2022-01-18 17:53:44 +03:00
|
|
|
|
{-# LANGUAGE ApplicativeDo #-}
|
2022-05-26 14:54:30 +03:00
|
|
|
|
{-# LANGUAGE TemplateHaskell #-}
|
2021-04-22 00:44:37 +03:00
|
|
|
|
{-# LANGUAGE UndecidableInstances #-}
|
2021-09-24 01:56:37 +03:00
|
|
|
|
{-# OPTIONS_GHC -fno-warn-orphans #-}
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
2022-02-08 12:24:34 +03:00
|
|
|
|
-- | Postgres Instances Schema
|
|
|
|
|
--
|
|
|
|
|
-- Defines a 'Hasura.GraphQL.Schema.Backend.BackendSchema' type class instance for Postgres.
|
2021-05-21 05:46:58 +03:00
|
|
|
|
module Hasura.Backends.Postgres.Instances.Schema
|
2022-05-26 17:05:13 +03:00
|
|
|
|
( updateOperators,
|
2021-09-24 01:56:37 +03:00
|
|
|
|
)
|
|
|
|
|
where
|
|
|
|
|
|
|
|
|
|
import Data.Aeson qualified as J
|
2022-06-08 18:31:28 +03:00
|
|
|
|
import Data.Aeson.Key qualified as K
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Data.Has
|
|
|
|
|
import Data.HashMap.Strict qualified as Map
|
|
|
|
|
import Data.HashMap.Strict.Extended qualified as M
|
|
|
|
|
import Data.List.NonEmpty qualified as NE
|
|
|
|
|
import Data.Parser.JSONPath
|
|
|
|
|
import Data.Text qualified as T
|
2022-05-26 14:54:30 +03:00
|
|
|
|
import Data.Text.Casing qualified as C
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Data.Text.Extended
|
2021-11-18 21:02:58 +03:00
|
|
|
|
import Hasura.Backends.Postgres.SQL.DML as PG hiding (CountType, incOp)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.Backends.Postgres.SQL.Types as PG hiding (FunctionName, TableName)
|
|
|
|
|
import Hasura.Backends.Postgres.SQL.Value as PG
|
2021-12-15 20:07:21 +03:00
|
|
|
|
import Hasura.Backends.Postgres.Schema.OnConflict
|
2022-06-29 19:03:25 +03:00
|
|
|
|
import Hasura.Backends.Postgres.Schema.Select
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.Backends.Postgres.Types.BoolExp
|
|
|
|
|
import Hasura.Backends.Postgres.Types.Column
|
2021-12-09 12:05:42 +03:00
|
|
|
|
import Hasura.Backends.Postgres.Types.Insert as PGIR
|
2021-11-18 21:02:58 +03:00
|
|
|
|
import Hasura.Backends.Postgres.Types.Update as PGIR
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.Base.Error
|
|
|
|
|
import Hasura.GraphQL.Schema.Backend
|
|
|
|
|
( BackendSchema,
|
2022-06-30 18:22:19 +03:00
|
|
|
|
BackendTableSelectSchema,
|
2021-09-24 01:56:37 +03:00
|
|
|
|
ComparisonExp,
|
|
|
|
|
MonadBuildSchema,
|
|
|
|
|
)
|
|
|
|
|
import Hasura.GraphQL.Schema.Backend qualified as BS
|
|
|
|
|
import Hasura.GraphQL.Schema.BoolExp
|
|
|
|
|
import Hasura.GraphQL.Schema.Build qualified as GSB
|
|
|
|
|
import Hasura.GraphQL.Schema.Common
|
2021-10-01 15:52:19 +03:00
|
|
|
|
import Hasura.GraphQL.Schema.Mutation qualified as GSB
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
import Hasura.GraphQL.Schema.Parser
|
|
|
|
|
( Definition,
|
|
|
|
|
FieldParser,
|
|
|
|
|
InputFieldsParser,
|
|
|
|
|
Kind (..),
|
|
|
|
|
MonadParse,
|
|
|
|
|
MonadSchema,
|
|
|
|
|
Parser,
|
|
|
|
|
memoize,
|
|
|
|
|
)
|
|
|
|
|
import Hasura.GraphQL.Schema.Parser qualified as P
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.GraphQL.Schema.Select
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
import Hasura.GraphQL.Schema.Table
|
2021-11-26 16:47:12 +03:00
|
|
|
|
import Hasura.GraphQL.Schema.Update qualified as SU
|
2022-06-23 12:14:24 +03:00
|
|
|
|
import Hasura.Name qualified as Name
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.Prelude
|
2021-12-09 12:05:42 +03:00
|
|
|
|
import Hasura.RQL.IR.BoolExp
|
|
|
|
|
import Hasura.RQL.IR.Root (RemoteRelationshipField)
|
|
|
|
|
import Hasura.RQL.IR.Select
|
|
|
|
|
( QueryDB (QDBConnection),
|
|
|
|
|
)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.RQL.IR.Select qualified as IR
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
import Hasura.RQL.IR.Update qualified as IR
|
2022-05-31 01:07:02 +03:00
|
|
|
|
import Hasura.RQL.IR.Value qualified as IR
|
2021-12-09 12:05:42 +03:00
|
|
|
|
import Hasura.RQL.Types.Backend (Backend (..))
|
|
|
|
|
import Hasura.RQL.Types.Column
|
|
|
|
|
import Hasura.RQL.Types.Function (FunctionInfo)
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
import Hasura.RQL.Types.Source
|
2022-05-26 14:54:30 +03:00
|
|
|
|
import Hasura.RQL.Types.SourceCustomization
|
Clean Relay's code, break schema cycles, introduce Node ID V2
## Motivation
This PR rewrites most of Relay to achieve the following:
- ~~fix a bug in which the same node id could refer to two different tables in the schema~~
- remove one of the few remaining uses of the source cache in the schema building code
In doing so, it also:
- simplifies the `BackendSchema` class by removing `node` from it,
- makes it much easier for other backends to support Relay,
- documents, re-organizes, and clarifies the code.
## Description
This PR introduces a new `NodeId` version ~~, and adapts the Postgres code to always generate this V2 version~~. This new id contains the source name, in addition to the table name, in order to disambiguate similar table names across different sources (which is now possible with source customization). In doing so, it now explicitly handles that case for V1 node ids, and returns an explicit error message instead of running the risk of _silently returning the wrong information_.
Furthermore, it adapts `nodeField` to support multiple backends; most of the code was trivial to generalize, and as a result it lowers the cost of entry for other backends, that now only need to support `AFNodeId` in their translation layer.
Finally, it removes one more cycle in the schema building code, by using the same trick we used for remote relationships instead of using the memoization trick of #4576.
## Remaining work
- ~~[ ]write a Changelog entry~~
- ~~[x] adapt all tests that were asserting on an old node id~~
## Future work
This PR was adapted from its original form to avoid a breaking change: while it introduces a Node ID V2, we keep generating V1 IDs and the parser rejects V2 IDs. It will be easy to make the switch at a later data in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4593
GitOrigin-RevId: 88e5cb91e8b0646900547fa8c7c0e1463de267a1
2022-06-07 16:35:26 +03:00
|
|
|
|
import Hasura.RQL.Types.Table (RolePermInfo (..), TableInfo, UpdPermInfo)
|
2021-12-09 12:05:42 +03:00
|
|
|
|
import Hasura.SQL.Backend (BackendType (Postgres), PostgresKind (Citus, Vanilla))
|
2022-06-30 18:22:19 +03:00
|
|
|
|
import Hasura.SQL.Tag (HasTag)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
import Hasura.SQL.Types
|
|
|
|
|
import Language.GraphQL.Draft.Syntax qualified as G
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
|
|
|
|
----------------------------------------------------------------
|
|
|
|
|
-- BackendSchema instance
|
|
|
|
|
|
2021-04-22 00:44:37 +03:00
|
|
|
|
-- | This class is an implementation detail of 'BackendSchema'.
|
|
|
|
|
-- Some functions of 'BackendSchema' differ across different Postgres "kinds",
|
|
|
|
|
-- or call to functions (such as those related to Relay) that have not been
|
|
|
|
|
-- generalized to all kinds of Postgres and still explicitly work on Vanilla
|
2021-11-18 21:02:58 +03:00
|
|
|
|
-- Postgres. This class allows each "kind" to specify its own specific
|
2021-04-22 00:44:37 +03:00
|
|
|
|
-- implementation. All common code is directly part of `BackendSchema`.
|
2021-11-18 21:02:58 +03:00
|
|
|
|
--
|
|
|
|
|
-- Note: Users shouldn't ever put this as a constraint. Use `BackendSchema
|
|
|
|
|
-- ('Postgres pgKind)` instead.
