introduces mapSchemaConfig utility function

[yaacovCR]

motivation:

1. rebuilds lexicographicSortSchema() on this generic utility, better demonstrating only the sorting
2. rebuilds extendSchemaImpl() on this utility
3. can be used as base to more easily enhance extendSchema()/buildASTSchema()/buildSchema() to take resolvers, etc
4. can be used to expose a generic safe mapSchemaConfig() utility
5. improved performance -- not the initial motivation, but turns out these changes bring an 80% speed improvement to buildSchema().

[netlify]

✅ Deploy Preview for compassionate-pike-271cb3 ready!

Name	Link
🔨 Latest commit	7298f67
🔍 Latest deploy log	https://app.netlify.com/sites/compassionate-pike-271cb3/deploys/67459968a7f8d6000836b710
😎 Deploy Preview	https://deploy-preview-4297--compassionate-pike-271cb3.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

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

[github-actions]

Hi @yaacovCR, 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

[yaacovCR]

note: diff is easier to scan when hiding whitespace

[yaacovCR]

This re-implementation brings a 24% performance boost to buildSchema() and a small drop in memory use.

I'm not sure yet what exactly brings the performance boost, I was just hoping for parity.

[yaacovCR]

Just poking around a bit with deopt-explorer, which, as usual, is beyond me, but the summary for main is:

and the summary for map is:

For clarity, within the map version, we rename getWrappedType() to typeFromAST() and getNamedType() to namedTypeFromAST() because it allows us to use getNamedType() in the SchemaContext closure without a clash -- and these functions essentially are a pre-GraphQLSchema-creation version of our typeFromAST() utility with the exact same algorithm.

For clarity, I also renamed buildType() to buildNamedType().

So the thing that leaps out at me is that the number of deoptimizations in buildType()/buildNamedType() and extendSchemaImpl() drop, the former significantly. I am not 100% on why. I can only guess that splitting the functions into smaller bits helped.

[yaacovCR]

UPDATE:

• I have separated out the commits introducing mapSchemaConfig() and those used to rebuild lexicographicSortSchema() and extendSchemaImpl(), the latter underlying buildSchema() and extendSchema().
• The initial commit has full test coverage on its own.
• I have introduced another "helper" method available to the mapper functions, getNamedTypes() which means that we don't have to manually track the added types and then concatenate them to the pre-existing types within extendSchemaImpl().

Benchmark results from current version are even better than before, an ~80% improvement on main:

[github-actions]

@github-actions publish-pr-on-npm

@yaacovCR The latest changes of this PR are available on NPM as
graphql@17.0.0-alpha.7.canary.pr.4297.42bcd01f7c733e255ff9eab84ec8d7500dfd208a
Note: no gurantees provided so please use your own discretion.

Also you can depend on latest version built from this PR:
npm install --save graphql@canary-pr-4297

[ab-pm]

It seems this is now duplicated between

graphql-js/src/utilities/extendSchema.ts

Lines 574 to 579 in ef042cf

	const stdTypeMap = new Map(
	[...specifiedScalarTypes, ...introspectionTypes].map((type) => [
	type.name,
	type,
	]),
	);

and

graphql-js/src/utilities/mapSchemaConfig.ts

Lines 459 to 464 in ef042cf

	const stdTypeMap = new Map(
	[...specifiedScalarTypes, ...introspectionTypes].map((type) => [
	type.name,
	type,
	]),
	);

One of them should suffice :-)