Renée
To match the `graphql-js` implementation in `printSchema`, it may be useful to have an option in `.serialize()` to omit non-semantic directives from the output. This would represent the schema in...
In introspection results (https://github.com/apollographql/apollo-rs/issues/675), default values for input types are encoded as strings containing GraphQL syntax: https://spec.graphql.org/draft/#sec-The-__InputValue-Type To build an `apollo_compiler::Schema` from an introspection response, we would need to parse...
## Description Many errors emitted by the lexer do not set the `error.index` property, so they point to the start of the document by default. ### Steps to reproduce Input...
Currently we only support `\u1234`, but the draft spec adds `\u{12345678}` syntax. We should add support for that in the lexer and the parser. https://spec.graphql.org/draft/#EscapedUnicode
We moved from miette to ariadne so we can create diagnostics that reference multiple files at once. This caused one major regression in our diagnostic display: ariadne often prints way...
ref https://github.com/apollographql/apollo-rs/pull/347#discussion_r1020325156 Right now parser errors show the index and the current token it's looking at, which isn't all that helpful to end users. IIRC Simon mentioned using the Python...
The parser does its best with any invalid input it gets. We have a few cases where it could be better. ```graphql type A { field1: Int |||garbage||| field2: Int...
Related to #325 and #446 If we have some input with completely invalid characters, for example: ```graphql type A { > } ``` `
This error came up on CITGM: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/1773/nodes=debian8-64/testReport/junit/(root)/citgm/clinic_v4_0_0/ (Ctrl+F for "not ok") It seems flakey: there are multiple tests in that same suite that do almost the same thing, but only...
I've been publishing alphas for a few years, it would be nice to wrap it up and have a proper thing that we can recommend to people to `npm install`....