|
2021-04-22 00:44:37 +03:00
|
|
|
|
class PostgresSchema (pgKind :: PostgresKind) where
|
2021-09-24 01:56:37 +03:00
|
|
|
|
pgkBuildTableRelayQueryFields ::
|
|
|
|
|
BS.MonadBuildSchema ('Postgres pgKind) r m n =>
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
SourceInfo ('Postgres pgKind) ->
|
2021-09-24 01:56:37 +03:00
|
|
|
|
TableName ('Postgres pgKind) ->
|
|
|
|
|
TableInfo ('Postgres pgKind) ->
|
2022-05-26 14:54:30 +03:00
|
|
|
|
C.GQLNameIdentifier ->
|
2021-09-24 01:56:37 +03:00
|
|
|
|
NESeq (ColumnInfo ('Postgres pgKind)) ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m [FieldParser n (QueryDB ('Postgres pgKind) (RemoteRelationshipField IR.UnpreparedValue) (IR.UnpreparedValue ('Postgres pgKind)))]
|
2021-09-24 01:56:37 +03:00
|
|
|
|
pgkBuildFunctionRelayQueryFields ::
|
|
|
|
|
BS.MonadBuildSchema ('Postgres pgKind) r m n =>
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
SourceInfo ('Postgres pgKind) ->
|
2021-09-24 01:56:37 +03:00
|
|
|
|
FunctionName ('Postgres pgKind) ->
|
|
|
|
|
FunctionInfo ('Postgres pgKind) ->
|
|
|
|
|
TableName ('Postgres pgKind) ->
|
|
|
|
|
NESeq (ColumnInfo ('Postgres pgKind)) ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m [FieldParser n (QueryDB ('Postgres pgKind) (RemoteRelationshipField IR.UnpreparedValue) (IR.UnpreparedValue ('Postgres pgKind)))]
|
2021-09-24 01:56:37 +03:00
|
|
|
|
pgkRelayExtension ::
|
|
|
|
|
Maybe (XRelay ('Postgres pgKind))
|
2021-04-22 00:44:37 +03:00
|
|
|
|
|
|
|
|
|
instance PostgresSchema 'Vanilla where
|
2021-09-24 01:56:37 +03:00
|
|
|
|
pgkBuildTableRelayQueryFields = buildTableRelayQueryFields
|
2021-04-22 00:44:37 +03:00
|
|
|
|
pgkBuildFunctionRelayQueryFields = buildFunctionRelayQueryFields
|
2021-06-09 16:02:15 +03:00
|
|
|
|
pgkRelayExtension = Just ()
|
2021-04-22 00:44:37 +03:00
|
|
|
|
|
2021-05-21 05:46:58 +03:00
|
|
|
|
instance PostgresSchema 'Citus where
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
pgkBuildTableRelayQueryFields _ _ _ _ _ = pure []
|
|
|
|
|
pgkBuildFunctionRelayQueryFields _ _ _ _ _ = pure []
|
2021-06-09 16:02:15 +03:00
|
|
|
|
pgkRelayExtension = Nothing
|
2021-05-21 05:46:58 +03:00
|
|
|
|
|
|
|
|
|
-- postgres schema
|
2021-04-22 00:44:37 +03:00
|
|
|
|
|
2022-06-30 18:22:19 +03:00
|
|
|
|
instance
|
|
|
|
|
( PostgresSchema pgKind,
|
|
|
|
|
Backend ('Postgres pgKind),
|
|
|
|
|
HasTag ('Postgres pgKind)
|
|
|
|
|
) =>
|
|
|
|
|
BS.BackendTableSelectSchema ('Postgres pgKind)
|
|
|
|
|
where
|
|
|
|
|
tableArguments = defaultTableArgs
|
|
|
|
|
selectTable = defaultSelectTable
|
|
|
|
|
selectTableAggregate = defaultSelectTableAggregate
|
|
|
|
|
tableSelectionSet = defaultTableSelectionSet
|
|
|
|
|
|
2021-04-22 00:44:37 +03:00
|
|
|
|
instance
|
2021-11-08 21:11:44 +03:00
|
|
|
|
( Backend ('Postgres pgKind),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
PostgresSchema pgKind
|
|
|
|
|
) =>
|
|
|
|
|
BackendSchema ('Postgres pgKind)
|
|
|
|
|
where
|
2021-02-23 20:37:27 +03:00
|
|
|
|
-- top level parsers
|
2022-06-07 08:32:08 +03:00
|
|
|
|
buildTableQueryAndSubscriptionFields = GSB.buildTableQueryAndSubscriptionFields
|
2021-09-24 01:56:37 +03:00
|
|
|
|
buildTableRelayQueryFields = pgkBuildTableRelayQueryFields
|
2022-04-22 22:53:12 +03:00
|
|
|
|
buildTableStreamingSubscriptionFields = GSB.buildTableStreamingSubscriptionFields
|
2021-12-09 12:05:42 +03:00
|
|
|
|
buildTableInsertMutationFields = GSB.buildTableInsertMutationFields backendInsertParser
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
buildTableUpdateMutationFields = pgkBuildTableUpdateMutationFields
|
2021-02-23 20:37:27 +03:00
|
|
|
|
buildTableDeleteMutationFields = GSB.buildTableDeleteMutationFields
|
2022-06-29 19:03:25 +03:00
|
|
|
|
buildFunctionQueryFields = buildFunctionQueryFieldsPG
|
2021-09-24 01:56:37 +03:00
|
|
|
|
buildFunctionRelayQueryFields = pgkBuildFunctionRelayQueryFields
|
2022-06-29 19:03:25 +03:00
|
|
|
|
buildFunctionMutationFields = buildFunctionMutationFieldsPG
|
2021-06-15 18:53:20 +03:00
|
|
|
|
|
2021-12-09 12:05:42 +03:00
|
|
|
|
mkRelationshipParser = GSB.mkDefaultRelationshipParser backendInsertParser ()
|
2021-06-15 18:53:20 +03:00
|
|
|
|
|
2021-02-23 20:37:27 +03:00
|
|
|
|
-- backend extensions
|
2021-09-24 01:56:37 +03:00
|
|
|
|
relayExtension = pgkRelayExtension @pgKind
|
|
|
|
|
nodesAggExtension = Just ()
|
2022-04-22 22:53:12 +03:00
|
|
|
|
streamSubscriptionExtension = Just ()
|
2021-06-15 18:53:20 +03:00
|
|
|
|
|
2021-02-23 20:37:27 +03:00
|
|
|
|
-- indivdual components
|
2021-09-24 01:56:37 +03:00
|
|
|
|
columnParser = columnParser
|
2022-05-03 11:58:56 +03:00
|
|
|
|
scalarSelectionArgumentsParser = pgScalarSelectionArgumentsParser
|
2022-05-26 14:54:30 +03:00
|
|
|
|
|
|
|
|
|
-- NOTE: We don't use @orderByOperators@ directly as this will cause memory
|
|
|
|
|
-- growth, instead we use separate functions, according to @jberryman on the
|
|
|
|
|
-- memory growth, "This is turning a CAF Into a function, And the output is
|
|
|
|
|
-- likely no longer going to be shared even for the same arguments, and even
|
|
|
|
|
-- though the domain is extremely small (just HasuraCase or GraphqlCase)."
|
2022-06-10 06:59:00 +03:00
|
|
|
|
orderByOperators _sourceInfo = \case
|
2022-05-26 14:54:30 +03:00
|
|
|
|
HasuraCase -> orderByOperatorsHasuraCase
|
|
|
|
|
GraphqlCase -> orderByOperatorsGraphqlCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
comparisonExps = comparisonExps
|
2022-01-18 17:53:44 +03:00
|
|
|
|
countTypeInput = countTypeInput
|
2021-02-23 20:37:27 +03:00
|
|
|
|
aggregateOrderByCountType = PG.PGInteger
|
2021-09-24 01:56:37 +03:00
|
|
|
|
computedField = computedFieldPG
|
2021-06-15 18:53:20 +03:00
|
|
|
|
|
2021-12-09 12:05:42 +03:00
|
|
|
|
backendInsertParser ::
|
|
|
|
|
forall pgKind m r n.
