Wikipedia:Bots/Requests for approval/Taxobot 2

The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA. The result of the discussion was Approved.

Taxobot 2

• (BRFA
• Approved BRFAs
• talk
• contribs
• count
• SUL
• logs
• page moves
• block log
• rights log
• flag)

Operator: Smith609 (talk · contribs)

Automatic or Manually assisted: Manually supervised by users

Programming language(s): PHP

Source code available: Available at Google Code

Function overview: This bot will help users to create the back-end templates that support Template:Automatic taxobox. The bot will transform a user's input into the correct syntax, and will suggest input to the user based on existing data in Wikipedia.

Links to relevant discussions (where appropriate): Template_talk:Taxobox#Usability

Edit period(s): When triggered by a user.

Estimated number of pages affected: A couple of pages per user activation; eventually (as use of Automatic Taxobox becomes more widespread) there will be one page per taxon.

Exclusion compliant (Y/N): Yes

Already has a bot flag (Y/N):

Function details:

This bot can be activated by a user when a new Automatic taxobox is created.

A link will be provided in the template to the HTML bot interface. The user will be asked to provide their username, which will be displayed in the edit summary; only valid usernames will be allowed to use the tool. (This system works well at User:Citation bot.)

• [*] The user clicks the link. The bot requests from the user:
  • The taxonomic rank of the taxon
  • The taxonomic parent of the taxon
  • The scientific name of the taxon (pre-filled to the name of the referring page).

• The bot validates the user's input and presents any possible errors (unrecognized rank? parent lacks WP page?) to the user.

• The user corrects and confirms the input

• The bot converts this input into a new page at Template:Taxonomy/taxon x, formatted per the correct syntax. (If the page already exists the user is presented with an error message.)

• The bot checks to see whether the parent taxon already has a page at Template:Taxonomy/parent taxon.

• • If so, the bot's work is done; the new taxon has been connected to the existing tree of life.

• • If not, the bot helps the user to create taxonomy for the parent taxon.

• • • The user is asked to provide the information listed at [*] above.
    • If the parent taxon has a WP page, the fields are pre-filled from the article on the parent taxon.
    • If the parent taxon lacks a WP page, WikiSpecies is consulted for this information.

• • The bot returns to [*], helping the user to create new back-end pages until the taxonomic hierarchy is linked to the existing tree of life.

Whilst the bot physically creates the pages, all the information entered has been manually verified by users and is entered in a rigid format. The bot will never amend existing data.

I propose that during the initial testing period, the only user authorised to activate the bot is myself (Smith609 (talk · contribs)). Once the bot is operating as I expect I suggest allowing other users to use the bot, with the output being scrutinized by myself (and the BRFA team?).

Discussion

• I'd be ok with this. I like the new Automatic taxobox format and I think this will greatly improve the conversion. Two questions related to this request: If the parent taxon does not have a Template:Taxonomy/parent taxon and the bot pre-fills data sourced from the parent taxon's Template:Taxobox on the article page, will there be appropriate instructions and warnings that the pre-filled fields should be checked and not assumed to be correct? I know of many instances where genus and family articles do not agree in all taxonomy fields. Will the bot then also prompt the user to replace the parent taxon's old taxobox with the shiny new automatic taxobox? And one unrelated question to think about for the future: How easy or self-evident will it be to insert a new taxon level within an existing automatic taxobox structure for the uninitiated? Say there's a genus and Wikipedia treats it as only containing species, but there's an accepted infrageneric classification and a editor wants to insert it between the genus and species levels. Would the user just edit an automatic taxobox on a species page to list the parent as the new taxon, thus prompting this bot to kick in and do its magic? Cheers, Rkitko ^(talk) 22:06, 16 September 2010 (UTC)

  The warnings that you suggest are definitely a good idea.

  Replacing old taxoboxes with automatic taxoboxes is a great idea, but for simplicity is not covered within this bot request. I will create a separate bot request when I have the time to implement this feature, if it is wanted.

  Modifying automatic taxoboxes is also a little tricky at present; again, this is something that I can submit a bot request for in the future. There are a couple of ways that "easy editing" can be added to the taxobox template and this is something that I indend to discuss once the template is a little more widely used.

