Adding generated common table expressions

1,093 views
Skip to first unread message

Ashley Waite

unread,
Mar 17, 2017, 7:28:17 AM3/17/17
to Django developers (Contributions to Django itself)

Hey all,


I'd like to suggest adding Common Table Expression (CTE) query generation as a feature to Django.

I've been working on a project that required manipulation of many records at once, and as with many ORMs found that this wasn't an ideal use-case in Django. As the rest of our code base and related projects are in Django, there was a strong preference to find a way to do it and keep to the same model-is-the-truth design.

I first did this by writing some hackish functions using raw querysets and generating my own CTE based queries, but it lacked ideal flexibility and maintainability. So I've now written some modifications into my Django to do this in a more Django-esque way and think that this functionality would be beneficial within the project itself, but am unsure exactly where to start the conversation about that.


Why generate CTE based queries from querysets?

By allowing querysets to be attached to each other, and setting appropriate WHERE clauses, arbitrary and nested SQL queries can be generated. Where the results of the queries are only necessary for the execution of following queries this saves a very substantial amount of time and database work. Once these features exist, other functionality can also transparently use these to generate more efficient queries (such as large IN clauses).

This allows several powerful use cases I think Django would benefit from:


Large 'IN' clauses, can be implemented as CTEs reducing expensive lookups to a single CTE INNER JOIN. For sets of thousands to match from tables of millions of records this can be a very substantial gain.


Composite 'IN' conditions, where multiple fields must match and you're matching against a large set of condition rows. In my usage this was "where the md5/sha hashes match one of the million md5/sha tuples in my match set". This is simply a CTE JOIN with two clauses in the WHERE.


Nested data creation, where the parent doesn't yet exist. Django doesn't currently do this as the primary keys are needed, and this makes normalised data structures unappealing. Using INSERTs as CTEs that supply those keys to following statements means that entire nested data structures of new information can be recreated in the database at once, efficiently and atomically.


Non-uniform UPDATEs, such that a modified set of objects can all be updated with different data at the same time by utilising a CTE values statement JOINed to the UPDATE statement. As there's currently no way to do this kind of bulk update the alternative is to update each instance individually, and this doesn't scale well.

These could also be used with aggregations and other calculated fields to create complex queries that aren't possible at the moment.


What my PoC looks like

With another mildly hackish PoC that creates a VALUEs set from a dict/namedtuple which can be used to provide large input data, my present modified version syntax looks a bit like this (not perfect queries):

class Hashes(models.Model):
md5 = models.UUIDField(verbose_name="MD5 hash (base16)", db_index=True)
sha2 = models.CharField(max_length=44, null=True, verbose_name="SHA256 hash (base64)")

# Mock QuerySet of values q_mo = Hashes.as_literal(input_hashes).values("md5", "sha2")
# A big IN query
q_in = Hashes.objects.attach(q_mo).filter(md5=q_mo.ref("md5"))

# Matched existing values with composite 'IN' (where md5 and sha2 match, or md5 matches and existing record lacks sha2) q_ex = Hashes.objects.attach(q_mo).filter(md5=q_mo.ref("md5")).filter(Q(sha160=q_mo.ref("sha160")) | Q(sha160=None)) # Create new records that don't exist q_cr = Hashes.objects.attach(q_mo, q_ex).filter(md5=q_mo.ref("md5")).exclude(md5=q_ex.ref("md5")).values("md5", "sha2").as_insert()
Returning the newly created records.

