Обсуждение: Improve error message for duplicate labels in enum types

Hi,

Currently, when creating an enum type, duplicate labels are caught by a unique
index on pg_enum, resulting in a generic error message. 

 postgres=# create type t as enum ('a','b','a');
 ERROR:  duplicate key value violates unique constraint "pg_enum_typid_label_index"
 DETAIL:  Key (enumtypid, enumlabel)=(16418, a) already exists.

I propose adding an explicit check for duplicate labels during enum creation,
so that a more user-friendly and descriptive error message can be produced,
similar to what is already done in ALTER TYPE ... ADD VALUE
or ALTER TYPE ... RENAME VALUE .. TO ....
 
With the attached patch applied, the error message becomes:

ERROR:  label "a" used more than once

Regards,
Yugo Nagata


-- 
Yugo Nagata <nagata@sraoss.co.jp>

Hi Rahila,

On Fri, 4 Jul 2025 07:42:58 +0530
Rahila Syed <rahilasyed90@gmail.com> wrote:

>  Thank you for sharing the patch.
> +1 to the idea of improving the error message.

Thank you for your review.

> Please take the following points mentioned into consideration.
> 
> 1. I’m considering whether there might be a more efficient way to handle
> this.
> The current method adds an extra loop to check for duplicates, in addition
> to the existing duplicate index check,
> even when no duplicates are present. Would it be possible to achieve this
> by wrapping the following
> insert call in a PG_TRY() and PG_CATCH() block and logging more descriptive
> error in the PG_CATCH() block?

Although this introduces a loop-in-loop for checkin the dulicates, I believe
the impact to the performance is not high because the number of values in an
enum would be not so large and creating an enum type is not executed so fequently.
I found check_duplicates_in_publist() and interpret_function_parameter_list take
similar ways, where duplicates are checked in nested loops.

Although this introduces a nested loop to check for duplicates, I believe the
performance impact is negligible, since the number of values in an enum is typically
small, and enum type creation is not a frequent operation.
I found that check_duplicates_in_publist() and interpret_function_parameter_list()
take a similar approach, using nested loops to check for duplicates.

If we were to use PG_TRY and PG_CATCH block, we wouldn't be able to identify exactly 
which label is duplicated.
 
> 2. If we choose to include the check in the 0001 patch you provided, would
> it make more sense to place
> it earlier in the function, before assigning OIDs to the labels and running
> qsort? This way, we could
> catch duplicates sooner and prevent unnecessary processing.

If the duplicate check were done before the loop, we would have to add an extra nested
loop, which seems wasteful. The loop is the first place where the actual label text is
accessed using strVal(), and there's already a check for the label length here.
So I believe this is also the appropriate place to check for duplicates.

Regards,
Yugo Nagata

-- 
Yugo Nagata <nagata@sraoss.co.jp>

Hi Yugo

On 03.07.25 17:04, Yugo Nagata wrote:
> Currently, when creating an enum type, duplicate labels are caught by a unique
> index on pg_enum, resulting in a generic error message. 
>
>  postgres=# create type t as enum ('a','b','a');
>  ERROR:  duplicate key value violates unique constraint "pg_enum_typid_label_index"
>  DETAIL:  Key (enumtypid, enumlabel)=(16418, a) already exists.
>
> I propose adding an explicit check for duplicate labels during enum creation,
> so that a more user-friendly and descriptive error message can be produced,
> similar to what is already done in ALTER TYPE ... ADD VALUE
> or ALTER TYPE ... RENAME VALUE .. TO ....
>  
> With the attached patch applied, the error message becomes:
>
> ERROR:  label "a" used more than once
>

The error message for already existing enum labels starts with "enum",
e.g.  ERROR:  enum label "bar" already exists. So, perhaps this new
error message should follow the same pattern?

I also wonder if we need to add tests for it, so that we make sure the
new error is triggered prior to the generic one, e.g. in create_type.sql

-- check for duplicate enum entries
CREATE TYPE den AS ENUM ('foo','bar','foo');
CREATE TYPE en AS ENUM ('foo','bar');
ALTER TYPE en ADD VALUE 'foo';
ALTER TYPE en RENAME VALUE 'foo' TO 'bar';
DROP TYPE en;

Regards, Jim

Hi Jim,

On Mon, 25 Aug 2025 10:08:23 +0200
Jim Jones <jim.jones@uni-muenster.de> wrote:

