protected] http://dankulp.com/blog --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