SQL can be generated that looks something like this:
WITH cte_1_0 (md5, sha2) AS (
	VALUES ('00002d30243bfe9d06673765c432c2bd'::uuid, 'fsA8okuCuq9KybxqcAzNdjlIyAx1QJjTPdf1ZFK/hDI='::varchar(44)),
('0000f20a46e4e60338697948a0917423', '6bVZgpYZtit1E32BlANWXoKnFFFDNierDSIi0SraND4=')), cte_1 AS ( SELECT "hashes"."id", "hashes"."md5", "hashes"."sha2" FROM "hashes" , "cte_1_0" WHERE ("hashes"."md5" = (cte_1_0.md5) AND ("hashes"."sha2" = (cte_1_0.sha2) OR "hashes"."sha2" IS NULL) )) SELECT "hashes"."md5" FROM "hashes" , "cte_1_0" , "cte_1" WHERE ("hashes"."md5" = (cte_1_0.md5) AND NOT ("hashes"."md5" = (cte_1.md5)))
That is:
  • A qs.as_insert() and qs.as_update() on queryset to create *lazy* insert and update queries.
  • A qs.attach() that allows querysets to be attached to other querysets, and will generate them as CTE statements.
  • A qs.ref() that returns an expression that when the query is compiled will be a field reference on the CTE that represents that queryset.
  • Additional compilers on the QuerySet subclasses that these return (so no changes to base compilers meaning no functionality impact to existing usage)
  • Generation of WITH clauses for attached querysets, and RETURN clauses for lazy UPDATE and INSERT querysets with fields requested (via values() in this case)

As these can be attached to querysets that are attached to querysets, that are... etc, many statements can be chained allowing substantial changes to be performed without needing Django to have to receive, process, and resend at every step.

I've had a read through the enhancement proposal docs etc, and I'm willing to do what's needed to make this functionality solid, and put forth a proposal to add it. But am first seeking feedback on it, and whether this is a feature that will be considered.


Thanks,
- Ashley

Tim Graham

unread,
Mar 17, 2017, 7:49:39 PM3/17/17
to Django developers (Contributions to Django itself)
Hi, I don't know anything about CTE, but did you see this third-party package? https://github.com/petrounias/django-cte-trees -- It seems to be PostgreSQL only.

I was going to write, "Considering that not all databases support CTE (MySQL doesn't), a third-party app might be the way to go rather than having it built-in to Django." -- however, then I noticed that supported is added in MySQL 8 [0], so maybe any database differences could be abstracted away with an appropriate API.

I'll leave it to others to comment on the technical details of your proposal. Meanwhile, if you have a chance to contribute smaller patches to Django, I think it's helpful to start with smaller patches before tackling something larger like this.

[0] http://mysqlserverteam.com/mysql-8-0-labs-recursive-common-table-expressions-in-mysql-ctes/

Matthias Kestenholz

unread,
Mar 18, 2017, 3:59:57 AM3/18/17
to django-d...@googlegroups.com
Hi,

On Sat, Mar 18, 2017 at 12:49 AM, Tim Graham <timog...@gmail.com> wrote:
> Hi, I don't know anything about CTE, but did you see this third-party
> package? https://github.com/petrounias/django-cte-trees -- It seems to be
> PostgreSQL only.

Just chiming in to point out a maintained and up-to-date friendly fork
of the project above:
https://github.com/matthiask/django-cte-forest

Ashley, is your PoC available somewhere? Is your implementation in any
way similar to the implementation in
django-cte-trees/django-cte-forest?

Thanks,
Matthias

Adam Johnson

unread,
Mar 18, 2017, 5:08:10 AM3/18/17
to django-d...@googlegroups.com
 supported is added in MySQL 8 [0]

Additionally MariaDB 10.2 supports them, and it's nearing release.


--
You received this message because you are subscribed to the Google Groups "Django developers  (Contributions to Django itself)" group.
To unsubscribe from this group and stop receiving emails from it, send an email to django-developers+unsubscribe@googlegroups.com.
To post to this group, send email to django-developers@googlegroups.com.
Visit this group at https://groups.google.com/group/django-developers.
To view this discussion on the web visit https://groups.google.com/d/msgid/django-developers/CANvPqgDWUPuUA6rw8uawa7_AsUkXFhuegFM1cJ-gjVUCxvBUwg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.



