Errors for Graph Type

[mnd999]

Error codes for graph types.

This work is likely to be feature flagged for some time, so we don't want to merge it to the public docs yet, but we do need to review the errors.

[Hunterness]

I haven't looked at the implementation PR but I did have some thoughts on the error messages from just looking at them in isolation.

Status: 0 open comment in this group and 1 in the next

[Hunterness]

do we want a second example where the node elem type and constraint have different properties (to show that it's not only a problem when the properties are the same)

[mnd999]

I don't think we need to, if somebody is looking at this page, they already have the error.

[Hunterness]

maybe, but I just felt that since both the node and the relationship example had the same property it might be nice to also show some other cases that would fail

ALTER CURRENT GRAPH TYPE SET {
    (p:Person => {name :: STRING}),
    CONSTRAINT FOR (p:Person) REQUIRE p.ssn IS NOT NULL
}


Fix to:
ALTER CURRENT GRAPH TYPE SET {
    (p:Person => {name :: STRING, ssn:: ANY NOT NULL})
}

(this both have another property and doesn't say we expect the label to be identifying in the constraint syntax)

[mnd999]

I don't think we need any more examples, most of the error codes do not have any.

[Hunterness]

yeah, cause they didn't have time to add any when doing the initial docs, they want there to be examples for all codes (I've been asked to add examples when I updated codes previously) so that's not really a metric to go after

[Hunterness]

but if you don't want new/additional example then update the existing one instead: I just want some more variation in the examples and not have the very same case in both the node and relationship version (just switching which constraint type is where)

[Hunterness]

New round of comments

Status: only open comments are from the groups above

[Hunterness]

maybe, but I just felt that since both the node and the relationship example had the same property it might be nice to also show some other cases that would fail

ALTER CURRENT GRAPH TYPE SET {
    (p:Person => {name :: STRING}),
    CONSTRAINT FOR (p:Person) REQUIRE p.ssn IS NOT NULL
}


Fix to:
ALTER CURRENT GRAPH TYPE SET {
    (p:Person => {name :: STRING, ssn:: ANY NOT NULL})
}

(this both have another property and doesn't say we expect the label to be identifying in the constraint syntax)

[renetapopova]

Many editorial comments, some of which require fixing the description in the codebase. When these are addressed, I'll build it locally and read it again. I also need to regenerate the index file to add the new codes.

[renetapopova]

Looks good. A few minor comments.

[Hunterness]

Noticed when double checking the messages and the impl updates

[Hunterness]

Review of new errors (and a brief look through the old)

Status: two open comments

[Hunterness]

Since Gowri have (multiple times) stumbled on the inline key constraints it might be nice with an example of that as well

Given
ALTER CURRENT GRAPH TYPE SET {
    (:Label => :ToBeRemoved {prop :: STRING IS KEY})
}

Then:
ALTER CURRENT GRAPH TYPE ALTER {
    (:Label => {prop :: STRING IS KEY})
}

[Hunterness]

I see that you have it for DROP, but I think Gowri have had more questions on it for ALTER 🤷

[Hunterness]

would we need some indication that these are two queries and not one containing two alter clauses? 🤔 (since this would fail parsing on the second ALTER I'd assume)

[mnd999]

How is this normally done?

[Hunterness]

I would guess either separating them into two code blocks or having ; after them, but I'd double check how the docs team would want it

[Hunterness]

Next group of comments

[mnd999]

@renetapopova I've updated this with the remaining set of error codes, and there is a couple of outstanding questions for docs.

[renetapopova]

@renetapopova I've updated this with the remaining set of error codes, and there is a couple of outstanding questions for docs.

I'll take a look next week.

[renetapopova]

Is this PR for 2025.10?

[mnd999]

@renetapopova No, I don't think this is ready for release yet.

[Hunterness]

I think you're adding and not removing a KEY constraint 🤔

[renetapopova]

This file is autogenerated. You don't need to update it.

[renetapopova]

I added some suggestions and a placeholder for the version.

[neo4j-docops-agent]

This PR includes documentation updates
View the updated docs at https://neo4j-docs-status-codes-341.surge.sh

New pages:

• 22NC1
• 22NC2
• 22NC3
• 22NC4
• 22NC5
• 22NC6
• 22NC7
• 22NC8
• 22NC9

Updated pages:

• List of GQLSTATUS error codes