|
|
|
|
|
MonadBuildSchema ('Postgres pgKind) r m n =>
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
SourceInfo ('Postgres pgKind) ->
|
2021-12-09 12:05:42 +03:00
|
|
|
|
TableInfo ('Postgres pgKind) ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (InputFieldsParser n (PGIR.BackendInsert pgKind (IR.UnpreparedValue ('Postgres pgKind))))
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
backendInsertParser sourceName tableInfo =
|
|
|
|
|
fmap BackendInsert <$> onConflictFieldParser sourceName tableInfo
|
2021-10-01 15:52:19 +03:00
|
|
|
|
|
2021-02-23 20:37:27 +03:00
|
|
|
|
----------------------------------------------------------------
|
|
|
|
|
-- Top level parsers
|
|
|
|
|
|
2021-09-24 01:56:37 +03:00
|
|
|
|
buildTableRelayQueryFields ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
MonadBuildSchema ('Postgres pgKind) r m n =>
|
2022-06-30 18:22:19 +03:00
|
|
|
|
BackendTableSelectSchema ('Postgres pgKind) =>
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
SourceInfo ('Postgres pgKind) ->
|
2021-09-24 01:56:37 +03:00
|
|
|
|
TableName ('Postgres pgKind) ->
|
|
|
|
|
TableInfo ('Postgres pgKind) ->
|
2022-05-26 14:54:30 +03:00
|
|
|
|
C.GQLNameIdentifier ->
|
2021-09-24 01:56:37 +03:00
|
|
|
|
NESeq (ColumnInfo ('Postgres pgKind)) ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m [FieldParser n (QueryDB ('Postgres pgKind) (RemoteRelationshipField IR.UnpreparedValue) (IR.UnpreparedValue ('Postgres pgKind)))]
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
buildTableRelayQueryFields sourceName tableName tableInfo gqlName pkeyColumns = do
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase <- asks getter
|
2021-11-26 00:07:53 +03:00
|
|
|
|
let fieldDesc = Just $ G.Description $ "fetch data from the table: " <>> tableName
|
2022-05-26 14:54:30 +03:00
|
|
|
|
rootFieldName <- mkRootFieldName $ applyFieldNameCaseIdentifier tCase (mkRelayConnectionField gqlName)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
fmap afold $
|
2021-12-09 12:05:42 +03:00
|
|
|
|
optionalFieldParser QDBConnection $
|
2022-05-26 14:54:30 +03:00
|
|
|
|
selectTableConnection sourceName tableInfo rootFieldName fieldDesc pkeyColumns
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
|
|
|
|
|
pgkBuildTableUpdateMutationFields ::
|
|
|
|
|
MonadBuildSchema ('Postgres pgKind) r m n =>
|
2022-06-30 18:22:19 +03:00
|
|
|
|
BackendTableSelectSchema ('Postgres pgKind) =>
|
2022-05-31 17:41:09 +03:00
|
|
|
|
Scenario ->
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
-- | The source that the table lives in
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
SourceInfo ('Postgres pgKind) ->
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
-- | The name of the table being acted on
|
|
|
|
|
TableName ('Postgres pgKind) ->
|
|
|
|
|
-- | table info
|
|
|
|
|
TableInfo ('Postgres pgKind) ->
|
|
|
|
|
-- | field display name
|
2022-05-26 14:54:30 +03:00
|
|
|
|
C.GQLNameIdentifier ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m [FieldParser n (IR.AnnotatedUpdateG ('Postgres pgKind) (RemoteRelationshipField IR.UnpreparedValue) (IR.UnpreparedValue ('Postgres pgKind)))]
|
2022-05-31 17:41:09 +03:00
|
|
|
|
pgkBuildTableUpdateMutationFields scenario sourceName tableName tableInfo gqlName =
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
concat . maybeToList <$> runMaybeT do
|
2022-04-27 15:16:09 +03:00
|
|
|
|
updatePerms <- MaybeT $ _permUpd <$> tablePermissions tableInfo
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
lift $
|
|
|
|
|
GSB.buildTableUpdateMutationFields
|
|
|
|
|
-- TODO: https://github.com/hasura/graphql-engine-mono/issues/2955
|
|
|
|
|
(\ti -> fmap BackendUpdate <$> updateOperators ti updatePerms)
|
2022-05-31 17:41:09 +03:00
|
|
|
|
scenario
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
sourceName
|
|
|
|
|
tableName
|
|
|
|
|
tableInfo
|
|
|
|
|
gqlName
|
2021-09-24 01:56:37 +03:00
|
|
|
|
|
|
|
|
|
buildFunctionRelayQueryFields ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
MonadBuildSchema ('Postgres pgKind) r m n =>
|
2022-06-30 18:22:19 +03:00
|
|
|
|
BackendTableSelectSchema ('Postgres pgKind) =>
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
SourceInfo ('Postgres pgKind) ->
|
2021-09-24 01:56:37 +03:00
|
|
|
|
FunctionName ('Postgres pgKind) ->
|
|
|
|
|
FunctionInfo ('Postgres pgKind) ->
|
|
|
|
|
TableName ('Postgres pgKind) ->
|
|
|
|
|
NESeq (ColumnInfo ('Postgres pgKind)) ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m [FieldParser n (QueryDB ('Postgres pgKind) (RemoteRelationshipField IR.UnpreparedValue) (IR.UnpreparedValue ('Postgres pgKind)))]
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
buildFunctionRelayQueryFields sourceName functionName functionInfo tableName pkeyColumns = do
|
2021-11-26 00:07:53 +03:00
|
|
|
|
let fieldDesc = Just $ G.Description $ "execute function " <> functionName <<> " which returns " <>> tableName
|
2021-09-24 01:56:37 +03:00
|
|
|
|
fmap afold $
|
2021-12-09 12:05:42 +03:00
|
|
|
|
optionalFieldParser QDBConnection $
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
selectFunctionConnection sourceName functionInfo fieldDesc pkeyColumns
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
|
|
|
|
----------------------------------------------------------------
|
|
|
|
|
-- Individual components
|
|
|
|
|
|
2021-09-24 01:56:37 +03:00
|
|
|
|
columnParser ::
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(MonadSchema n m, MonadError QErr m, MonadReader r m, Has P.MkTypename r, Has NamingCase r) =>
|
2021-09-24 01:56:37 +03:00
|
|
|
|
ColumnType ('Postgres pgKind) ->
|
|
|
|
|
G.Nullability ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (Parser 'Both n (IR.ValueWithOrigin (ColumnValue ('Postgres pgKind))))
|
2022-05-26 14:54:30 +03:00
|
|
|
|
columnParser columnType (G.Nullability isNullable) = do
|
|
|
|
|
tCase <- asks getter
|
2021-02-23 20:37:27 +03:00
|
|
|
|
-- TODO(PDV): It might be worth memoizing this function even though it isn’t
|
|
|
|
|
-- recursive simply for performance reasons, since it’s likely to be hammered
|
|
|
|
|
-- during schema generation. Need to profile to see whether or not it’s a win.
|
server: remove remnants of query plan caching (fix #1795)
Query plan caching was introduced by - I believe - hasura/graphql-engine#1934 in order to reduce the query response latency. During the development of PDV in hasura/graphql-engine#4111, it was found out that the new architecture (for which query plan caching wasn't implemented) performed comparably to the pre-PDV architecture with caching. Hence, it was decided to leave query plan caching until some day in the future when it was deemed necessary.
Well, we're in the future now, and there still isn't a convincing argument for query plan caching. So the time has come to remove some references to query plan caching from the codebase. For the most part, any code being removed would probably not be very well suited to the post-PDV architecture of query execution, so arguably not much is lost.
Apart from simplifying the code, this PR will contribute towards making the GraphQL schema generation more modular, testable, and easier to profile. I'd like to eventually work towards a situation in which it's easy to generate a GraphQL schema parser *in isolation*, without being connected to a database, and then parse a GraphQL query *in isolation*, without even listening any HTTP port. It is important that both of these operations can be examined in detail, and in isolation, since they are two major performance bottlenecks, as well as phases where many important upcoming features hook into.
Implementation
The following have been removed:
- The entirety of `server/src-lib/Hasura/GraphQL/Execute/Plan.hs`
- The core phases of query parsing and execution no longer have any references to query plan caching. Note that this is not to be confused with query *response* caching, which is not affected by this PR. This includes removal of the types:
- - `Opaque`, which is replaced by a tuple. Note that the old implementation was broken and did not adequately hide the constructors.
- - `QueryReusability` (and the `markNotReusable` method). Notably, the implementation of the `ParseT` monad now consists of two, rather than three, monad transformers.
- Cache-related tests (in `server/src-test/Hasura/CacheBoundedSpec.hs`) have been removed .
- References to query plan caching in the documentation.
- The `planCacheOptions` in the `TenantConfig` type class was removed. However, during parsing, unrecognized fields in the YAML config get ignored, so this does not cause a breaking change. (Confirmed manually, as well as in consultation with @sordina.)
- The metrics no longer send cache hit/miss messages.
There are a few places in which one can still find references to query plan caching:
- We still accept the `--query-plan-cache-size` command-line option for backwards compatibility. The `HASURA_QUERY_PLAN_CACHE_SIZE` environment variable is not read.
https://github.com/hasura/graphql-engine-mono/pull/1815
GitOrigin-RevId: 17d92b254ec093c62a7dfeec478658ede0813eb7
2021-07-27 14:51:52 +03:00
|
|
|
|
peelWithOrigin . fmap (ColumnValue columnType) <$> case columnType of
|
2021-09-24 01:56:37 +03:00
|
|
|
|
ColumnScalar scalarType ->
|
|
|
|
|
possiblyNullable scalarType <$> do
|
|
|
|
|
-- We convert the value to JSON and use the FromJSON instance. This avoids
|
|
|
|
|
-- having two separate ways of parsing a value in the codebase, which
|
|
|
|
|
-- could lead to inconsistencies.
|
|
|
|
|
--
|
|
|
|
|
-- The mapping from postgres type to GraphQL scalar name is done by
|
|
|
|
|
-- 'mkScalarTypeName'. This is confusing, and we might want to fix it
|
|
|
|
|
-- later, as we will parse values differently here than how they'd be
|
|
|
|
|
-- parsed in other places using the same scalar name; for instance, we
|
|
|
|
|
-- will accept strings for postgres columns of type "Integer", despite the
|
|
|
|
|
-- fact that they will be represented as GraphQL ints, which otherwise do
|
|
|
|
|
-- not accept strings.
|
|
|
|
|
--
|
|
|
|
|
-- TODO: introduce new dedicated scalars for Postgres column types.
|
Refactor type name customization
Source typename customization (hasura/graphql-engine@aac64f2c81faa6a3aef4d0cf5fae97289ac4383e) introduced a mechanism to change certain names in the GraphQL schema that is exposed. In particular it allows last-minute modification of:
1. the names of some types, and
2. the names of some root fields.
The above two items are assigned distinct customization algorithms, and at times both algorithms are in scope. So a need to distinguish them is needed.