--
Adam

Josh Smeaton

unread,
Mar 18, 2017, 5:28:53 AM3/18/17
to Django developers (Contributions to Django itself)
Thanks for bringing this up Ashley, and for all of the detail you provided. I'd certainly like to see CTEs make their way into Django, provided we could come up with a nice enough API. From the look of it, you've already got something that works with an okay API so I'm hopeful.

I'd be very interested in seeing your POC too if you're able to share.

From looking very briefly at django-cte-trees it doesn't aim to support user defined CTEs for anything other than recursive queries. I'd be interested in seeing, as part of a DEP, how CTE inclusion in django core could support the cte-trees project from an API perspective.

Anssi Kääriäinen

unread,
Mar 20, 2017, 3:23:44 AM3/20/17
to Django developers (Contributions to Django itself)
+1 for the INSERT and UPDATE parts.

For the .attach() method, I wonder if it would be better to work on having generic subquery support in the ORM instead of targeting CTEs specifically. From Django's perspective there isn't any big difference between:
    WITH fooquery AS (SELECT * FROM foo WHERE ...)
    SELECT * FROM baz JOIN fooquery ON ...
and
    SELECT * FROM baz JOIN (SELECT * FROM foo WHERE ...) fooquery ON ...

The great thing about the subquery form is that it works on every database, and to my knowledge the subquery support is surprisingly standard on all database Django needs to care about.

If we want full support for either of the above cases, then we need some way to tell to the Django's ORM what the fooquery (either as subquery or with CTE) produces. For simple cases this won't be complex (say, for .values() it's just a list of expressions), but what to do when the subquery/CTE contains .select_related() for example? Then we'd have a situation where the subquery produces a list of expressions but those expressions might refer to different models in the CTE.

So, the starting point should be to allow only "simple" queries in .attach(). This means that the attached query must be either .values() queryset, or a queryset containing expressions from single model only (no .extra(), .annotate() or select_related() added). In addition I think .attach() should support subqueries instead of using only CTEs. We get support on every backend instead of PostgreSQL only for pretty much the same amount of effort.

 - Anssi

Ashley Waite

unread,
Mar 21, 2017, 8:08:49 PM3/21/17
to Django developers (Contributions to Django itself)
I'm nesting the query generators as they currently exist and adding awareness of that nesting into another compiler, so that queries that reference other models, contain annotations, etc should still in theory work fine.

I do agree that the best method would be to generalise this to incorporate subqueries as well, but I considered those to be a subset of general CTE queries which can effectively be any valid query.

With some minor modifications to the existing INSERT/UPDATE/SELECT compilers via subclassing I think that subqueries as special cases of general statements that have some restrictions is probably the easiest route.
My current fiddling has changed INSERT/UPDATE to have returning clauses using subclassed compilers:
https://github.com/ashleywaite/django/blob/cte-dev/django/db/models/sql/subqueries.py#L174-L209
https://github.com/ashleywaite/django/blob/cte-dev/django/db/models/sql/compiler.py#L1283-L1321

Such that regular non-RETURN aware uses aren't impacted by that code - but I suspect that making INSERT and UPDATE be RETURN aware and lazily executed as a default case might be better, and more consistent with the way that SELECT based querysets work.

I'm using a workaround to feed in VALUES clauses, which again I think would probably be better as a proper QuerySet subclass that is used by INSERT statements and CTEs alike.
The language is a bit confusing because .values() is a method in Django, so mock querysets that define literal values intended to generate a VALUES clause don't really have a good name, but they would introduce some consistency as that's what INSERT statements are doing, and what non-uniform UPDATEs and composite IN style queries would want to do.

The leading CTE WITH clauses I'm generating are done via it's own compiler as well, such that only queries that are aware of having such queries attached will generate them:
https://github.com/ashleywaite/django/blob/cte-dev/django/db/models/sql/compiler.py#L1324-L1374

These could fairly easily be adapted to use a more general case for attaching one query to another, as my relatively haskish PoC was mostly to show that it can be done without substantial complexity, and that modification of other systems to better incorporate these patterns is probably a good idea.

