Commons:Bots/Work requests

From Wikimedia Commons, the free media repository
< Commons:Bots(Redirected from Commons:BWR)
Jump to: navigation, search

Shortcut: COM:BR · COM:BWR

Bot policy and list · Requests to operate a bot · Requests for work to be done by a bot · Changes to allow localization  · Requests for batch uploads

Filing cabinet icon.svg
SpBot archives all sections tagged with {{Section resolved|1=~~~~}} after 1 day.

Music copyright detection bot[edit]

Wikipedia Zero's made us attractive as a piracy host (T129845). To combat music piracy I suggest setting up an audio fingerprinting system like Echoprint or AcoustID. These are open sourced (unlike Shazam or Gracenote) and supposedly easy to get running. Tracking this as T132650. Dispenser (talk) 23:19, 14 April 2016 (UTC)

This is a great idea. Sadly, we would be using AcoustID under "commercial use".. while open source, it would cost wikimedia money to use it legally. [1] . For echoprint, "If you download our data and then add to it, you are required to contribute data back to us." - This seems like not a big issue, as common files can be used freely (usually with Attribution). Public domain songs would be no issue to get to them. Cheers! CoolCanuck eh? 20:16, 9 May 2016 (UTC)
User:CoolCanuck, would you mind re-posting that to phabricator to keep the discussion centralized? Dispenser (talk) 14:02, 16 May 2016 (UTC)

Easy: scrape Unsplash photo metadata[edit]

Unsplash photos usually have metadata in /info pages, eg. However, the images themselves have no metadata embedded in them. It would be useful to get a bot to fetch the metadata in /info pages and categorize Images from Unsplash accordingly. ~pikolas [[mia diskuto]] 19:41, 15 April 2016 (UTC)

@Pikolas: Example diffs please. :) Riley Huntley (talk) 00:22, 24 April 2016 (UTC)
@Riley Huntley: Hey Riley, thanks for replying. I'm not sure what you meant to say though. I was thinking of having the bot adding categories such as "Photos taken with camera XXXX". Maybe it would be possible to alter the actual metadata of the image, but I'm not sure if this is currently possible. ~pikolas [[mia diskuto]] 04:11, 24 April 2016 (UTC)
@Pikolas: Altering the actual EXIF data requires reuploading the images, this is not an "easy" task by any means and painfully displays it's information. Seems like a lot of work for less than 500 images, but I'll leave the judging to a bot operator whos wanting to do this task. Riley Huntley (talk) 00:58, 26 April 2016 (UTC)

Automatic upload of USGS Earthquake shakemaps[edit]

See also: Category:ShakeMaps

Not unimportant: there should be a bot polling USGS-Earthquake , then in case of Earthquake with Magnitude > 7 it should instantly upload the shakemap (site,file) . --Itu (talk) 20:24, 29 April 2016 (UTC)

Isn't this something better suited for Wikidata? Or is it more than a single point with a radius gradient? Dispenser (talk) 21:20, 29 April 2016 (UTC)
It's definitely more complicated than a simple gradient. -- King of ♠ 23:12, 29 April 2016 (UTC)
?? - It's not about having raw data or deriving data from images. It's about instant automatic transfer of images generated by USGS earthquake authority to commons whenever a relevant earthquake occurs. --Itu (talk) 01:39, 30 April 2016 (UTC)

Commons:Template i18n/bot replacements[edit]

If anyone wants to help, I'll be running my bot through Commons:Template i18n/bot replacements later today. Some of these are being cleaned up on-going, many of them aren't. I've just noticed the following as well that I'll add to the list when I get a chance:


I appreciate any help. :) Riley Huntley (talk) 17:47, 2 May 2016 (UTC)

Invalid EXIF dates[edit]

Regrettably, UploadWizard completes the date field of {{Information}} template with untrustworthy dates without distinguishing their source and meaning by corresponding templates. This situation requieres a lot of work on detection of suspected, invalid and confusing dates.

Some of candidates for {{Invalid EXIF date}} template are pages with automatically filled dates with time 00:00, especially in combination with date of any first day of the month (see an example). Some systematic seeking of suspected dates should be designed. --ŠJů (talk) 22:26, 14 May 2016 (UTC)

They can be pulled off the database. Before writing a bot, you could use Quarry to test out some SQL that parses the metadata using the img_metadata field of the image table. E.g. 8020. -- (talk) 06:47, 20 May 2016 (UTC)

Moving infos from category to description[edit]

We uploaded about 17,000 images from Paolo Monti Archive. We wanted to reuse as much infos as possible from the original catalogue, so we have a lot of categories to fix. We need in particular to tranfer a string from a category to the description. Most of images from Monti's Archive have an unexisting (red), long category. That category has those characteristics:

  1. uncreated (red)
  2. shared only by an image (or two)
  3. formed by words separated by commas (not by "-" as the other original categories).
  4. 20 chars long or more

We need to trasfer that string (deleting the category) to the beginning of the description field.
Example: File:Paolo Monti - Serie fotografica (Anzola d'Ossola, 1966) - BEIC 6329139.jpg has a category called "Anzola d'Ossola. Riale. Alberi." (note the commas). The text "Anzola d'Ossola. Riale. Alberi." must be transferred at the very beginning of the description (deleting that category and leaving intact the other categories and the previous description).
A query made by Federico Leva shows we have about 7,000 categories of more than 75 chars that are used in less than 3 images (and 1,000 used in 2). I suggest to consider every red category with 20 chars or more and with commas between words. Thank you for your help. --Marco Chemello (BEIC) (talk) 20:59, 23 May 2016 (UTC)