jeluf wrote:

The extension needs to be reviewed first. From a first look at it, I'd suggest

• A primary key is always unique. There's no need for the UNIQUE KEY.
• Covering indexes should be used to speed up select queries.
• The preg_replace list in DYMNorm.php looks very expensive. Wouldn't it be faster to use the array form of preg_replace for this, or even Tim's custom replace routines.

Changing keywords from "shell" to "need-review"

treauth wrote:

*** Bug 12329 has been marked as a duplicate of this bug. ***

wikt.3.connelm wrote:

(In reply to comment #1)

The extension needs to be reviewed first. From a first look at it, I'd suggest

• A primary key is always unique. There's no need for the UNIQUE KEY.
• Covering indexes should be used to speed up select queries.
• The preg_replace list in DYMNorm.php looks very expensive. Wouldn't it be faster to use the array form of preg_replace for this, or even Tim's custom replace routines.

Changing keywords from "shell" to "need-review"

I'm sorry, but that is fundamentally wrong. In honor of the 100th (or is it 1 millionth? Who counts?) request, I'll try re-implementing this in Javascript. It will be painfully slow, and horrifically brutal on the WMF servers, but maybe, just maybe, the lookups people do will direct them to the entries that actually want.

Removing "need-review" flag since Rob Church, Tim Starling, and Brion Vibber all reviewed the code before it was committed to svn. It also ran on Brion's own public test wiki for close to a year with no known bug reports.

It was however my first MediaWiki extension and my first work using SQL so it can surely be improved but none of the points are blockers. Also the original developer, myself, is without a MediaWiki development environment since Brion's cross-country move. Having the current code go live is much more likely to attract improvements than letting it stagnate here for another year.

• The superfluous UNIQUE KEY is superficial and does not affect functionality nor performance.
• I thought the only ways the data was queried were covered by indeces. I did not add any superfluous indeces. However if I missed something I would regard that a minor issue that in any case may require a more experienced SQL developer than me.
• The preg_replace code is certainly ungainly but given that all of the regular expressions are of the very simplest type and the strings to be transformed are short Wiktionary page titles. They are a mustard seed in comparison with the regular expression processing done by the parser. Tim's custom replace routines would be much nicer here but implementing them would require testing whereas the current code is known to work. I'll file an enhancement request but it is no blocker.

Putting need-review back since I haven't looked at it in ages and would definitely want to look it over again before considering putting it live.

The extension can be tested on a full Wiktioanry mirror from the latest dump at http://70.79.106.1

Sadly this extension is no longer live anywhere I'm aware of.

r45305 on core and extension MWSearch states: "Make "did you mean" more Googley. It wasn't prominent enough; the red text and bolding should make it more visible, and also more familiar (since I took the styling straight from Google). I also moved it up way to the top: it *needs* to be one of the first things the user sees to be really useful. This was particularly terrible on enwiki, where the enormous bloated help message means you have to scroll down through half a page of largely useless nonsense before even seeing the suggestion."

Doesn't that outdate this request?

See https://www.mediawiki.org/wiki/Writing_an_extension_for_deployment for information on what is needed to get an extension reviewed before potentially deploying it on a wikisite.

Hello, this is a quasi-automated-but-not-really message:

I am reviewing all tracking bugs for extensions to review and deploy to WMF servers. See the list here:
https://bugzilla.wikimedia.org/showdependencytree.cgi?id=31235&hide_resolved=1

The [[mw:Review queue]] page lists the steps necessary to complete the review. I have copied them below and done some initial filling out based on what I can easily gleen from this bug and any linked to sources that are obvious. If I miss something/state something false, please do correct me.

Also, if you haven't yet done so, please review the information on and linked to from:
https://www.mediawiki.org/wiki/Writing_an_extension_for_deployment

TODO/Check list

Extension page on mediawiki.org:
Bugzilla component:
Extension in Gerrit:
Design Review:
Archeticecture/Performance Review:
Security Review:
Screencast (if applicable):
Community support: Yes? Not really anymore?

Next

Given the state of this request, I am inclined to close this bug in the next week if I don't hear back that it is indeed still wanted/useful.

Thanks!

Closing. Feel free to reopen if there is any maintained code to review/community consensus to install.