Wikidata:Requests for permissions/Bot/NinoBot
- The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
- Approved--Ymblanter (talk) 19:41, 24 March 2024 (UTC)[reply]
NinoBot (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)
Operator: Ignacio Rodríguez (talk • contribs • logs)
Task/s: Semi-automatically create version, edition or translation (Q3331189) entries for Spanish Wikisource Index pages.
Code: Gitlab. My repository is highly messy but the main program is "crawler.py" and most of the magic occurs in "UTILS.py"
Function details: My bot is an interactive commandline script that will parse Commons file pages that are linked to Spanish Wikisource, and don't have a Wikidata entry. An example item can be seen here: Q124806220. Currently:
- It parses a
{{:commons:Template:Book}}
template, and extract and process all relevant data. - It checks Wikidata for items with the same identifiers (as today: Internet Archive ID (P724), Google Books ID (P675), Galiciana work ID (P3004) and BDH edition ID (P4956), but adding more is trivial)
- Creates the Wikidata entry and fills the relevant info. Label and description must be manually confirmed.
I'm sorry if I conducted a few more tests than permitted. I needed live feedback and info from real works to actually build the many, many layers of filters that this script is made of (bibliographic data in Commons can be VERY messy). As all the edits were visually inspected and manually confirmed, I don't consider this to be strictly a bot, but bot flag would be nice (>8-12 edits per item). --Ignacio Rodríguez (talk) 14:39, 10 March 2024 (UTC)[reply]
- @Ignacio Rodríguez: it's really a great idea, I'd love to see the code (or at least some part of it) so we could replicate it for all Wikisources . Do you also plan to remove the data on Commons and recall them from Wikidata at the end? or at the least, adding the Q-ID on Commons, and like-wise: add the Q-ID on the Wikisource index? And if you have question on your side, don't hesitate to ask too! Cheers, VIGNERON (talk) 18:38, 11 March 2024 (UTC)[reply]
- @VIGNERON Yes! I can send you the code (careful, it's radioactive and not apt for human consumption). I think I saw you at the Wikisource Telegram. About the second part: it's very feasible, and in fact in the beginning it did that... In the end I realized that if I overwrote the info at Commons, I might be losing some info, so I decided to not touch it too much. Still, I have a safeguard: see author (P50) of Q124814257 (Q124814257), it has a little subject named as (P1810), so to be able to detect some errors. Ignacio Rodríguez (talk) 19:04, 11 March 2024 (UTC)[reply]
- PS: Of course I have another bot at Spanish WS that checks daily Index without QID and adds it from Commons. I could do it directly and I might. Ignacio Rodríguez (talk) 22:07, 11 March 2024 (UTC)[reply]