In the original design, this was addressed by introducing a newtype wrapper `Typename` around GraphQL `Name`s, dedicated to the names of types. However, in the majority of the codebase, type names are also represented by `Name`. For this reason, it was unavoidable to allow for easy conversion. This was supported by a `HasName Typename` instance, as well as by publishing the constructors of `Typename`.
This means that the type safety that newtypes can add is lost. In particular, it is now very easy to confuse type name customization with root field name customization.
This refactors the above design by instead introducing newtypes around the customization operations:
```haskell
newtype MkTypename = MkTypename {runMkTypename :: Name -> Name}
deriving (Semigroup, Monoid) via (Endo Name)
newtype MkRootFieldName = MkRootFieldName {runMkRootFieldName :: Name -> Name}
deriving (Semigroup, Monoid) via (Endo Name)
```
The `Monoid` instance allows easy composition of customization operations, piggybacking off of the type of `Endo`maps.
This design allows safe co-existence of the two customization algorithms, while avoiding the syntactic overhead of packing and unpacking newtypes.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/2989
GitOrigin-RevId: da3a353a9b003ee40c8d0a1e02872e99d2edd3ca
2021-11-30 12:51:46 +03:00
|
|
|
|
name <- mkScalarTypeName scalarType
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
let schemaType = P.TNamed P.NonNullable $ P.Definition name Nothing Nothing P.TIScalar
|
2021-09-24 01:56:37 +03:00
|
|
|
|
pure $
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
P.Parser
|
2021-09-24 01:56:37 +03:00
|
|
|
|
{ pType = schemaType,
|
|
|
|
|
pParser =
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
P.valueToJSON (P.toGraphQLType schemaType) >=> \case
|
|
|
|
|
J.Null -> P.parseError $ "unexpected null value for type " <>> name
|
2021-09-24 01:56:37 +03:00
|
|
|
|
value ->
|
|
|
|
|
runAesonParser (parsePGValue scalarType) value
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
`onLeft` (P.parseErrorWith ParseFailed . qeError)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
}
|
2022-04-13 13:07:06 +03:00
|
|
|
|
ColumnEnumReference (EnumReference tableName enumValues tableCustomName) ->
|
2021-02-23 20:37:27 +03:00
|
|
|
|
case nonEmpty (Map.toList enumValues) of
|
|
|
|
|
Just enumValuesList -> do
|
2022-04-13 13:07:06 +03:00
|
|
|
|
tableGQLName <- qualifiedObjectToName tableName
|
|
|
|
|
name <- addEnumSuffix tableGQLName tableCustomName
|
2022-05-26 14:54:30 +03:00
|
|
|
|
pure $ possiblyNullable PGText $ P.enum name Nothing (mkEnumValue tCase <$> enumValuesList)
|
2021-02-23 20:37:27 +03:00
|
|
|
|
Nothing -> throw400 ValidationFailed "empty enum values"
|
|
|
|
|
where
|
|
|
|
|
possiblyNullable scalarType
|
|
|
|
|
| isNullable = fmap (fromMaybe $ PGNull scalarType) . P.nullable
|
2021-09-24 01:56:37 +03:00
|
|
|
|
| otherwise = id
|
2022-05-26 14:54:30 +03:00
|
|
|
|
mkEnumValue :: NamingCase -> (EnumValue, EnumValueInfo) -> (P.Definition P.EnumValueInfo, PGScalarValue)
|
|
|
|
|
mkEnumValue tCase (EnumValue value, EnumValueInfo description) =
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
( P.Definition (applyEnumValueCase tCase value) (G.Description <$> description) Nothing P.EnumValueInfo,
|
2021-09-24 01:56:37 +03:00
|
|
|
|
PGValText $ G.unName value
|
2021-02-23 20:37:27 +03:00
|
|
|
|
)
|
|
|
|
|
|
2022-05-03 11:58:56 +03:00
|
|
|
|
pgScalarSelectionArgumentsParser ::
|
2021-09-24 01:56:37 +03:00
|
|
|
|
MonadParse n =>
|
|
|
|
|
ColumnType ('Postgres pgKind) ->
|
2022-05-03 11:58:56 +03:00
|
|
|
|
InputFieldsParser n (Maybe (ScalarSelectionArguments ('Postgres pgKind)))
|
|
|
|
|
pgScalarSelectionArgumentsParser columnType
|
2021-02-23 20:37:27 +03:00
|
|
|
|
| isScalarColumnWhere PG.isJSONType columnType =
|
2021-09-24 01:56:37 +03:00
|
|
|
|
P.fieldOptional fieldName description P.string `P.bindFields` fmap join . traverse toColExp
|
2021-02-23 20:37:27 +03:00
|
|
|
|
| otherwise = pure Nothing
|
|
|
|
|
where
|
2022-06-23 12:14:24 +03:00
|
|
|
|
fieldName = Name._path
|
2021-02-23 20:37:27 +03:00
|
|
|
|
description = Just "JSON select path"
|
|
|
|
|
toColExp textValue = case parseJSONPath textValue of
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
Left err -> P.parseError $ T.pack $ "parse json path error: " ++ err
|
2021-09-24 01:56:37 +03:00
|
|
|
|
Right [] -> pure Nothing
|
2022-05-03 11:58:56 +03:00
|
|
|
|
Right jPaths -> pure $ Just $ PG.ColumnOp PG.jsonbPathOp $ PG.SEArray $ map elToColExp jPaths
|
2022-06-08 18:31:28 +03:00
|
|
|
|
elToColExp (Key k) = PG.SELit $ K.toText k
|
2021-02-23 20:37:27 +03:00
|
|
|
|
elToColExp (Index i) = PG.SELit $ tshow i
|
|
|
|
|
|
2022-05-26 14:54:30 +03:00
|
|
|
|
orderByOperatorsHasuraCase ::
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(G.Name, NonEmpty (Definition P.EnumValueInfo, (BasicOrderType ('Postgres pgKind), NullsOrderType ('Postgres pgKind))))
|
2022-05-26 14:54:30 +03:00
|
|
|
|
orderByOperatorsHasuraCase = orderByOperators HasuraCase
|
|
|
|
|
|
|
|
|
|
orderByOperatorsGraphqlCase ::
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(G.Name, NonEmpty (Definition P.EnumValueInfo, (BasicOrderType ('Postgres pgKind), NullsOrderType ('Postgres pgKind))))
|
2022-05-26 14:54:30 +03:00
|
|
|
|
orderByOperatorsGraphqlCase = orderByOperators GraphqlCase
|
|
|
|
|
|
|
|
|
|
-- | Do NOT use this function directly, this should be used via
|
|
|
|
|
-- @orderByOperatorsHasuraCase@ or @orderByOperatorsGraphqlCase@
|
2021-09-24 01:56:37 +03:00
|
|
|
|
orderByOperators ::
|
2022-05-26 14:54:30 +03:00
|
|
|
|
NamingCase ->
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(G.Name, NonEmpty (Definition P.EnumValueInfo, (BasicOrderType ('Postgres pgKind), NullsOrderType ('Postgres pgKind))))
|
2022-05-26 14:54:30 +03:00
|
|
|
|
orderByOperators tCase =
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(Name._order_by,) $
|
2022-06-10 06:59:00 +03:00
|
|
|
|
NE.fromList
|
2022-06-23 12:14:24 +03:00
|
|
|
|
[ ( define (applyFieldNameCaseCust tCase Name._asc) "in ascending order, nulls last",
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(PG.OTAsc, PG.NLast)
|
|
|
|
|
),
|
2022-06-23 12:14:24 +03:00
|
|
|
|
( define (applyFieldNameCaseCust tCase Name._asc_nulls_first) "in ascending order, nulls first",
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(PG.OTAsc, PG.NFirst)
|
|
|
|
|
),
|
2022-06-23 12:14:24 +03:00
|
|
|
|
( define (applyFieldNameCaseCust tCase Name._asc_nulls_last) "in ascending order, nulls last",
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(PG.OTAsc, PG.NLast)
|
|
|
|
|
),
|
2022-06-23 12:14:24 +03:00
|
|
|
|
( define (applyFieldNameCaseCust tCase Name._desc) "in descending order, nulls first",
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(PG.OTDesc, PG.NFirst)
|
|
|
|
|
),
|
2022-06-23 12:14:24 +03:00
|
|
|
|
( define (applyFieldNameCaseCust tCase Name._desc_nulls_first) "in descending order, nulls first",
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(PG.OTDesc, PG.NFirst)
|
|
|
|
|
),
|
2022-06-23 12:14:24 +03:00
|
|
|
|
( define (applyFieldNameCaseCust tCase Name._desc_nulls_last) "in descending order, nulls last",
|
2022-06-10 06:59:00 +03:00
|
|
|
|
(PG.OTDesc, PG.NLast)
|
|
|
|
|
)
|
|
|
|
|
]
|
2021-02-23 20:37:27 +03:00
|
|
|
|
where
|
server: Metadata origin for definitions (type parameter version v2)
The code that builds the GraphQL schema, and `buildGQLContext` in particular, is partial: not every value of `(ServerConfigCtx, GraphQLQueryType, SourceCache, HashMap RemoteSchemaName (RemoteSchemaCtx, MetadataObject), ActionCache, AnnotatedCustomTypes)` results in a valid GraphQL schema. When it fails, we want to be able to return better error messages than we currently do.
The key thing that is missing is a way to trace back GraphQL type information to their origin from the Hasura metadata. Currently, we have a number of correctness checks of our GraphQL schema. But these correctness checks only have access to pure GraphQL type information, and hence can only report errors in terms of that. Possibly the worst is the "conflicting definitions" error, which, in practice, can only be debugged by Hasura engineers. This is terrible DX for customers.
This PR allows us to print better error messages, by adding a field to the `Definition` type that traces the GraphQL type to its origin in the metadata. So the idea is simple: just add `MetadataObjId`, or `Maybe` that, or some other sum type of that, to `Definition`.
However, we want to avoid having to import a `Hasura.RQL` module from `Hasura.GraphQL.Parser`. So we instead define this additional field of `Definition` through a new type parameter, which is threaded through in `Hasura.GraphQL.Parser`. We then define type synonyms in `Hasura.GraphQL.Schema.Parser` that fill in this type parameter, so that it is not visible for the majority of the codebase.
The idea of associating metadata information to `Definition`s really comes to fruition when combined with hasura/graphql-engine-mono#4517. Their combination would allow us to use the API of fatal errors (just like the current `MonadError QErr`) to report _inconsistencies_ in the metadata. Such inconsistencies are then _automatically_ ignored. So no ad-hoc decisions need to be made on how to cut out inconsistent metadata from the GraphQL schema. This will allow us to report much better errors, as well as improve the likelihood of a successful HGE startup.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4770
Co-authored-by: Samir Talwar <47582+SamirTalwar@users.noreply.github.com>
GitOrigin-RevId: 728402b0cae83ae8e83463a826ceeb609001acae
2022-06-28 18:52:26 +03:00
|
|
|
|
define name desc = P.Definition name (Just desc) Nothing P.EnumValueInfo
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
2021-09-24 01:56:37 +03:00
|
|
|
|
comparisonExps ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
( BackendSchema ('Postgres pgKind),
|
|
|
|
|
MonadSchema n m,
|
|
|
|
|
MonadError QErr m,
|
|
|
|
|
MonadReader r m,
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
Has SchemaOptions r,
|
2022-05-26 14:54:30 +03:00
|
|
|
|
Has MkTypename r,
|
|
|
|
|
Has NamingCase r
|
2021-09-24 01:56:37 +03:00
|
|
|
|
) =>
|
|
|
|
|
ColumnType ('Postgres pgKind) ->
|
|
|
|
|
m (Parser 'Input n [ComparisonExp ('Postgres pgKind)])
|
2022-05-31 01:07:02 +03:00
|
|
|
|
comparisonExps = memoize 'comparisonExps \columnType -> do
|
2021-02-23 20:37:27 +03:00
|
|
|
|
-- see Note [Columns in comparison expression are never nullable]
|
Remove circular dependency in schema building code
### Description
The main goal of this PR is, as stated, to remove the circular dependency in the schema building code. This cycle arises from the existence of remote relationships: when we build the schema for a source A, a remote relationship might force us to jump to the schema of a source B, or some remote schema. As a result, we end up having to do a dispatch from a "leaf" of the schema, similar to the one done at the root. In turn, this forces us to carry along in the schema a lot of information required for that dispatch, AND it forces us to import the instances in scope, creating an import loop.
As discussed in #4489, this PR implements the "dependency injection" solution: we pass to the schema a function to call to do the dispatch, and to get a generated field for a remote relationship. That way, this function can be chosen at the root level, and the leaves need not be aware of the overall context.
This PR grew a bit bigger than that, however; in an attempt to try and remove the `SourceCache` from the schema altogether, it changed a lot of functions across the schema building code, to thread along the `SourceInfo b` of the source being built. This avoids having to do cache lookups within a given source. A few cases remain, such as relay, that we might try to tackle in a subsequent PR.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/4557
GitOrigin-RevId: 9388e48372877520a72a9fd1677005df9f7b2d72
2022-05-27 20:21:22 +03:00
|
|
|
|
collapseIfNull <- retrieve soDangerousBooleanCollapse
|
2021-04-08 11:25:11 +03:00
|
|
|
|
|
|
|
|
|
-- parsers used for comparison arguments
|
2021-09-24 01:56:37 +03:00
|
|
|
|
geogInputParser <- geographyWithinDistanceInput
|
|
|
|
|
geomInputParser <- geometryWithinDistanceInput
|
|
|
|
|
ignInputParser <- intersectsGeomNbandInput
|
|
|
|
|
ingInputParser <- intersectsNbandGeomInput
|
|
|
|
|
typedParser <- columnParser columnType (G.Nullability False)
|
2021-02-23 20:37:27 +03:00
|
|
|
|
nullableTextParser <- columnParser (ColumnScalar PGText) (G.Nullability True)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
textParser <- columnParser (ColumnScalar PGText) (G.Nullability False)
|
2021-02-25 14:05:51 +03:00
|
|
|
|
-- `lquery` represents a regular-expression-like pattern for matching `ltree` values.
|
2021-09-24 01:56:37 +03:00
|
|
|
|
lqueryParser <- columnParser (ColumnScalar PGLquery) (G.Nullability False)
|
2021-02-25 14:05:51 +03:00
|
|
|
|
-- `ltxtquery` represents a full-text-search-like pattern for matching `ltree` values.
|
2021-09-24 01:56:37 +03:00
|
|
|
|
ltxtqueryParser <- columnParser (ColumnScalar PGLtxtquery) (G.Nullability False)
|
|
|
|
|
maybeCastParser <- castExp columnType
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase <- asks getter
|
2022-06-23 12:14:24 +03:00
|
|
|
|
let name = applyTypeNameCaseCust tCase $ P.getName typedParser <> Name.__comparison_exp
|
2021-09-24 01:56:37 +03:00
|
|
|
|
desc =
|
|
|
|
|
G.Description $
|
|
|
|
|
"Boolean expression to compare columns of type "
|
|
|
|
|
<> P.getName typedParser
|
|
|
|
|
<<> ". All fields are combined with logical 'AND'."
|
2022-05-31 01:07:02 +03:00
|
|
|
|
textListParser = fmap IR.openValueOrigin <$> P.list textParser
|
|
|
|
|
columnListParser = fmap IR.openValueOrigin <$> P.list typedParser
|
2022-05-26 14:54:30 +03:00
|
|
|
|
-- Naming conventions
|
2021-09-24 01:56:37 +03:00
|
|
|
|
pure $
|
|
|
|
|
P.object name (Just desc) $
|
|
|
|
|
fmap catMaybes $
|
|
|
|
|
sequenceA $
|
|
|
|
|
concat
|
|
|
|
|
[ flip (maybe []) maybeCastParser $ \castParser ->
|
2022-06-23 12:14:24 +03:00
|
|
|
|
[ P.fieldOptional Name.__cast Nothing (ACast <$> castParser)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
],
|
|
|
|
|
-- Common ops for all types
|
|
|
|
|
equalityOperators
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(IR.mkParameter <$> typedParser)
|
2022-06-29 16:35:59 +03:00
|
|
|
|
(mkListParameter columnType <$> columnListParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
-- Comparison ops for non Raster types
|
|
|
|
|
guard (isScalarColumnWhere (/= PGRaster) columnType)
|
|
|
|
|
*> comparisonOperators
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
-- Ops for Raster types
|
|
|
|
|
guard (isScalarColumnWhere (== PGRaster) columnType)
|
|
|
|
|
*> [ mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "intersects", "rast"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
Nothing
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTIntersectsRast . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "intersects", "nband", "geom"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
Nothing
|
|
|
|
|
(ABackendSpecific . ASTIntersectsNbandGeom <$> ingInputParser),
|
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "intersects", "geom", "nband"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
Nothing
|
|
|
|
|
(ABackendSpecific . ASTIntersectsGeomNband <$> ignInputParser)
|
|
|
|
|
],
|
|
|
|
|
-- Ops for String like types
|
|
|
|
|
guard (isScalarColumnWhere isStringType columnType)
|
|
|
|
|
*> [ mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__like)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column match the given pattern")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ALIKE . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__nlike)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column NOT match the given pattern")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ANLIKE . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__ilike)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column match the given case-insensitive pattern")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AILIKE . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__nilike)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column NOT match the given case-insensitive pattern")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ANILIKE . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__similar)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column match the given SQL regular expression")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASIMILAR . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__nsimilar)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column NOT match the given SQL regular expression")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ANSIMILAR . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__regex)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column match the given POSIX regular expression, case sensitive")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AREGEX . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__iregex)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column match the given POSIX regular expression, case insensitive")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AIREGEX . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__nregex)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column NOT match the given POSIX regular expression, case sensitive")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ANREGEX . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__niregex)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column NOT match the given POSIX regular expression, case insensitive")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ANIREGEX . IR.mkParameter <$> typedParser)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
],
|
|
|
|
|
-- Ops for JSONB type
|
|
|
|
|
guard (isScalarColumnWhere (== PGJSONB) columnType)
|
|
|
|
|
*> [ mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__contains)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column contain the given json value at the top level")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AContains . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_contained", "in"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the column contained in the given json value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AContainedIn . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_has", "key"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the string exist as a top-level key in the column")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AHasKey . IR.mkParameter <$> nullableTextParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_has", "keys", "any"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "do any of these strings exist as top-level keys in the column")
|
|
|
|
|
(ABackendSpecific . AHasKeysAny . mkListLiteral (ColumnScalar PGText) <$> textListParser),
|
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_has", "keys", "all"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "do all of these strings exist as top-level keys in the column")
|
|
|
|
|
(ABackendSpecific . AHasKeysAll . mkListLiteral (ColumnScalar PGText) <$> textListParser)
|
|
|
|
|
],
|
|
|
|
|
-- Ops for Geography type
|
|
|
|
|
guard (isScalarColumnWhere (== PGGeography) columnType)
|
|
|
|
|
*> [ mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "intersects"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column spatially intersect the given geography value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTIntersects . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "d", "within"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the column within a given distance from the given geography value")
|
|
|
|
|
(ABackendSpecific . ASTDWithinGeog <$> geogInputParser)
|
|
|
|
|
],
|
|
|
|
|
-- Ops for Geometry type
|
|
|
|
|
guard (isScalarColumnWhere (== PGGeometry) columnType)
|
|
|
|
|
*> [ mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "contains"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column contain the given geometry value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTContains . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "crosses"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column cross the given geometry value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTCrosses . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "equals"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the column equal to given geometry value (directionality is ignored)")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTEquals . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "overlaps"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column 'spatially overlap' (intersect but not completely contain) the given geometry value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTOverlaps . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "touches"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column have atleast one point in common with the given geometry value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTTouches . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "within"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the column contained in the given geometry value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTWithin . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "intersects"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column spatially intersect the given geometry value")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ASTIntersects . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "3d", "intersects"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does the column spatially intersect the given geometry value in 3D")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AST3DIntersects . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "d", "within"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the column within a given distance from the given geometry value")
|
|
|
|
|
(ABackendSpecific . ASTDWithinGeom <$> geomInputParser),
|
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_st", "3d", "d", "within"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the column within a given 3D distance from the given geometry value")
|
|
|
|
|
(ABackendSpecific . AST3DDWithinGeom <$> geomInputParser)
|
|
|
|
|
],
|
|
|
|
|
-- Ops for Ltree type
|
|
|
|
|
guard (isScalarColumnWhere (== PGLtree) columnType)
|
|
|
|
|
*> [ mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__ancestor)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the left argument an ancestor of right (or equal)?")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AAncestor . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_ancestor", "any"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does array contain an ancestor of `ltree`?")
