Yaacov Rydzinski

original `execute` and `subscribe` should throw if defer/stream are used

5

Currently, when the defer/stream directives are not present in the schema, defer/stream will not be triggered, and so the directives will be ignored even when used with the original function....

incrementalDelivery: original execute should not throw on new directives

7

original execute and executeSync should ignore `@defer` and `@stream` rather than throw. extracted from #3726

incrementalDelivery: use experimental options rather than experimental functions

7

Using TS conditional types, when `enableExperimental` is set to true, return type of `execute` will correspond to the new types; otherwise, `execute` will return the old types. When the option...

replace raw mapAsyncIterable with Repeater based approach

7

The basic idea behind this PR is that graphql-js should not export something with type signature of an async generator that sometimes does not behave like an async generator. Repeaters...

Add support for new graphql tools selectionSet hints

1

Rather than fragment replacements...

allow nested defers at the same level

4

Implements Option A from https://github.com/graphql/defer-stream-wg/discussions/80

Add test for consolidating grouped field sets properly into deferred fragments

3

Group field sets should be properly consolidated when some of the fields in a sibling defer are masked by parent fields.

do not emit pending for empty non-published subsequent results

2

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...

draft: add support for fine grained aborting promises on error

2

abort async resolvers on errors TODO: connect return() and throw() events on returned iterators to abortion of resolvers for subscriptions/incremental delivery

add checks for actual deferral

3

depends on #3993 adds judicious explicit checking for the expected extra ticks when fields are deferred. @robrichard -- these new tests all also pass with #3982 without any changes =>...