graphql-engine/server/src-lib/Hasura/RQL/DDL/Schema/Cache/Fields.hs
Vamshi Surabhi e8e4f30dd6 server: support remote relationships on SQL Server and BigQuery (#1497)
Remote relationships are now supported on SQL Server and BigQuery. The major change though is the re-architecture of remote join execution logic. Prior to this PR, each backend is responsible for processing the remote relationships that are part of their AST.

This is not ideal as there is nothing specific about a remote join's execution that ties it to a backend. The only backend specific part is whether or not the specification of the remote relationship is valid (i.e, we'll need to validate whether the scalars are compatible).

The approach now changes to this:

1. Before delegating the AST to the backend, we traverse the AST, collect all the remote joins while modifying the AST to add necessary join fields where needed.

1. Once the remote joins are collected from the AST, the database call is made to fetch the response. The necessary data for the remote join(s) is collected from the database's response and one or more remote schema calls are constructed as necessary.

1. The remote schema calls are then executed and the data from the database and from the remote schemas is joined to produce the final response.

### Known issues

1. Ideally the traversal of the IR to collect remote joins should return an AST which does not include remote join fields. This operation can be type safe but isn't taken up as part of the PR.

1. There is a lot of code duplication between `Transport/HTTP.hs` and `Transport/Websocket.hs` which needs to be fixed ASAP. This too hasn't been taken up by this PR.

1. The type which represents the execution plan is only modified to handle our current remote joins and as such it will have to be changed to accommodate general remote joins.

1. Use of lenses would have reduced the boilerplate code to collect remote joins from the base AST.

1. The current remote join logic assumes that the join columns of a remote relationship appear with their names in the database response. This however is incorrect as they could be aliased. This can be taken up by anyone, I've left a comment in the code.

### Notes to the reviewers

I think it is best reviewed commit by commit.

1. The first one is very straight forward.

1. The second one refactors the remote join execution logic but other than moving things around, it doesn't change the user facing functionality.  This moves Postgres specific parts to `Backends/Postgres` module from `Execute`. Some IR related code to `Hasura.RQL.IR` module.  Simplifies various type class function signatures as a backend doesn't have to handle remote joins anymore

1. The third one fixes partial case matches that for some weird reason weren't shown as warnings before this refactor

1. The fourth one generalizes the validation logic of remote relationships and implements `scalarTypeGraphQLName` function on SQL Server and BigQuery which is used by the validation logic. This enables remote relationships on BigQuery and SQL Server.

https://github.com/hasura/graphql-engine-mono/pull/1497

GitOrigin-RevId: 77dd8eed326602b16e9a8496f52f46d22b795598
2021-06-11 03:27:39 +00:00

