issue_comments

Demos:

• https://latest.datasette.io/fixtures/neighborhood_search
• https://timezones-api.now.sh/timezones/by_point

Documentation: http://datasette.readthedocs.io/en/latest/sql_queries.html#canned-queries

So it would look like this: { "databases": { "timezones": { "license": "ODbL", "license_url": "http://opendatacommons.org/licenses/odbl/", "queries": { "by_point": { "title": "Timezones by point", "description": "Find the timezone for a latitude/longitude point", "sql": "select tzid\nfrom\n timezones\nwhere\n within(GeomFromText('POINT(' || :longitude || ' ' || :latitude || ')'), timezones.Geometry)\n and rowid in (\n SELECT pkid FROM idx_timezones_Geometry\n where xmin < :longitude\n and xmax > :longitude\n and ymin < :latitude\n and ymax > :latitude\n )" } } } } }

https://timezones-api.now.sh/-/metadata currently shows this: { "databases": { "timezones": { "license": "ODbL", "license_url": "http://opendatacommons.org/licenses/odbl/", "queries": { "by_point": "select tzid\nfrom\n timezones\nwhere\n within(GeomFromText(\u0027POINT(\u0027 || :longitude || \u0027 \u0027 || :latitude || \u0027)\u0027), timezones.Geometry)\n and rowid in (\n SELECT pkid FROM idx_timezones_Geometry\n where xmin \u003c :longitude\n and xmax \u003e :longitude\n and ymin \u003c :latitude\n and ymax \u003e :latitude\n )" }, "source": "timezone-boundary-builder", "source_url": "https://github.com/evansiroky/timezone-boundary-builder", "tables": { "timezones": { "license": "ODbL", "license_url": "http://opendatacommons.org/licenses/odbl/", "sortable_columns": [ "tzid" ], "source": "timezone-boundary-builder", "source_url": "https://github.com/evansiroky/timezone-boundary-builder" } } } }, "license": "ODbL", "license_url": "http://opendatacommons.org/licenses/odbl/", "source": "timezone-boundary-builder", "source_url": "https://github.com/evansiroky/timezone-boundary-builder", "title": "OpenStreetMap Time Zone Boundaries" }

We could support the value part of the "queries" array optionally being a dictionary with the same set of metadata fields supported for a table, plus a new "sql" key to hold the SQL for the query.

That's a good idea. We already do this for tables - e.g. on https://fivethirtyeight.datasettes.com/fivethirtyeight-ac35616/most-common-name%2Fsurnames - so having it as an option for canned queries definitely makes sense.