  Martin (Smith609 – Talk) 00:06, 17 September 2010 (UTC)

• I've got a draft version of the bot going at http://toolserver.org/~verisimilus/Bot/taxobot/taxobot.php. Obviously, no edits will be committed until the approval process is complete and I have checked for bugs and added an input-verification system, but this should give interested parties an idea of what to expect. Martin (Smith609 – Talk) 21:45, 17 September 2010 (UTC)

  Hey Martin, thanks for your replies above. All sounds good to me! I tried out the tool using Stylidium graminifolium as an example. When it pre-filled the data from the taxobox for the parent, it just displayed "S" instead of picking up the section Linearis as the parent. Not a big issue, but might be confusing to some. Any ideas why it did that? Cheers, Rkitko ^(talk) 21:42, 21 September 2010 (UTC)

   Fixed – Stylidium sect. Lineares is now returned as the parent. Martin (Smith609 – Talk) 16:49, 22 September 2010 (UTC)

  Ah, also, running the script on Stylidium sect. Debilia, it chose Stylidium for the pre-filled parent taxon, ignoring Stylidium subg. Tolypangium. I assume it's picking "major" taxa ranks to pre-fill. Not a bad idea, but should be explained. Rkitko ^(talk) 21:51, 21 September 2010 (UTC)

   Fixed – now selects "Stylidium subg. Tolypangium" as parent genus. Martin (Smith609 – Talk) 16:39, 22 September 2010 (UTC)

  Hah, ok, last one: Try running Quercus berberidifolia. It pre-fills the taxon field with just "Quercus" and not the full binomial. It also pre-fills just a "Q" in the parent field. Minor issues, I know, since the function of the bot is not to do everything for you. Just curious. Cheers, Rkitko ^(talk) 22:09, 21 September 2010 (UTC)

   Fixed – Now recognizes "Section Quercus" as parent. Martin (Smith609 – Talk) 16:47, 22 September 2010 (UTC)

• Thanks for all these reports. It looks like there was a problem with handling the semi-duplicate data provided in binomial / trinomial parameters. Also with non-alphanumeric characters (e.g. spaces) in taxonomic names. These should be readily fixed; I've made a start but will complete the process anon. Martin (Smith609 – Talk) 14:44, 22 September 2010 (UTC)

  All these should now be fixed. Thanks for the reports, and do let me know any other unusual cases. Martin (Smith609 – Talk) 16:49, 22 September 2010 (UTC)

• I have a rare period of free time coming up so am eager to move on with the bot approvals process if possible. If it's possible to approve a trial quickly, or to suggest any necessary amendments, that would be great! Martin (Smith609 – Talk) 16:24, 20 September 2010 (UTC)
• There don't seem to have been any objections to the principle of the bot's operation; therefore I'm requesting a trial period of fifty edits. Martin (Smith609 – Talk) 22:35, 24 September 2010 (UTC)

• Looks great as far as finding that parent taxon goes. I did a couple tests using taxa I had created articles for a few years ago (though I don't have any recent ones to try out) and they worked perfectly. Can't wait to get this working...btw, you know the video game Spore's wikia website already HAS a template that does all this automatic taxonomy stuff? Lucky them.... Bob the Wikipedian ^{(talk • contribs)} 21:58, 26 September 2010 (UTC)

Trial

• Approved for trial (50 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. MBisanz ^talk 23:09, 30 September 2010 (UTC)

• Edits underway and available for inspection at Special:Contributions/Taxobot; I've added automatic taxoboxes created via this method at Leptomitus and Vauxia (using novel taxonomic information from an original source) and Bactroceras (using taxonomic information from existing taxoboxes). As always, comments are very welcome! Martin (Smith609 – Talk) 01:01, 1 October 2010 (UTC)

• WOAH. That template actually looks a bit scary. So the implementation is clear to me-- all one does is they place {{automatic taxobox}} on the page and add the fossil range, authority, and subdivisions, as needed. The part that's scaring me, though, is where is the data coming from? Bob the Wikipedian ^{(talk • contribs)} 21:43, 2 October 2010 (UTC)

Never mind-- I figured it out by reading the documentation for the template. Looks like the bot is doing a good job so far. Bob the Wikipedian ^{(talk • contribs)} 22:49, 2 October 2010 (UTC)

Well done. Tim1357 ^talk

 Approved.. Tim1357 ^talk 14:12, 22 October 2010 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA.