[pull] main from graphql:main #41
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
![pull[bot]](https://avatars.githubusercontent.com/in/12910?s=60&v=4){kind=small}
In GraphQL.js working group we discussed and put process in place. But we never officially wrote down in docs. You can watch this recording https://youtu.be/BBxc_HmB71I?si=cvdUMI47vBsT9F2y&t=1316 or check the transcript. These are the changes we agreed in the working group, just putting them in writing. --------- Co-authored-by: Benjie <[email protected]>
|
Hi @pull[bot], I'm @github-actions bot happy to help you with this PR 👋 Supported commandsPlease post this commands in separate comments and only one per comment:
|
The publish method checks to see if a subsequent result is empty; this same logic should be employed to suppress pending notices for empty records. This has already been achieved for subsequent results that are children of the initial result, as we generated the pending notices from the list of initially published records. For subsequent results that are children of other subsequent results, we previously generated the pending notice prior to actually publishing. This change integrates the logic: the publishing method itself returns a pending notice as required. This results in a bug-fix for subsequent records of other subsequent records as well as a reduction of code for subsequent results to the initial result.
Implements Option A from graphql/defer-stream-wg#80
see: nodejs/node#3675 and in particular: nodejs/node#3675 (comment)
changed in v8 11.3 / node 20 see: https://chromium.googlesource.com/v8/v8/+/73c45833472086ca1c6ab4140cdb058d1890e930%5E%21/src/flags/flag-definitions.h fixes benchmarking, see https://github.com/graphql/graphql-js/actions/runs/8314729475/job/22752257524#step:6:142
inline much of the original executeOperation leave out executeRootGroupedFieldSet, will help reduce the diff for #4026
Fixes the error message created by the DeferStreamDirectiveOnValidOperationsRule for the stream directive.
following graphql/graphql-spec#1077 now part of the following PR stack, with the laters PRs extracted from this one #4026: incremental: introduce GraphQLWrappedResult to avoid filtering #4050: perf: allow skipping of field plan generation #4051: perf: introduce completePromisedListItemValue #4052: refactor: introduce completeIterableValue #4053: perf: optimize completion loops #4046: perf: use undefined for empty
in the non-deferred case. depends on #4026 --------- Co-authored-by: Rob Richard <[email protected]>
refactoring that will streamline when we introduce two versions of this function to optimize the loop when not streaming depends on #4051
As convention, we will lazily instantiate arrays/sets when adding the first item. This applies to arrays/sets on execution/incremental context, as well as the second member of the GraphQLWrappedResult tuple holding the array of incremental data records.
…fragments (#3997) Group field sets should be properly consolidated when some of the fields in a sibling defer overlap with a child deferred fragment. The child deferred fragment should not prompt creation of an additional incremental data record for that field, because the field would always be sent with the parent.
Test logic was inadvertently altered in #4026. Tests still pass, just fixing logic.
… set results to reduce mutation
…grouped field sets error
…QLEnumValue (#4288) this extracts logic from #3044 and #3145 (later rebased as #3807 and #3808) to implement more informative error messages without implementing [the full schema coordinate RFC](graphql/graphql-spec#794) This is a BREAKING CHANGE because these schema elements are now longer plain objects and function differently in various scenarios, for example with `String(<schemaElement>` `JSON.stringifu(<schemaElement>` and `.toString()` and `.toJSON()` --------- Co-authored-by: Jovi De Croock <[email protected]>
I was just trying to understand the new flows for how scalar values are represented/parsed/serialized, and noticed that the comments for a couple of these methods looked backwards. Really hoping I am not just confused about the terminology, and that these aren't actually correct 😅
New Spec PR: graphql/graphql-spec#1098 Old Spec PR: graphql/graphql-spec#947 Original issue raised by @benjaminjkraft : #3592
this way, if (when?) we re-enable incemental delivery support with subscriptions, the signature of collect fields will not need to change
#3814 added default value validation and coercion, deprecating `astFromValue()` (which was unsafe, used `serialize()` to uncoerce input values provided in the internal format) and replacing it with a call to `valueToLiteral()` which safely operates on external values. This PR makes that change backwards compatible by reintroducing it as a new config option of `default` instead of replacing the existing option of `defaultValue`, where the type of `default` is: ```ts export type GraphQLDefaultInput = | { value: unknown; literal?: never } | { literal: ConstValueNode; value?: never }; ``` `default.value` is the external default value, while old config option of `defaultValue` is the internal value. Instead of removing it in v17, it is deprecated, and will be removed in v18. Side note: `GraphqlDefaultInput` renamed from `GraphQlDefaultValueUsage` introduced within the #3814 PR stack. Co-authored-by: Michael Hayes <[email protected]>
…erred (#4320) In the "old" duplicating version of incremental delivery, `collectFields()` was responsible for branching, and it therefore allowed processing a deferred named fragment even if that named fragment had already been visited as a regular non-deferred named fragment. One could have argued against that, as the old version had the concept of inlining, and so we could have just skipped the named fragment and considered it inlined, but chose not to do that. Now that we have an algorithm without duplication of fields, revisiting a named fragment will have no effect, as within the `buildExecutionPlan()` step, all the duplicated fields will be removed. So we can just remove the exception and go back to the version pre-incremental delivery where we always skip a visited named fragment. Note that we still only consider a named fragment to have been visited if it was not marked as deferred, because in the case of overlapping deferred and non-deferred spreads of the same named fragment, we need to visit it as the non-deferred spread to actually realize that it is non-deferred. [As an aside, looks like I made a mistake in #3994 => we would never have wanted to ignore the result of `shouldIncludeNode()` => since we are removing all ignoring of these conditions with respect to defer, this PR "fixes" that mistake as well.]