276 lines
12 KiB
Haskell
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

{-# LANGUAGE Arrows #-}
module Hasura.RQL.DDL.Schema.Cache.Fields
(addNonColumnFields)
where
import Hasura.Prelude
import qualified Data.HashMap.Strict.Extended as M
import qualified Data.HashSet as HS
import qualified Data.Sequence as Seq
import qualified Language.GraphQL.Draft.Syntax as G
import Control.Arrow.Extended
import Control.Lens (_3, _4, (^.))
import Data.Aeson
import Data.Text.Extended
import qualified Hasura.Incremental as Inc
import qualified Hasura.SQL.AnyBackend as AB
import Hasura.Base.Error
import Hasura.RQL.DDL.ComputedField
import Hasura.RQL.DDL.Relationship
import Hasura.RQL.DDL.RemoteRelationship
import Hasura.RQL.DDL.Schema.Cache.Common
import Hasura.RQL.DDL.Schema.Function
import Hasura.RQL.Types
addNonColumnFields
:: forall b arr m
. ( ArrowChoice arr, Inc.ArrowDistribute arr, ArrowWriter (Seq CollectedInfo) arr
, ArrowKleisli m arr, MonadError QErr m, BackendMetadata b)
=> ( SourceName
, HashMap (TableName b) (TableCoreInfoG b (ColumnInfo b) (ColumnInfo b))
, FieldInfoMap (ColumnInfo b)
, RemoteSchemaMap
, DBFunctionsMetadata b
, NonColumnTableInputs b
) `arr` FieldInfoMap (FieldInfo b)
addNonColumnFields = proc ( source
, rawTableInfo
, columns
, remoteSchemaMap
, pgFunctions
, NonColumnTableInputs{..}
) -> do
objectRelationshipInfos
<- buildInfoMapPreservingMetadata
(_rdName . (^. _4))
(\(s, _, t, c) -> mkRelationshipMetadataObject @b ObjRel (s,t,c))
buildObjectRelationship
-< (_tciForeignKeys <$> rawTableInfo, map (source, columns, _nctiTable,) _nctiObjectRelationships)
arrayRelationshipInfos
<- buildInfoMapPreservingMetadata
(_rdName . (^. _3))
(mkRelationshipMetadataObject @b ArrRel)
buildArrayRelationship
-< (_tciForeignKeys <$> rawTableInfo, map (source, _nctiTable,) _nctiArrayRelationships)
let relationshipInfos = objectRelationshipInfos <> arrayRelationshipInfos
computedFieldInfos
<- buildInfoMapPreservingMetadata
(_cfmName . (^. _4))
(\(s, _, t, c) -> mkComputedFieldMetadataObject (s, t, c))
buildComputedField
-< (HS.fromList $ M.keys rawTableInfo, map (source, pgFunctions, _nctiTable,) _nctiComputedFields)
rawRemoteRelationshipInfos
<- buildInfoMapPreservingMetadata
(_rrmName . (^. _3))
(mkRemoteRelationshipMetadataObject @b)
buildRemoteRelationship
-< ((M.elems columns, remoteSchemaMap), map (source, _nctiTable,) _nctiRemoteRelationships)
let relationshipFields = mapKeys fromRel relationshipInfos
computedFieldFields = mapKeys fromComputedField computedFieldInfos
remoteRelationshipFields = mapKeys fromRemoteRelationship rawRemoteRelationshipInfos
-- First, check for conflicts between non-column fields, since we can raise a better error
-- message in terms of the two metadata objects that define them.
(align relationshipFields computedFieldFields >- returnA)
>-> (| Inc.keyed (\fieldName fields -> (fieldName, fields) >- noFieldConflicts FIRelationship FIComputedField) |)
-- Second, align with remote relationship fields
>-> (\fields -> align (M.catMaybes fields) remoteRelationshipFields >- returnA)
>-> (| Inc.keyed (\fieldName fields -> (fieldName, fields) >- noFieldConflicts id FIRemoteRelationship) |)
-- Next, check for conflicts with custom field names. This is easiest to do before merging with
-- the column info itself because we have access to the information separately, and custom field
-- names are not currently stored as a separate map (but maybe should be!).
