html_url,issue_url,id,node_id,user,created_at,updated_at,author_association,body,reactions,issue,performed_via_github_app https://github.com/simonw/datasette/issues/1080#issuecomment-720695174,https://api.github.com/repos/simonw/datasette/issues/1080,720695174,MDEyOklzc3VlQ29tbWVudDcyMDY5NTE3NA==,9599,2020-11-02T20:05:26Z,2020-11-02T20:05:26Z,OWNER,"URL design: `/database/table/-/facet/colname` And for other types of facet (to be supported later): `/database/table/-/facet/colname?_type=m2m`","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631, https://github.com/simonw/datasette/issues/1080#issuecomment-720696827,https://api.github.com/repos/simonw/datasette/issues/1080,720696827,MDEyOklzc3VlQ29tbWVudDcyMDY5NjgyNw==,9599,2020-11-02T20:08:49Z,2020-11-02T20:13:56Z,OWNER,"Implementing pagination for facets will be interesting. Would be easier if I had a nicer reusable internal pagination mechanism, which is also needed for #856 (pagination of canned queries).","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631, https://github.com/simonw/datasette/issues/1080#issuecomment-720697226,https://api.github.com/repos/simonw/datasette/issues/1080,720697226,MDEyOklzc3VlQ29tbWVudDcyMDY5NzIyNg==,9599,2020-11-02T20:09:38Z,2020-11-02T20:09:38Z,OWNER,"Maybe this ends up being code that defers to a simulated canned query, rendered using the existing `query.html` template.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631, https://github.com/simonw/datasette/issues/1080#issuecomment-720698577,https://api.github.com/repos/simonw/datasette/issues/1080,720698577,MDEyOklzc3VlQ29tbWVudDcyMDY5ODU3Nw==,9599,2020-11-02T20:12:26Z,2020-11-02T20:12:26Z,OWNER,"For regular column faceting, here's the query that is used: https://github.com/simonw/datasette/blob/13d1228d80c91d382a05b1a9549ed02c300ef851/datasette/facets.py#L196-L204 Since it uses `order by count desc, value` maybe those values could be used to implement cursor-based pagination. That wouldn't be robust in the face of changing data, but I'm not sure it's possible to implement paginated faceting in a way that survives ongoing changes to the underlying data.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631, https://github.com/simonw/datasette/issues/1080#issuecomment-720699160,https://api.github.com/repos/simonw/datasette/issues/1080,720699160,MDEyOklzc3VlQ29tbWVudDcyMDY5OTE2MA==,9599,2020-11-02T20:13:42Z,2020-11-02T20:13:42Z,OWNER,Also relevant to this issue: #830 - redesigning the facet plugin hook in preparation for Datasette 1.0. And #972 supporting faceting against arbitrary queries.,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631, https://github.com/simonw/datasette/issues/1080#issuecomment-1029691693,https://api.github.com/repos/simonw/datasette/issues/1080,1029691693,IC_kwDOBm6k_c49X9kt,9599,2022-02-04T06:16:45Z,2022-02-04T06:16:45Z,OWNER,"Had a new, different idea for how this could work: support a `?_group_count=colname` parameter to the table view, which turns the page into a `select colname, count(*) ... group by colname` query - but keeps things like the filter interface, facet selection, search box and so on.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631, https://github.com/simonw/datasette/issues/1080#issuecomment-1029695083,https://api.github.com/repos/simonw/datasette/issues/1080,1029695083,IC_kwDOBm6k_c49X-Zr,9599,2022-02-04T06:24:40Z,2022-02-04T06:25:18Z,OWNER,"An initial prototype of that in my local `group-count` branch quickly started running into problems: ```diff diff --git a/datasette/views/table.py b/datasette/views/table.py index be9e9c3..d30efe1 100644 --- a/datasette/views/table.py +++ b/datasette/views/table.py @@ -105,8 +105,12 @@ class RowTableShared(DataView): type_ = ""integer"" notnull = 0 else: - type_ = column_details[r[0]].type - notnull = column_details[r[0]].notnull + try: + type_ = column_details[r[0]].type + notnull = column_details[r[0]].notnull + except KeyError: # Probably count(*) + type_ = ""integer"" + notnull = False columns.append( { ""name"": r[0], @@ -613,6 +617,15 @@ class TableView(RowTableShared): offset=offset, ) + # If ?_group_count we convert the SQL query here + group_count = request.args.getlist(""_group_count"") + if group_count: + wrapped_sql = ""select {cols}, count(*) from ({sql}) group by {cols}"".format( + cols="", "".join(group_count), + sql=sql, + ) + sql = wrapped_sql + if request.args.get(""_timelimit""): extra_args[""custom_time_limit""] = int(request.args.get(""_timelimit"")) ``` Resulted in errors like this one: ``` pk_path = path_from_row_pks(row, pks, not pks, False) File ""/Users/simon/Dropbox/Development/datasette/datasette/utils/__init__.py"", line 82, in path_from_row_pks bits = [ File ""/Users/simon/Dropbox/Development/datasette/datasette/utils/__init__.py"", line 83, in row[pk][""value""] if isinstance(row[pk], dict) else row[pk] for pk in pks IndexError: No item with that key ```","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",734777631,