The script currently writes a package.json with a dependency on parse-server version 2.2. This should probably always be automatically updated to the latest version using some CI magic.
The script uses double square brackets, which are a non-standard extension to `[]`. Some shells (e.g. dash, the default shell under Debian) do not support double square brackets. The shebang line should reflect that.
* feat: add allowHeaders to Options
This allows developers to use custom headers in their API requests, and they will be accepted by their mounted app.
* refactor: convert allowCrossDomain to generator to add appId in scope
This is necessary as the middleware may run in OPTIONS request that do not contain the appId within the header.
* chore: update Definitions and docs
* fix: update test to use new allowCrossDomain params
* chore: add tests for allowCustomDomain middleware re: allowHeadrs
* Fix: aggregate not matching null values
* Exclude Postgres from this new test - it does not even support and is not working correctly - should be addressed separately
* chore(package): update apollo-link-http to version 1.5.16
* chore(package): update apollo-link-ws to version 1.0.19
* chore(package): update apollo-link to version 1.2.13
* chore(package): update lockfile package-lock.json
* chore(package): update @babel/cli to version 7.6.0
* chore(package): update @babel/core to version 7.6.0
* chore(package): update @babel/preset-env to version 7.6.0
* chore(package): update lockfile package-lock.json
* refactor(GraphQL): Pointer constraint input type as ID
Redefines the Pointer constraint input type from a custom scalar to
a simple ID.
* fix: PR review requested changes
The method `createSessionToken` in RestWrite.js is assumed to always return a promise (see Line 961 in `handleFollowUp`) - this was throwing an error `cannot read 'then' of undefined`. This simply one word change fixes that error.
* Remove nested operations
* Improve error log
* Fix bug schema to load
* Fix ParseGraphQLSchema tests
* Fix tests
* Fix failing tests
* First verstion not complete of create class mutation
* Fix bug caused by circular dependency
* Renaming files
* Schema types should be loaded before parse classes
* Fix tests
* Create class mutation boilerplate
* Improve CreateClassSchemaInput fields names
* Remove fields
* Pointer and relation fields
* Improve pointer default type
* Class type
* Create class mutation resolver
* Schema field transformers
* Class types transformations
* First test
* Numbers test
* Boolean tests
* Date test
* Fix some get tests
* Test for created at and updated at
* File tests
* Test for objects
* Renaming reducerFabric to reducerGenerator
* Changing get tests for file and object
* Object composed queries test
* Array test
* Null field test
* Bytes test
* Geo Point test
* Polygons tests
* Remove create generic mutation
* Fix tests
* Create class test - isRequired and defaultValue will be added back later
* Enforce master key
* Fix tests
* Duplicated field test
* updateClass mutation
* Remove update generic mutation tests
* Remove update generic mutation
* deleteClass mutation
* Remove delete generic mutation tests
* Remove delete generic mutation
* class query
* Classes query
* Remove get generic query from tests
* Remove remaining generic operations and fix tests
* Fix last test
* Try to fix redis tests
* Fix postgres tests
* Update objectsMutations and objectsQueries files locations
* Rename classSchema files to schema files
* Rename ClassObject to ParseObject
* Fix names and paths
* Still some wrong names
* refactor(GraphQL): Rename objectId to id
Renames `objectId` to `id` for the GraphQL API. Queries, mutations,
custom and generic types were updated.
Removes `RELATION_INPUT` and `POINTER_INPUT`. Now the user just need
to provide the ID of the object to link.
* fix: Column "id" not found on Postgres
* fix: Avoid deleting Parse class objectId
* fix: Undo objectId removal on mutations
* fix: Handle generic mutation id