As they stand now, CTEs used in the way I am using them could with little effort be modified to support aggregations, and annotations on the base query to pull in values from complex generated CTE's is probably the ideal way to perform some multi-table calculations.

- Ashley

Ashley Waite

unread,
Mar 21, 2017, 8:14:10 PM3/21/17
to Django developers (Contributions to Django itself)
CTE Forest is a specific use case of CTEs to store self-referential tree data sets.

It's quite a different use to what I'm proposing, though its implementation might become simpler if Django is generally CTE aware.

- Ashley

Ashley Waite

unread,
Mar 21, 2017, 8:32:09 PM3/21/17
to Django developers (Contributions to Django itself)
Here's the code changes I've made, noting that some of them were to shove in a generalised VALUES clause that mocks being a queryset, so that it plays with the same interface.

https://github.com/django/django/compare/master...ashleywaite:cte-dev#files_bucket

I've had a glance at cte-trees/cte-forest and once general CTEs are worked out expanding that to include recursive CTEs wouldn't be too difficult, and that would greatly simplify the implementation of cte-forest to the extent that it might be viable as a django data/reference type.

- Ashley

Marc Tamlyn

unread,
Apr 6, 2017, 4:53:32 AM4/6/17
to django-d...@googlegroups.com
Hi!

This looks generally very good, and I'm quite excited about it.

In terms of taking it forwards, I think a DEP is a very good idea, and there are at least 3 core devs who are keen to see a solution. Whether you have the right solution yet I'm not so sure, but it sounds like you're not either.

Regarding Anssi's comments about SubQuery, we do now have that in core as of 1.11 [0]. It does look like an .attach() approach might actually have been a nicer version of this, but on the other hand it's currently implementable solely with the Expressions API. It seems like the OuterRef is very similar to your queryset.ref(). An even nicer approach using attach could be to say qs.attach(q1=some_qs).filter(a=F('q1__b'))?

Looking forwards to seeing a DEP!


--
You received this message because you are subscribed to the Google Groups "Django developers (Contributions to Django itself)" group.
To unsubscribe from this group and stop receiving emails from it, send an email to django-developers+unsubscribe@googlegroups.com.
To post to this group, send email to django-developers@googlegroups.com.
Visit this group at https://groups.google.com/group/django-developers.

Ashley Waite

unread,
Apr 6, 2017, 5:11:21 AM4/6/17
to django-d...@googlegroups.com
Oh right, I feel silly for not thinking of named arguments as a resolution there, that's a damn good idea! 

It resolves a couple of minor issues I'd anticipated in a complete implementation, so consider the idea stolen!  ;)

I mean, ummm, that was totally what I had planned! 

- Ashley

You received this message because you are subscribed to a topic in the Google Groups "Django developers (Contributions to Django itself)" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/django-developers/b370mxfKCHg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to django-developers+unsubscribe@googlegroups.com.

To post to this group, send email to django-developers@googlegroups.com.
Visit this group at https://groups.google.com/group/django-developers.

Anssi Kääriäinen

unread,
Apr 6, 2017, 9:14:01 AM4/6/17
to Django developers (Contributions to Django itself)
On Thursday, April 6, 2017 at 11:53:32 AM UTC+3, Marc Tamlyn wrote:
Regarding Anssi's comments about SubQuery, we do now have that in core as of 1.11 [0]. It does look like an .attach() approach might actually have been a nicer version of this, but on the other hand it's currently implementable solely with the Expressions API. It seems like the OuterRef is very similar to your queryset.ref(). An even nicer approach using attach could be to say qs.attach(q1=some_qs).filter(a=F('q1__b'))?

Hmmh, we have one form of SubQuery, but that's actually for SELECT clause, not for FROM clause. I believe the same class won't work for the CTE or subquery in FROM clause case.

