Page MenuHomePhabricator

Add Australia to Monuments Database
Closed, ResolvedPublic

Description

The data will live on Wikidata so this is blocked by:

Preliminary mapping

{
    "country": "au",
    "lang": "en",
    "description": "Commonwealth and National heritage sites in Australia in English",
    "type": "sparql",
    "project": "",
    "table": "monuments_au_(en)",
    "commonsTemplate": "Cultural Heritage Australia",
    "commonsTrackerCategory": "Cultural heritage monuments in Australia with known IDs",
    "commonsCategoryBase": "Cultural heritage monuments in Australia",
    "unusedImagesPage": "",
    "imagesWithoutIdPage": "",
    "missingCommonscatPage": "",
    "sparql": "{ ?item wdt:P3008 ?id } UNION { ?item wdt:P1435 wd:Q30108476 } UNION { ?item wdt:P1435 wd:Q30108476 }",
    "sql_lang": "English  # Wikidata",
    "sql_country": "Australia",
    "sql_data": {
        "dataset": {
            "value": "au",
            "type": "Text"
        },
        "lang": {
            "value": "en",
            "type": "Text"
        },
        "adm0": {
            "value": "au",
            "type": "Text"
        }
    }
}

Event Timeline

LilyOfTheWest moved this task from Backlog to Next Up on the Wiki-Loves-Monuments (2017) board.

Missing are

  • project
  • unusedImagesPage
  • imagesWithoutIdPage
  • missingCommonscatPage
  • lang (can't remember what we need if we want it on e.g. commons)

So I submitted a patch with a working config.

Because of T172973: Handle multiple values for a property we get multiple versions of the same item returned from the query. (e.g. same id but different coordinates)

If we want the report pages to be on en.wikipedia then most things should work. If we however want them to be on commons (where the lists are) then, similar to T173783: Get Malaysia into the monuments database, we get and issue with commons not having a langugage code. Additionally the query attempts to look for monuments_article on commons since project is used in the query.

Change 375537 had a related patch set uploaded (by Lokal Profil; owner: Lokal Profil):
[labs/tools/heritage@master] [WIP] Add monuments_config for Australia

https://gerrit.wikimedia.org/r/375537

I need the desired names of the report pages and to know on which wiki they should live before i can take this further.

@Gnangarra Could you let us know where to put the report pages? On these pages Erfgoedbot will report what images could potentially be added to which monument items (unusedImagesPage), which images don't have an identifier (imagesWithoutIdPage) and missingCommonscatPage (Andre/JF: please remind me what this was doing?). This page could live on Commons or enwp.

@Lokal_Profil I'm not sure if after that the workflow of adding images to the items is entirely clear. You may want to put those instructions on top of those pages (this is different for Wikidata than wikipedia-lists).

@Gnangarra the report pages are still needed.

@Effeietsanders I created T176200 to deal with the needed instructions.

[...] missingCommonscatPage (Andre/JF: please remind me what this was doing?).

This one is less commonly used but IIRC it checks if any commons category has been tagged with a monuments template and if so suggests that this be added as a commonscat for that monument (in list or on wikidata)

I've now added these as sub-pages to the contact's (@Gnangarra ) user page on en.wiki. If you want them somewhere else let me know.

The pages are on en.wiki rather than commons for the same reason as the complexities in T173783: Get Malaysia into the monuments database.

The sub-pages are:

Change 375537 merged by jenkins-bot:
[labs/tools/heritage@master] Add monuments_config for Australia

https://gerrit.wikimedia.org/r/375537

Mentioned in SAL (#wikimedia-cloud) [2017-09-25T09:49:08Z] <JeanFred> Deploy latest from Git master: 1ab75f9 (T174333)

Harvesting was already in progress so I sneakily ran harvesting on that dataset in parallel manually. It should make it into monuments_all in the next 4 hours.

Lokal_Profil claimed this task.

Woho the first real sparql harvest in production =)