Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add parser support for string literal aliases #4023

Open
wants to merge 2 commits into
base: main
Choose a base branch
from

Conversation

cuhtis
Copy link

@cuhtis cuhtis commented Feb 26, 2024

These are the graphql-js changes for a proposed spec change that changes the definition of an alias from:

Alias : Name :

...to:

Alias
 - Name :
 - StringValue :

At Meta, our GraphQL clients for both web and mobile care a lot about local data consistency. Local data consistency is when the client is able to reconcile data from multiple GraphQL responses in a local client-side cache which can be subscribed to. When data changes, we issue updates to these subscribers — which allows us to keep various screens "in-sync" and show the same view of the data.

The canonical example of this on Facebook is if you navigate to the Groups page from a Feed post and join the group, then when you navigate back to Feed we should update the "Join Group" button in the post to "Joined".

Local data consistency is supported today by Relay for JS clients (widely adopted within industry) and by Meta's internal mobile GraphQL client for Android/iOS clients.

To support consistency, we need to be able to remap aliases/field names into their "canonical names". The canonical name is what we use to keep two fields consistent, as it represents the true definition of a field.

Given a field selection:

alias: field(arg: "foo")

the canonical name would be:

field(arg:"foo")

Internally, we've explored various ways of doing this. We currently need to embed a lot of information about the schema and query in our clients; for example Relay creates a "normalization AST" with this information, and our internal mobile client requires all the information to be pre-compiled into the app binary. This leads to additional costs, e.g. binary size bloat or wire size costs.

We've found that we can more efficiently deliver the canonical name information by embedding it into our response instead, which removes the need for pre-compiled metadata. We run a transform on our queries to add aliases for each field, and use the canonical name as the alias. However since the canonical name may contain non-supported characters (e.g. (, ), $), we are proposing a spec change to allow StringValue tokens as the alias.

This would allow syntax such as:

"field(arg:\"foo\")": field(arg: "foo")

We've attempted alternative ways of implementing this, such as doing this transform during server-side execution (a spec violation, and high server overhead!) or delivering encoded aliases to abide by the NameValue specification (client parsing overhead for decoding!).

Potential side effects / outcomes of this change:

  1. Selection set conflict validation should still work out of the box, even if one selection is a string literal and the other is a normal NameValue.
  2. Here we parse the StringValue into a NameNode (same as when parsing a NameValue), which doesn't have any requirements on the name itself as far as I can tell.
  3. We allow string literal aliases here but not number literals, which abides by the JSON specification. According to the JSON spec, map/object keys must be strings.

Summary: These are the graphql-js changes for a proposed spec change that changes the definition of an alias from:
```
Alias : Name :
```
...to:
```
Alias :
 - Name :
 - StringValue :
```

At Meta, our GraphQL clients for both web and mobile care a lot about local data consistency. Local data consistency is when the client is able to reconcile data from multiple GraphQL responses in a local client-side cache which can be subscribed to. When data changes, we issue updates to these subscribers — which allows us to keep various screens "in-sync" and show the same view of the data.

The canonical example of this on Facebook is if you navigate to the Groups page from a Feed post and join the group, then when you navigate back to Feed we should update the "Join Group" button in the post to "Joined".

Local data consistency is supported today by Relay for JS clients (widely adopted within industry) and by Meta's internal mobile GraphQL client for Android/iOS clients.

To support consistency, we need to be able to remap aliases/field names into their "canonical names". The canonical name is what we use to keep two fields consistent, as it represents the true definition of a field.

Given a field selection:
```
alias: field(arg: "foo")
```
the canonical name would be:
```
field(arg:"foo")
```

Internally, we've explored various ways of doing this. We currently need to embed a lot of information about the schema and query in our clients; for example Relay creates a "normalization AST" with this information, and our internal mobile client requires all the information to be pre-compiled into the app binary. This leads to additional costs, e.g. binary size bloat or wire size costs.

We've found that we can more efficiently deliver the canonical name information by embedding it into our response instead, which removes the need for pre-compiled metadata. We run a transform on our queries to add aliases for each field, and use the canonical name as the alias. However since the canonical name may contain non-supported characters (e.g. `(`, `)`, `$`), we are proposing a spec change to allow `StringValue` tokens as the alias.

This would allow syntax such as:
```
"field(arg:\"foo\")": field(arg: "foo")
```

We've attempted alternative ways of implementing this, such as doing this transform during server-side execution (a spec violation, and high server overhead!) or delivering encoded aliases to abide by the `NameValue` specification (client parsing overhead for decoding!).

Potential side effects / outcomes of this change:
1. Selection set conflict validation should still work out of the box, even if one selection is a string literal and the other is a normal `NameValue`.
2. Here we parse the `StringValue` into a `NameNode` (same as when parsing a `NaveValue`), which doesn't have any requirements on the name itself as far as I can tell.
3. We allow string literal aliases here but not number literals, which abides by the JSON specification. According to the JSON spec, map/object keys must be strings.


Test Plan: `npm test`
Copy link

linux-foundation-easycla bot commented Feb 26, 2024

CLA Missing ID CLA Not Signed

Copy link

netlify bot commented Feb 26, 2024

Deploy Preview for compassionate-pike-271cb3 ready!

Name Link
🔨 Latest commit 3fc725d
🔍 Latest deploy log https://app.netlify.com/sites/compassionate-pike-271cb3/deploys/65dcf6458561b900085fb092
😎 Deploy Preview https://deploy-preview-4023--compassionate-pike-271cb3.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify site configuration.

@zpao
Copy link

zpao commented Feb 26, 2024

/easycla

Copy link

Hi @cuhtis, I'm @github-actions bot happy to help you with this PR 👋

Supported commands

Please post this commands in separate comments and only one per comment:

  • @github-actions run-benchmark - Run benchmark comparing base and merge commits for this PR
  • @github-actions publish-pr-on-npm - Build package from this PR and publish it on NPM

@cuhtis cuhtis marked this pull request as ready for review February 26, 2024 20:38
@cuhtis
Copy link
Author

cuhtis commented Feb 26, 2024

/easycla

@yaacovCR
Copy link
Contributor

yaacovCR commented Sep 26, 2024

Spec PR: graphql/graphql-spec#1082

@yaacovCR yaacovCR added the spec RFC Implementation of a proposed change to the GraphQL specification label Sep 26, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
spec RFC Implementation of a proposed change to the GraphQL specification
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants