[Moult]

(AI-written, not verified by Dion.)

SQLAlchemy 2.1.0 (released 2026-09-24) removed the implicit coercion of a Select into a FROM clause that 1.4 and 2.0 had deprecated. union_select() relies on it for the count query behind __len__():

expression.select(*[functions.count().label('aCount')]).select_from(
    expression.select(*cols).where(whereClause).distinct().select_from(table))

Under 2.0 this emits SADeprecationWarning: Implicit coercion of SELECT and textual SELECT constructs into FROM clauses is deprecated; please call .subquery() ...; under 2.1 it raises ArgumentError: FROM expression, such as a Table or alias() object expected, got <Select> from sql.py:65, so len(graph) fails on every store (reached, for us, through brickschema's Graph.__len__).

The fix is the .subquery() the warning asks for. The SQL is identical under both versions:

SELECT count(*) AS "aCount" FROM (SELECT DISTINCT ... WHERE ...) AS anon_1 UNION ALL ...

Verified with the package's tests (DB=sqlite) on SQLAlchemy 2.0.54 and 2.1.0: same results, and no deprecation warning any more on 2.0. Note that on the current develop the suite needs #116 to import on Python 3.12+ (no pkg_resources), and three tests fail under rdflib 7 because the store is not graph-aware; I have a fix for that too and can open it as a follow-up PR if wanted.

[Andrej730]

Can confirm this is the correct solution, we've merged it downstream - gtfierro#2

If this will be merged - please consider using the version of this from downstream, it has some additional code simplifications for readability, just so downstream wouldn't have this extra diff they'll need to resolve later.

[mwatts15]

@Andrej730 can you do me a favor and make a PR on this repo with your downstream changes? Cheers, Mark W.

…

On Sat, Sep 26, 2026, 01:02 Andrej730 ***@***.***> wrote: *Andrej730* left a comment (RDFLib/rdflib-sqlalchemy#117) <#117 (comment)> Can confirm this is the correct solution, we've merged it downstream - gtfierro#2 <gtfierro#2> If this will be merged - please consider using the version of this from downstream, it has some additional code simplifications for readability, just so downstream wouldn't have this extra diff they'll need to resolve later. — Reply to this email directly, view it on GitHub <#117?email_source=notifications&email_token=AALLFSDOLCGXBADASJZ6MZL5Q5LWRA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOBUGM3TEMRYGI22M4TFMFZW63VKON2WE43DOJUWEZLEUVSXMZLOOSWGM33PORSXEX3DNRUWG2Y#issuecomment-5843722825>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AALLFSGUQGUIWNSHWCJS4AT5Q5LWRAVCNFSNUABDKJSXA33TNF2G64TZHMZTMNJWGEYTEO2JONZXKZJ3GU2TONJVGEYDANBYUF3AE> . Triage notifications, keep track of coding agent tasks and review pull requests on the go with GitHub Mobile for iOS <https://github.com/notifications/mobile/ios/AALLFSAR5WMUBPEW3XU4VLD5Q5LWRA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOBUGM3TEMRYGI22M4TFMFZW63VKON2WE43DOJUWEZLEUVSXMZLOOSVGM33PORSXEX3JN5ZQ> and Android <https://github.com/notifications/mobile/android/AALLFSDOIPSB6YENAGBA6P35Q5LWRA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTKOBUGM3TEMRYGI22M4TFMFZW63VKON2WE43DOJUWEZLEUVSXMZLOOSXGM33PORSXEX3BNZSHE33JMQ>. Download it today! You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[Andrej730]

@mwatts15 sure, submitted the PR (#119), thank you!

[mwatts15]

Merged an equivalent change to develop. Thanks for your contribution.