|
|
|
|
|
(ABackendSpecific . AAncestorAny . mkListLiteral columnType <$> columnListParser),
|
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__descendant)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "is the left argument a descendant of right (or equal)?")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . ADescendant . IR.mkParameter <$> typedParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_descendant", "any"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does array contain a descendant of `ltree`?")
|
|
|
|
|
(ABackendSpecific . ADescendantAny . mkListLiteral columnType <$> columnListParser),
|
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-06-23 12:14:24 +03:00
|
|
|
|
(C.fromName Name.__matches)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does `ltree` match `lquery`?")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AMatches . IR.mkParameter <$> lqueryParser),
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_matches", "any"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does `ltree` match any `lquery` in array?")
|
|
|
|
|
(ABackendSpecific . AMatchesAny . mkListLiteral (ColumnScalar PGLquery) <$> textListParser),
|
|
|
|
|
mkBoolOperator
|
2022-05-26 14:54:30 +03:00
|
|
|
|
tCase
|
2021-09-24 01:56:37 +03:00
|
|
|
|
collapseIfNull
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(C.fromTuple $$(G.litGQLIdentifier ["_matches", "fulltext"]))
|
2021-09-24 01:56:37 +03:00
|
|
|
|
(Just "does `ltree` match `ltxtquery`?")
|
2022-05-31 01:07:02 +03:00
|
|
|
|
(ABackendSpecific . AMatchesFulltext . IR.mkParameter <$> ltxtqueryParser)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
]
|
|
|
|
|
]
|
2021-02-23 20:37:27 +03:00
|
|
|
|
where
|
2022-05-31 01:07:02 +03:00
|
|
|
|
mkListLiteral :: ColumnType ('Postgres pgKind) -> [ColumnValue ('Postgres pgKind)] -> IR.UnpreparedValue ('Postgres pgKind)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
mkListLiteral columnType columnValues =
|
2022-05-31 01:07:02 +03:00
|
|
|
|
IR.UVLiteral $
|
2021-09-24 01:56:37 +03:00
|
|
|
|
SETyAnn
|
|
|
|
|
(SEArray $ txtEncoder . cvValue <$> columnValues)
|
|
|
|
|
(mkTypeAnn $ CollectableTypeArray $ unsafePGColumnToBackend columnType)
|
2022-06-29 16:35:59 +03:00
|
|
|
|
mkListParameter :: ColumnType ('Postgres pgKind) -> [ColumnValue ('Postgres pgKind)] -> IR.UnpreparedValue ('Postgres pgKind)
|
|
|
|
|
mkListParameter columnType columnValues = do
|
|
|
|
|
let scalarType = unsafePGColumnToBackend columnType
|
|
|
|
|
IR.UVParameter Nothing $
|
|
|
|
|
ColumnValue
|
|
|
|
|
(ColumnScalar $ PG.PGArray scalarType)
|
|
|
|
|
(PG.PGValArray $ cvValue <$> columnValues)
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
2022-05-31 01:07:02 +03:00
|
|
|
|
castExp :: ColumnType ('Postgres pgKind) -> m (Maybe (Parser 'Input n (CastExp ('Postgres pgKind) (IR.UnpreparedValue ('Postgres pgKind)))))
|
2021-02-23 20:37:27 +03:00
|
|
|
|
castExp sourceType = do
|
|
|
|
|
let maybeScalars = case sourceType of
|
|
|
|
|
ColumnScalar PGGeography -> Just (PGGeography, PGGeometry)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
ColumnScalar PGGeometry -> Just (PGGeometry, PGGeography)
|
2022-03-10 10:32:04 +03:00
|
|
|
|
ColumnScalar PGJSONB -> Just (PGJSONB, PGText)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
_ -> Nothing
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
|
|
|
|
forM maybeScalars $ \(sourceScalar, targetScalar) -> do
|
2022-06-23 12:14:24 +03:00
|
|
|
|
sourceName <- mkScalarTypeName sourceScalar <&> (<> Name.__cast_exp)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
targetName <- mkScalarTypeName targetScalar
|
2021-02-23 20:37:27 +03:00
|
|
|
|
targetOpExps <- comparisonExps $ ColumnScalar targetScalar
|
2021-09-24 01:56:37 +03:00
|
|
|
|
let field = P.fieldOptional targetName Nothing $ (targetScalar,) <$> targetOpExps
|
Refactor type name customization
Source typename customization (hasura/graphql-engine@aac64f2c81faa6a3aef4d0cf5fae97289ac4383e) introduced a mechanism to change certain names in the GraphQL schema that is exposed. In particular it allows last-minute modification of:
1. the names of some types, and
2. the names of some root fields.
The above two items are assigned distinct customization algorithms, and at times both algorithms are in scope. So a need to distinguish them is needed.
In the original design, this was addressed by introducing a newtype wrapper `Typename` around GraphQL `Name`s, dedicated to the names of types. However, in the majority of the codebase, type names are also represented by `Name`. For this reason, it was unavoidable to allow for easy conversion. This was supported by a `HasName Typename` instance, as well as by publishing the constructors of `Typename`.
This means that the type safety that newtypes can add is lost. In particular, it is now very easy to confuse type name customization with root field name customization.
This refactors the above design by instead introducing newtypes around the customization operations:
```haskell
newtype MkTypename = MkTypename {runMkTypename :: Name -> Name}
deriving (Semigroup, Monoid) via (Endo Name)
newtype MkRootFieldName = MkRootFieldName {runMkRootFieldName :: Name -> Name}
deriving (Semigroup, Monoid) via (Endo Name)
```
The `Monoid` instance allows easy composition of customization operations, piggybacking off of the type of `Endo`maps.
This design allows safe co-existence of the two customization algorithms, while avoiding the syntactic overhead of packing and unpacking newtypes.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/2989
GitOrigin-RevId: da3a353a9b003ee40c8d0a1e02872e99d2edd3ca
2021-11-30 12:51:46 +03:00
|
|
|
|
pure $ P.object sourceName Nothing $ M.fromList . maybeToList <$> field
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
2021-09-24 01:56:37 +03:00
|
|
|
|
geographyWithinDistanceInput ::
|
2021-10-29 17:42:07 +03:00
|
|
|
|
forall pgKind m n r.
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(MonadSchema n m, MonadError QErr m, MonadReader r m, Has MkTypename r, Has NamingCase r) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (Parser 'Input n (DWithinGeogOp (IR.UnpreparedValue ('Postgres pgKind))))
|
2021-02-23 20:37:27 +03:00
|
|
|
|
geographyWithinDistanceInput = do
|
|
|
|
|
geographyParser <- columnParser (ColumnScalar PGGeography) (G.Nullability False)
|
|
|
|
|
-- FIXME
|
|
|
|
|
-- It doesn't make sense for this value to be nullable; it only is for
|
|
|
|
|
-- backwards compatibility; if an explicit Null value is given, it will be
|
|
|
|
|
-- forwarded to the underlying SQL function, that in turns treat a null value
|
|
|
|
|
-- as an error. We can fix this by rejecting explicit null values, by marking
|
|
|
|
|
-- this field non-nullable in a future release.
|
2021-09-24 01:56:37 +03:00
|
|
|
|
booleanParser <- columnParser (ColumnScalar PGBoolean) (G.Nullability True)
|
|
|
|
|
floatParser <- columnParser (ColumnScalar PGFloat) (G.Nullability False)
|
|
|
|
|
pure $
|
2022-06-23 12:14:24 +03:00
|
|
|
|
P.object Name._st_d_within_geography_input Nothing $
|
|
|
|
|
DWithinGeogOp <$> (IR.mkParameter <$> P.field Name._distance Nothing floatParser)
|
|
|
|
|
<*> (IR.mkParameter <$> P.field Name._from Nothing geographyParser)
|
|
|
|
|
<*> (IR.mkParameter <$> P.fieldWithDefault Name._use_spheroid Nothing (G.VBoolean True) booleanParser)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
|
|
|
|
|
geometryWithinDistanceInput ::
|
2021-10-29 17:42:07 +03:00
|
|
|
|
forall pgKind m n r.
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(MonadSchema n m, MonadError QErr m, MonadReader r m, Has MkTypename r, Has NamingCase r) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (Parser 'Input n (DWithinGeomOp (IR.UnpreparedValue ('Postgres pgKind))))
|
2021-02-23 20:37:27 +03:00
|
|
|
|
geometryWithinDistanceInput = do
|
|
|
|
|
geometryParser <- columnParser (ColumnScalar PGGeometry) (G.Nullability False)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
floatParser <- columnParser (ColumnScalar PGFloat) (G.Nullability False)
|
|
|
|
|
pure $
|
2022-06-23 12:14:24 +03:00
|
|
|
|
P.object Name._st_d_within_input Nothing $
|
|
|
|
|
DWithinGeomOp <$> (IR.mkParameter <$> P.field Name._distance Nothing floatParser)
|
|
|
|
|
<*> (IR.mkParameter <$> P.field Name._from Nothing geometryParser)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
|
|
|
|
|
intersectsNbandGeomInput ::
|
2021-10-29 17:42:07 +03:00
|
|
|
|
forall pgKind m n r.
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(MonadSchema n m, MonadError QErr m, MonadReader r m, Has MkTypename r, Has NamingCase r) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (Parser 'Input n (STIntersectsNbandGeommin (IR.UnpreparedValue ('Postgres pgKind))))
|
2021-02-23 20:37:27 +03:00
|
|
|
|
intersectsNbandGeomInput = do
|
|
|
|
|
geometryParser <- columnParser (ColumnScalar PGGeometry) (G.Nullability False)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
integerParser <- columnParser (ColumnScalar PGInteger) (G.Nullability False)
|
|
|
|
|
pure $
|
2022-06-23 12:14:24 +03:00
|
|
|
|
P.object Name._st_intersects_nband_geom_input Nothing $
|
|
|
|
|
STIntersectsNbandGeommin <$> (IR.mkParameter <$> P.field Name._nband Nothing integerParser)
|
|
|
|
|
<*> (IR.mkParameter <$> P.field Name._geommin Nothing geometryParser)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
|
|
|
|
|
intersectsGeomNbandInput ::
|
2021-10-29 17:42:07 +03:00
|
|
|
|
forall pgKind m n r.
|
2022-05-26 14:54:30 +03:00
|
|
|
|
(MonadSchema n m, MonadError QErr m, MonadReader r m, Has MkTypename r, Has NamingCase r) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (Parser 'Input n (STIntersectsGeomminNband (IR.UnpreparedValue ('Postgres pgKind))))
|
2021-02-23 20:37:27 +03:00
|
|
|
|
intersectsGeomNbandInput = do
|
|
|
|
|
geometryParser <- columnParser (ColumnScalar PGGeometry) (G.Nullability False)
|
2021-09-24 01:56:37 +03:00
|
|
|
|
integerParser <- columnParser (ColumnScalar PGInteger) (G.Nullability False)
|
|
|
|
|
pure $
|
2022-06-23 12:14:24 +03:00
|
|
|
|
P.object Name._st_intersects_geom_nband_input Nothing $
|
2021-09-24 01:56:37 +03:00
|
|
|
|
STIntersectsGeomminNband
|
2022-06-23 12:14:24 +03:00
|
|
|
|
<$> (IR.mkParameter <$> P.field Name._geommin Nothing geometryParser)
|
|
|
|
|
<*> (fmap IR.mkParameter <$> P.fieldOptional Name._nband Nothing integerParser)
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
2022-01-18 17:53:44 +03:00
|
|
|
|
countTypeInput ::
|
|
|
|
|
MonadParse n =>
|
|
|
|
|
Maybe (Parser 'Both n (Column ('Postgres pgKind))) ->
|
|
|
|
|
InputFieldsParser n (IR.CountDistinct -> CountType ('Postgres pgKind))
|
|
|
|
|
countTypeInput = \case
|
|
|
|
|
Just columnEnum -> do
|
2022-06-23 12:14:24 +03:00
|
|
|
|
columns <- P.fieldOptional Name._columns Nothing (P.list columnEnum)
|
2022-01-18 17:53:44 +03:00
|
|
|
|
pure $ flip mkCountType columns
|
|
|
|
|
Nothing -> pure $ flip mkCountType Nothing
|
|
|
|
|
where
|
|
|
|
|
mkCountType :: IR.CountDistinct -> Maybe [Column ('Postgres pgKind)] -> CountType ('Postgres pgKind)
|
|
|
|
|
mkCountType _ Nothing = PG.CTStar
|
|
|
|
|
mkCountType IR.SelectCountDistinct (Just cols) = PG.CTDistinct cols
|
|
|
|
|
mkCountType IR.SelectCountNonDistinct (Just cols) = PG.CTSimple cols
|
2021-02-23 20:37:27 +03:00
|
|
|
|
|
2021-11-18 21:02:58 +03:00
|
|
|
|
-- | Update operator that prepends a value to a column containing jsonb arrays.
|
|
|
|
|
--
|
|
|
|
|
-- Note: Currently this is Postgres specific because json columns have not been ported
|
|
|
|
|
-- to other backends yet.
|
|
|
|
|
prependOp ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
( BackendSchema ('Postgres pgKind),
|
|
|
|
|
MonadReader r m,
|
|
|
|
|
MonadError QErr m,
|
|
|
|
|
MonadSchema n m,
|
2022-05-26 14:54:30 +03:00
|
|
|
|
Has MkTypename r,
|
|
|
|
|
Has NamingCase r
|
2021-11-18 21:02:58 +03:00
|
|
|
|
) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
SU.UpdateOperator ('Postgres pgKind) m n (IR.UnpreparedValue ('Postgres pgKind))
|
2021-11-26 16:47:12 +03:00
|
|
|
|
prependOp = SU.UpdateOperator {..}
|
2021-11-18 21:02:58 +03:00
|
|
|
|
where
|
2022-01-19 11:37:50 +03:00
|
|
|
|
updateOperatorApplicableColumn = isScalarColumnWhere (== PGJSONB) . ciType
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
2022-04-18 22:43:00 +03:00
|
|
|
|
updateOperatorParser tableGQLName _tableName columns = do
|
2021-11-18 21:02:58 +03:00
|
|
|
|
let typedParser columnInfo =
|
2022-05-31 01:07:02 +03:00
|
|
|
|
fmap IR.mkParameter
|
2021-11-18 21:02:58 +03:00
|
|
|
|
<$> BS.columnParser
|
2022-01-19 11:37:50 +03:00
|
|
|
|
(ciType columnInfo)
|
|
|
|
|
(G.Nullability $ ciIsNullable columnInfo)
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
|
|
|
|
desc = "prepend existing jsonb value of filtered columns with new jsonb value"
|
|
|
|
|
|
2021-11-26 16:47:12 +03:00
|
|
|
|
SU.updateOperator
|
2021-11-18 21:02:58 +03:00
|
|
|
|
tableGQLName
|
2022-06-23 12:14:24 +03:00
|
|
|
|
Name.__prepend
|
2021-11-18 21:02:58 +03:00
|
|
|
|
typedParser
|
|
|
|
|
columns
|
|
|
|
|
desc
|
|
|
|
|
desc
|
|
|
|
|
|
|
|
|
|
-- | Update operator that appends a value to a column containing jsonb arrays.
|
|
|
|
|
--
|
|
|
|
|
-- Note: Currently this is Postgres specific because json columns have not been ported
|
|
|
|
|
-- to other backends yet.