> Hi Yugo
> 
> On 03.07.25 17:04, Yugo Nagata wrote:
> > Currently, when creating an enum type, duplicate labels are caught by a unique
> > index on pg_enum, resulting in a generic error message. 
> >
> >  postgres=# create type t as enum ('a','b','a');
> >  ERROR:  duplicate key value violates unique constraint "pg_enum_typid_label_index"
> >  DETAIL:  Key (enumtypid, enumlabel)=(16418, a) already exists.
> >
> > I propose adding an explicit check for duplicate labels during enum creation,
> > so that a more user-friendly and descriptive error message can be produced,
> > similar to what is already done in ALTER TYPE ... ADD VALUE
> > or ALTER TYPE ... RENAME VALUE .. TO ....
> >  
> > With the attached patch applied, the error message becomes:
> >
> > ERROR:  label "a" used more than once
> >
> 
> The error message for already existing enum labels starts with "enum",
> e.g.  ERROR:  enum label "bar" already exists. So, perhaps this new
> error message should follow the same pattern?

Thank you for taking a look. That makes sense, so I updated the message to:

 ERROR:  enum label "a" used more than once


> I also wonder if we need to add tests for it, so that we make sure the
> new error is triggered prior to the generic one, e.g. in create_type.sql
> 
> -- check for duplicate enum entries
> CREATE TYPE den AS ENUM ('foo','bar','foo');
> CREATE TYPE en AS ENUM ('foo','bar');
> ALTER TYPE en ADD VALUE 'foo';
> ALTER TYPE en RENAME VALUE 'foo' TO 'bar';
> DROP TYPE en;

I also added a test for duplicate enum entries to enum.sql,
since tests for existing entries are already there.

Regards,
Yugo Nagata

-- 
Yugo Nagata <nagata@sraoss.co.jp>

On 26.08.25 04:55, Yugo Nagata wrote:
> Thank you for taking a look. That makes sense, so I updated the message to:
>
>  ERROR:  enum label "a" used more than once

Nice.

> I also added a test for duplicate enum entries to enum.sql,
> since tests for existing entries are already there.

+1

LGTM; I'll mark the CF entry as Ready for Committer.


Regards, Jim

Jim Jones <jim.jones@uni-muenster.de> writes:
> LGTM; I'll mark the CF entry as Ready for Committer.

Pushed with some trivial cosmetic adjustments, including the
perhaps-not-so-trivial fix of removing the comment you falsified.

I was concerned about Rahila's upthread worry about the performance
of this approach, but in some quick testing it seemed to add only
barely-noticeable overhead even at 1000 enum labels.  At 10000
labels it's slightly annoying: my machine goes from ~80ms to ~250ms.
But that seems well beyond what anybody would be likely to use,
so I judge it not worth trying to be smarter.

The obvious solution if we did wish to avoid the O(N^2) behavior would
be to qsort the labels and then compare only adjacent ones.  That'd
require a temporary array though, and I'd bet it's actually slower
than this way for normal-sized enums.  Another possibility perhaps is
to apply the check only when there are fewer than say 1000 labels,
reasoning that anything bigger is probably machine-generated anyhow.

            regards, tom lane

On Tue, 2 Sep 2025 14:00:52 -0400
Tom Lane <tgl@sss.pgh.pa.us> wrote:

> Jim Jones <jim.jones@uni-muenster.de> writes:
> > LGTM; I'll mark the CF entry as Ready for Committer.
> 
> Pushed with some trivial cosmetic adjustments, including the
> perhaps-not-so-trivial fix of removing the comment you falsified.

Thank you for committing the patch and for fixing it.

> I was concerned about Rahila's upthread worry about the performance
> of this approach, but in some quick testing it seemed to add only
> barely-noticeable overhead even at 1000 enum labels.  At 10000
> labels it's slightly annoying: my machine goes from ~80ms to ~250ms.
> But that seems well beyond what anybody would be likely to use,
> so I judge it not worth trying to be smarter.
> 
> The obvious solution if we did wish to avoid the O(N^2) behavior would
> be to qsort the labels and then compare only adjacent ones.  That'd
> require a temporary array though, and I'd bet it's actually slower
> than this way for normal-sized enums.  Another possibility perhaps is
> to apply the check only when there are fewer than say 1000 labels,
> reasoning that anything bigger is probably machine-generated anyhow.

I also thought the O(N^2) behavior was acceptable, since it seemed unlikely
that users would try to create such a large number of enum labels.
If any complaints arise in the future, we can address them then.

Regards,
Yugo Nagata

-- 
Yugo Nagata <nagata@sraoss.co.jp>