2009/1/20 Platonides platonides@gmail.com:
Andrew Dunbar wrote:
2009/1/20 Platonides platonides@gmail.com:
Andrew Dunbar wrote:
Just the couple I needed for checking for and fetching raw articles: /w/index.php?title=...&action=raw /w/api.php?format=json&action=query&titles=...|...&redirects
And what's the point of Firefox retrieving article contents when you're offline?
Err well being able to read articles when you're offline.
But also analysing Wiktionary. Checking whether it had entries for certain words etc.
Then why the API? I'd prefer a graphical interface, not just wikitext into XML. :)
Emulating the API made it trivial to have several classes which worked in exactly the same way. I could look up the real Wiktionary online, a local MySQL db, or a an offline dump without changing the client code.
I could also work on developing Wiktionary tools without paying for internet all day when I was travelling.
Graphical interfaces for just reading wikis already existed. But I'm a programmer who wanted to do more than just read articles.
Do you place all articles on the same page using AJAX?
I used it for a couple of personal web apps. I could collect strange new words when reading and enter them as a list which gave me red or blue links for each term so I could see whether they needed a Wiktionary entry or not. I also had a second back end which looked up a reverse engineered Spanish dictionary CD rom with the same API. I was planning to extend it then add all those redlinks to Wiktionary's requested entries page next time I was online. The other web app subscribed to a bunch of word-of-the-day RSS feeds and checked all the words for presence in Wiktionary.
plus an extension which returned a list of languages for which there were articles in an English Wiktionary page
For outside Firefox I also had a Perl CGI version but on my Eee PC that was way more expensive to run
Probably because the perl is interpreted and the extension compiled.
Well Firefox extensions are in JavaScript, also an interpreted language, but the cost of starting up the perl interpreter was expensive, especially doing a few simultaneous operations.
I thought it was a binary one. Firefox extensions can be in javascript or XPCOM. Still, it caches the opcodes and the trunk version is even faster.
Well XPCOM extensions are platform specific and I'm a big fan of cross-platform tools. I could've built a Windows version but no *nix or OS X version on my little Eee PC. And the js version was faster at looking up offline Wiktionary then looking up the online Wiktionary with a slow connection.
The use case I would deem more common would be to access page contents from outside for processing, eg. to show it on a page, as http://wiki-web.es/mediawiki-offline-reader/
Perhaps. I wasn't making something for the more common use case, I was making something I had a use for (-:
Well, showing mediawiki pages is exactly what that app does. It's just that I wouldn't use the dumps into offline Firefox to feed the API. Seems you have different uses (or perhaps I have completed misunderstood it :)
(-: Andrew Dunbar (hippietrail)
Mediawiki-api mailing list Mediawiki-api@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/mediawiki-api