>-> (\fields -> (columns, M.catMaybes fields) >- (noCustomFieldConflicts))
-- Finally, check for conflicts with the columns themselves.
>-> (\fields -> align columns (M.catMaybes fields) >- returnA)
>-> (| Inc.keyed (\_ fields -> fields >- noColumnConflicts) |)
where
noFieldConflicts this that = proc (fieldName, fields) -> case fields of
This (thisField, metadata) -> returnA -< Just (this thisField, metadata)
That (thatField, metadata) -> returnA -< Just (that thatField, metadata)
These (_, thisMetadata) (_, thatMetadata) -> do
tellA -< Seq.singleton $ CIInconsistency $ ConflictingObjects
("conflicting definitions for field " <>> fieldName)
[thisMetadata, thatMetadata]
returnA -< Nothing
noCustomFieldConflicts = proc (columns, nonColumnFields) -> do
let columnsByGQLName = mapFromL pgiName $ M.elems columns
(| Inc.keyed (\_ (fieldInfo, metadata) ->
(| withRecordInconsistency (do
(| traverseA_ (\fieldGQLName -> case M.lookup fieldGQLName columnsByGQLName of
-- Only raise an error if the GQL name isnt the same as the Postgres column name.
-- If they are the same, `noColumnConflicts` will catch it, and it will produce a
-- more useful error message.
Just columnInfo | toTxt (pgiColumn columnInfo) /= G.unName fieldGQLName ->
throwA -< err400 AlreadyExists
$ "field definition conflicts with custom field name for postgres column "
<>> pgiColumn columnInfo
_ -> returnA -< ())
|) (fieldInfoGraphQLNames fieldInfo)
returnA -< (fieldInfo, metadata))
|) metadata)
|) nonColumnFields
noColumnConflicts = proc fields -> case fields of
This columnInfo -> returnA -< FIColumn columnInfo
That (fieldInfo, _) -> returnA -< fieldInfo
These columnInfo (_, fieldMetadata) -> do
recordInconsistency -< ((Nothing, fieldMetadata), "field definition conflicts with postgres column")
returnA -< FIColumn columnInfo
mkRelationshipMetadataObject
:: forall b a
. (ToJSON a, Backend b)
=> RelType
-> (SourceName, TableName b, RelDef a)
-> MetadataObject
mkRelationshipMetadataObject relType (source, table, relDef) =
let objectId = MOSourceObjId source
$ AB.mkAnyBackend
$ SMOTableObj @b table
$ MTORel (_rdName relDef) relType
in MetadataObject objectId $ toJSON $ WithTable @b source table relDef
buildObjectRelationship
:: ( ArrowChoice arr
, ArrowWriter (Seq CollectedInfo) arr
, Backend b
)
=> ( HashMap (TableName b) (HashSet (ForeignKey b))
, ( SourceName
, FieldInfoMap (ColumnInfo b)
, TableName b
, ObjRelDef b
)
) `arr` Maybe (RelInfo b)
buildObjectRelationship = proc (fkeysMap, (source, columns, table, relDef)) -> do
let buildRelInfo def = objRelP2Setup source table fkeysMap def columns
buildRelationship -< (source, table, buildRelInfo, ObjRel, relDef)
buildArrayRelationship
:: ( ArrowChoice arr
, ArrowWriter (Seq CollectedInfo) arr
, Backend b
)
=> ( HashMap (TableName b) (HashSet (ForeignKey b))
, ( SourceName
, TableName b
, ArrRelDef b
)
) `arr` Maybe (RelInfo b)
buildArrayRelationship = proc (fkeysMap, (source, table, relDef)) -> do
let buildRelInfo def = arrRelP2Setup fkeysMap source table def
buildRelationship -< (source, table, buildRelInfo, ArrRel, relDef)
buildRelationship
:: forall b arr a
. ( ArrowChoice arr
, ArrowWriter (Seq CollectedInfo) arr
, ToJSON a
, Backend b
)
=> ( SourceName
, TableName b
, RelDef a -> Either QErr (RelInfo b, [SchemaDependency])
, RelType
, RelDef a
) `arr` Maybe (RelInfo b)
buildRelationship = proc (source, table, buildRelInfo, relType, relDef) -> do
let relName = _rdName relDef
metadataObject = mkRelationshipMetadataObject @b relType (source, table, relDef)