Updated to reflect spec draft graphql/graphql-spec#1132 Also changed the argument order to match the spec draft
preserving content motivation: this clears the way to upgrade our eslint configuration to v9, which should hopefully let us update our custom rule configuration currently, when attempting to frontport the next js website from v16 to main, we get eslint errors surrounding our custom rules. if we re-implement our custom rules according to the latest eslint configuration, it should make it easier this change allows us to do that inter alia by removing the eslint typedoc plugin
This converts the existing website to nextra just like https://github.com/graphql/graphql.github.io. This is a first step in moving the documentation here and having a redirect from graphql.org to graphql-js.org. Not sure yet why codecov started failing 😅 when I run `testonly:cover` locally it tells me we are 100% covered. WDYT about isolating the dependencies for the website in that folder? As seen in 9c7d615 this prevents the weird CI leaks that we're seeing Resolves #4200 --------- Co-authored-by: Yaacov Rydzinski <[email protected]>
Checked for broken styles and fixed them and also went over the links by using linkinator. I did check a lot of tools but looks like there isn't a good way at the moment to check for broken links that is maintained. Resolves #4242
Ultimately I was trying this out to see whether we can tweak the docs easily, it made me realize that our docs are tailored to general GraphQL rather than how do we use this library. It made me come up with a few suggestions - We should have a toggle on code examples to switch between `buildSchema` and programatically creating the schema with i.e. `GraphQLObjectType` - Our documentation starts with a tutorial, this ultimately feels like a mistake, we should lead with an explanation of what GraphQL.JS is and what it aims to do, clearly outlining the goals of this project - We should line out use-cases for building on this library and best practices of how to go to production Resolves #2941 Resolves #2567
Not quite sure yet about contents of the overview page, also the header is pretty odd, feels like a nextra bug or I goofed the CSS up 😅 generally though it looks like the extra-button and search/... aren't in their own container preventing a good space-between. The absence of links seems to cause thsi.
This provides people with the option to choose between the template approach or the classes approach. This is a proposal to tackle #1368 [Preview](https://graphql-7w0ort26u-the-graph-ql-foundation.vercel.app/) This has been applied throughout the codebase now, however one of the things I am uncertain about is how we offer `buildSchema` with the GraphQLDefer/... directives? Should we add an option to `buildSchema`? The exports defined in that chapter seem to only exist in v17 so we should explicitly flag that.
Following typescript documentation, it's not possible override interface property, we can only add new props. Since it's declared as unknown dict, even if we merge `GraphQLErrorExtensions`, we can't access to our extensions without workaround and verbose typescript in source-code. It's annoying since apollo expose `GraphQLFormattedError` instead `GraphQLError` Refs: https://www.typescriptlang.org/docs/handbook/declaration-merging.html Refs: apollographql/apollo-client#11789
Currently input-unions and by extension the `@oneOf` directive aren't present in the documentation. I have opted to put this into the advanced section. The copy might be up for improvement, honestly fire away if there's more cases to cover, just wanted to get the ball rolling here. CC @benjie --------- Co-authored-by: Benjie <[email protected]>
CC @dimaMachina Is it possible for the sidebar on `api-v16` to restart? currently it inherits the root one
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot]
Can you help keep this open source service alive? 💖 Please sponsor : )