As for the attach(), seems like a really nice syntax. We do need something for generating the join clause for the JOIN. If you look at an example:
    q1 = Book.objects.values('author_id').annotate(avg_price=Avg('price'))
    q2 = Author.objects.attach(q1=q1)
it needs to create something like:
WITH q1 AS (
    SELECT author_id, avg(price) FROM book GROUP BY author_id
)
   FROM author
   LEFT JOIN q1 ON author.id = q1.author_id;

Or, equivalently without the CTE:
   FROM author
   LEFT JOIN ( SELECT author_id, avg(price) FROM book GROUP BY author_id) ON author.id = q1.author_id;

Now, the main points are:
   1. There is no need to design this to be about CTEs. That just limits the feature from backends that don't have CTEs without any real benefit. From Django's perspective the two above queries are the same.
   2. We do need something for the JOIN ON condition. In some cases Django could guess this, but there needs to be an explicit way to express the join condition.

If we allow usage of expressions from the attached queryset, but don't try to go for cases where model instance are created from the attached queryset, this will be both possible to implement without having to write a change-everything patch, and this will also be a really nice feature.

For recursive CTEs, I'd leave that strictly as a later step. The only thing we need to check right now is that we don't do something that prevents a good recursive CTEs implementation later on.

 - Anssi
To unsubscribe from this group and stop receiving emails from it, send an email to django-develop...@googlegroups.com.
To post to this group, send email to django-d...@googlegroups.com.

matthew.pava

unread,
Nov 29, 2017, 3:09:39 PM11/29/17
to Django developers (Contributions to Django itself)
I would like to know what the status of this is (adding CTE functionality to Django).  It is supported by all current databases that Django supports.

Jacob Rief

unread,
Jul 22, 2019, 5:22:12 AM7/22/19
to Django developers (Contributions to Django itself)
For my better understanding, could you please elaborate on what you mean by Hashes.as_literal(input_hashes) ?

Javier Buzzi

unread,
Oct 17, 2019, 5:43:49 PM10/17/19
to Django developers (Contributions to Django itself)
What do you think of this syntax instead?

q1 = Book.objects.values('author_id').annotate(avg_price=Avg('price'))

q2
= Author.objects.attach('book_prices', q1, id=F('book_prices__author_id'))


def attach(name, queryset, **params):
   
# Would look something like this.
   
...


Same sql output.

Gaga Ro

unread,
May 10, 2022, 10:01:11 AM5/10/22
to Django developers (Contributions to Django itself)
Hello everyone,

I'm often using django-cte and I'd be thrilled to have CTE in the core.

If I'm not mistaken, the only DB currently supported by Django and not having CTE support is MySQL 5.7 (with an end of life in October 2023). I don't know if Django 4.2 will support it, but it should be dropped for Django 5.0 as it will be released in 2023. So we should have all supported DB supporting CTE when this feature would be over.

The ticket (https://code.djangoproject.com/ticket/28919) has been stalled for a few years now, this thread as well. I am willing to work on this but I would like more information first.

If I try to list all the requirements, we should have:

* A way to add one or more CTE.
* A way to reference the columns from the CTE.
* A way to join them in the main query.
* Setting a CTE as recursive?
* Choosing if a CTE is materialized or not (Not all DB support that, and I'm not sure if they all handle it the same way)?
* Insert / delete CTE with returning data?

Do we have a better idea now of what the API should look like?

Thanks.

Matthew Pava

unread,
May 10, 2022, 10:08:06 AM5/10/22
to django-d...@googlegroups.com

I will always suggest that we use the Subquery API to make CTEs. To make them recursive, just add a keyword argument (recursive=True) and/or use a union.

 

It’s been a while since I looked at CTEs, so I might be missing something.

 

I would hate to see us create an entirely separate API for CTEs.

Gaga Ro

unread,
May 10, 2022, 11:18:27 AM5/10/22
to Django developers (Contributions to Django itself)
As far as I know, CTE and subquery are equivalent when used only in a single place in the query.

CTE should be better when a single query is used several times.

If we want to reuse the Subquery API, we should find a way to be able to reuse a subquery (hash its content and use that as a key to detect the duplication?).
Reply all
Reply to author
Forward
0 new messages