schemaObject = SOSourceObj source
$ AB.mkAnyBackend
$ SOITableObj @b table
$ TORel relName
addRelationshipContext e = "in relationship " <> relName <<> ": " <> e
(| withRecordInconsistency (
(| modifyErrA (do
(info, dependencies) <- liftEitherA -< buildRelInfo relDef
recordDependencies -< (metadataObject, schemaObject, dependencies)
returnA -< info)
|) (addTableContext @b table . addRelationshipContext))
|) metadataObject
mkComputedFieldMetadataObject
:: forall b
. (Backend b)
=> (SourceName, TableName b, ComputedFieldMetadata b)
-> MetadataObject
mkComputedFieldMetadataObject (source, table, ComputedFieldMetadata{..}) =
let objectId = MOSourceObjId source
$ AB.mkAnyBackend
$ SMOTableObj @b table
$ MTOComputedField _cfmName
definition = AddComputedField source table _cfmName _cfmDefinition _cfmComment
in MetadataObject objectId (toJSON definition)
buildComputedField
:: forall b arr m
. ( ArrowChoice arr, ArrowWriter (Seq CollectedInfo) arr
, ArrowKleisli m arr, MonadError QErr m, BackendMetadata b )
=> ( HashSet (TableName b)
, (SourceName, DBFunctionsMetadata b, TableName b, ComputedFieldMetadata b)
) `arr` Maybe (ComputedFieldInfo b)
buildComputedField = proc (trackedTableNames, (source, pgFunctions, table, cf@ComputedFieldMetadata{..})) -> do
let addComputedFieldContext e = "in computed field " <> _cfmName <<> ": " <> e
function = _cfdFunction _cfmDefinition
funcDefs = fromMaybe [] $ M.lookup function pgFunctions
(| withRecordInconsistency (
(| modifyErrA (do
rawfi <- bindErrorA -< handleMultipleFunctions @b (_cfdFunction _cfmDefinition) funcDefs
bindErrorA -< buildComputedFieldInfo trackedTableNames table _cfmName _cfmDefinition rawfi _cfmComment)
|) (addTableContext @b table . addComputedFieldContext))
|) (mkComputedFieldMetadataObject (source, table, cf))
mkRemoteRelationshipMetadataObject
:: forall b
. Backend b
=> (SourceName, TableName b, RemoteRelationshipMetadata)
-> MetadataObject
mkRemoteRelationshipMetadataObject (source, table, RemoteRelationshipMetadata{..}) =
let objectId = MOSourceObjId source
$ AB.mkAnyBackend
$ SMOTableObj @b table
$ MTORemoteRelationship _rrmName
RemoteRelationshipDef{..} = _rrmDefinition
in MetadataObject objectId $ toJSON $
RemoteRelationship @b _rrmName source table _rrdHasuraFields _rrdRemoteSchema _rrdRemoteField
buildRemoteRelationship
:: forall b arr m
. ( ArrowChoice arr, ArrowWriter (Seq CollectedInfo) arr
, ArrowKleisli m arr, MonadError QErr m, BackendMetadata b)
=> ( ([ColumnInfo b], RemoteSchemaMap)
, (SourceName, TableName b, RemoteRelationshipMetadata)
) `arr` Maybe (RemoteFieldInfo b)
buildRemoteRelationship = proc ( (pgColumns, remoteSchemaMap)
, (source, table, rrm@RemoteRelationshipMetadata{..})
) -> do
let metadataObject = mkRemoteRelationshipMetadataObject @b (source, table, rrm)
schemaObj = SOSourceObj source
$ AB.mkAnyBackend
$ SOITableObj @b table
$ TORemoteRel _rrmName
addRemoteRelationshipContext e = "in remote relationship" <> _rrmName <<> ": " <> e
RemoteRelationshipDef{..} = _rrmDefinition
remoteRelationship = RemoteRelationship @b _rrmName source table _rrdHasuraFields
_rrdRemoteSchema _rrdRemoteField
(| withRecordInconsistency (
(| modifyErrA (do
(remoteField, dependencies) <- bindErrorA -< buildRemoteFieldInfo remoteRelationship pgColumns remoteSchemaMap
recordDependencies -< (metadataObject, schemaObj, dependencies)
returnA -< remoteField)
|)(addTableContext @b table . addRemoteRelationshipContext))
|) metadataObject