|
|
|
|
|
appendOp ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
( BackendSchema ('Postgres pgKind),
|
|
|
|
|
MonadReader r m,
|
|
|
|
|
MonadError QErr m,
|
|
|
|
|
MonadSchema n m,
|
2022-05-26 14:54:30 +03:00
|
|
|
|
Has MkTypename r,
|
|
|
|
|
Has NamingCase r
|
2021-11-18 21:02:58 +03:00
|
|
|
|
) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
SU.UpdateOperator ('Postgres pgKind) m n (IR.UnpreparedValue ('Postgres pgKind))
|
2021-11-26 16:47:12 +03:00
|
|
|
|
appendOp = SU.UpdateOperator {..}
|
2021-11-18 21:02:58 +03:00
|
|
|
|
where
|
2022-01-19 11:37:50 +03:00
|
|
|
|
updateOperatorApplicableColumn = isScalarColumnWhere (== PGJSONB) . ciType
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
2022-04-18 22:43:00 +03:00
|
|
|
|
updateOperatorParser tableGQLName _tableName columns = do
|
2021-11-18 21:02:58 +03:00
|
|
|
|
let typedParser columnInfo =
|
2022-05-31 01:07:02 +03:00
|
|
|
|
fmap IR.mkParameter
|
2021-11-18 21:02:58 +03:00
|
|
|
|
<$> BS.columnParser
|
2022-01-19 11:37:50 +03:00
|
|
|
|
(ciType columnInfo)
|
|
|
|
|
(G.Nullability $ ciIsNullable columnInfo)
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
|
|
|
|
desc = "append existing jsonb value of filtered columns with new jsonb value"
|
2021-11-26 16:47:12 +03:00
|
|
|
|
SU.updateOperator
|
2021-11-18 21:02:58 +03:00
|
|
|
|
tableGQLName
|
2022-06-23 12:14:24 +03:00
|
|
|
|
Name.__append
|
2021-11-18 21:02:58 +03:00
|
|
|
|
typedParser
|
|
|
|
|
columns
|
|
|
|
|
desc
|
|
|
|
|
desc
|
|
|
|
|
|
|
|
|
|
-- | Update operator that deletes a value at a specified key from a column
|
|
|
|
|
-- containing jsonb objects.
|
|
|
|
|
--
|
|
|
|
|
-- Note: Currently this is Postgres specific because json columns have not been ported
|
|
|
|
|
-- to other backends yet.
|
|
|
|
|
deleteKeyOp ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
( BackendSchema ('Postgres pgKind),
|
|
|
|
|
MonadReader r m,
|
|
|
|
|
MonadError QErr m,
|
|
|
|
|
MonadSchema n m,
|
2022-05-26 14:54:30 +03:00
|
|
|
|
Has MkTypename r,
|
|
|
|
|
Has NamingCase r
|
2021-11-18 21:02:58 +03:00
|
|
|
|
) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
SU.UpdateOperator ('Postgres pgKind) m n (IR.UnpreparedValue ('Postgres pgKind))
|
2021-11-26 16:47:12 +03:00
|
|
|
|
deleteKeyOp = SU.UpdateOperator {..}
|
2021-11-18 21:02:58 +03:00
|
|
|
|
where
|
2022-01-19 11:37:50 +03:00
|
|
|
|
updateOperatorApplicableColumn = isScalarColumnWhere (== PGJSONB) . ciType
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
2022-04-18 22:43:00 +03:00
|
|
|
|
updateOperatorParser tableGQLName _tableName columns = do
|
2022-05-31 01:07:02 +03:00
|
|
|
|
let nullableTextParser _ = fmap IR.mkParameter <$> columnParser (ColumnScalar PGText) (G.Nullability True)
|
2021-11-18 21:02:58 +03:00
|
|
|
|
desc = "delete key/value pair or string element. key/value pairs are matched based on their key value"
|
|
|
|
|
|
2021-11-26 16:47:12 +03:00
|
|
|
|
SU.updateOperator
|
2021-11-18 21:02:58 +03:00
|
|
|
|
tableGQLName
|
2022-06-23 12:14:24 +03:00
|
|
|
|
Name.__delete_key
|
2021-11-18 21:02:58 +03:00
|
|
|
|
nullableTextParser
|
|
|
|
|
columns
|
|
|
|
|
desc
|
|
|
|
|
desc
|
|
|
|
|
|
|
|
|
|
-- | Update operator that deletes a value at a specific index from a column
|
|
|
|
|
-- containing jsonb arrays.
|
|
|
|
|
--
|
|
|
|
|
-- Note: Currently this is Postgres specific because json columns have not been ported
|
|
|
|
|
-- to other backends yet.
|
|
|
|
|
deleteElemOp ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
( BackendSchema ('Postgres pgKind),
|
|
|
|
|
MonadReader r m,
|
|
|
|
|
MonadError QErr m,
|
|
|
|
|
MonadSchema n m,
|
2022-05-26 14:54:30 +03:00
|
|
|
|
Has MkTypename r,
|
|
|
|
|
Has NamingCase r
|
2021-11-18 21:02:58 +03:00
|
|
|
|
) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
SU.UpdateOperator ('Postgres pgKind) m n (IR.UnpreparedValue ('Postgres pgKind))
|
2021-11-26 16:47:12 +03:00
|
|
|
|
deleteElemOp = SU.UpdateOperator {..}
|
2021-11-18 21:02:58 +03:00
|
|
|
|
where
|
2022-01-19 11:37:50 +03:00
|
|
|
|
updateOperatorApplicableColumn = isScalarColumnWhere (== PGJSONB) . ciType
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
2022-04-18 22:43:00 +03:00
|
|
|
|
updateOperatorParser tableGQLName _tableName columns = do
|
2022-05-31 01:07:02 +03:00
|
|
|
|
let nonNullableIntParser _ = fmap IR.mkParameter <$> columnParser (ColumnScalar PGInteger) (G.Nullability False)
|
2021-11-18 21:02:58 +03:00
|
|
|
|
desc =
|
|
|
|
|
"delete the array element with specified index (negative integers count from the end). "
|
|
|
|
|
<> "throws an error if top level container is not an array"
|
|
|
|
|
|
2021-11-26 16:47:12 +03:00
|
|
|
|
SU.updateOperator
|
2021-11-18 21:02:58 +03:00
|
|
|
|
tableGQLName
|
2022-06-23 12:14:24 +03:00
|
|
|
|
Name.__delete_elem
|
2021-11-18 21:02:58 +03:00
|
|
|
|
nonNullableIntParser
|
|
|
|
|
columns
|
|
|
|
|
desc
|
|
|
|
|
desc
|
|
|
|
|
|
|
|
|
|
-- | Update operator that deletes a field at a certan path from a column
|
|
|
|
|
-- containing jsonb objects.
|
|
|
|
|
--
|
|
|
|
|
-- Note: Currently this is Postgres specific because json columns have not been ported
|
|
|
|
|
-- to other backends yet.
|
|
|
|
|
deleteAtPathOp ::
|
|
|
|
|
forall pgKind m n r.
|
|
|
|
|
( BackendSchema ('Postgres pgKind),
|
|
|
|
|
MonadReader r m,
|
|
|
|
|
MonadError QErr m,
|
|
|
|
|
MonadSchema n m,
|
2022-05-26 14:54:30 +03:00
|
|
|
|
Has MkTypename r,
|
|
|
|
|
Has NamingCase r
|
2021-11-18 21:02:58 +03:00
|
|
|
|
) =>
|
2022-05-31 01:07:02 +03:00
|
|
|
|
SU.UpdateOperator ('Postgres pgKind) m n [IR.UnpreparedValue ('Postgres pgKind)]
|
2021-11-26 16:47:12 +03:00
|
|
|
|
deleteAtPathOp = SU.UpdateOperator {..}
|
2021-11-18 21:02:58 +03:00
|
|
|
|
where
|
2022-01-19 11:37:50 +03:00
|
|
|
|
updateOperatorApplicableColumn = isScalarColumnWhere (== PGJSONB) . ciType
|
2021-11-18 21:02:58 +03:00
|
|
|
|
|
2022-04-18 22:43:00 +03:00
|
|
|
|
updateOperatorParser tableGQLName _tableName columns = do
|
2022-05-31 01:07:02 +03:00
|
|
|
|
let nonNullableTextListParser _ = P.list . fmap IR.mkParameter <$> columnParser (ColumnScalar PGText) (G.Nullability False)
|
2021-11-18 21:02:58 +03:00
|
|
|
|
desc = "delete the field or element with specified path (for JSON arrays, negative integers count from the end)"
|
|
|
|
|
|
2021-11-26 16:47:12 +03:00
|
|
|
|
SU.updateOperator
|
2021-11-18 21:02:58 +03:00
|
|
|
|
tableGQLName
|
2022-06-23 12:14:24 +03:00
|
|
|
|
Name.__delete_at_path
|
2021-11-18 21:02:58 +03:00
|
|
|
|
nonNullableTextListParser
|
|
|
|
|
columns
|
|
|
|
|
desc
|
|
|
|
|
desc
|
|
|
|
|
|
2021-12-29 06:51:01 +03:00
|
|
|
|
-- | The update operators that we support on Postgres.
|
2021-11-18 21:02:58 +03:00
|
|
|
|
updateOperators ::
|
|
|
|
|
forall pgKind m n r.
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
MonadBuildSchema ('Postgres pgKind) r m n =>
|
2021-11-18 21:02:58 +03:00
|
|
|
|
TableInfo ('Postgres pgKind) ->
|
|
|
|
|
UpdPermInfo ('Postgres pgKind) ->
|
2022-05-31 01:07:02 +03:00
|
|
|
|
m (InputFieldsParser n (HashMap (Column ('Postgres pgKind)) (UpdateOpExpression (IR.UnpreparedValue ('Postgres pgKind)))))
|
Role-invariant schema constructors
We build the GraphQL schema by combining building blocks such as `tableSelectionSet` and `columnParser`. These building blocks individually build `{InputFields,Field,}Parser` objects. Those object specify the valid GraphQL schema.
Since the GraphQL schema is role-dependent, at some point we need to know what fragment of the GraphQL schema a specific role is allowed to access, and this is stored in `{Sel,Upd,Ins,Del}PermInfo` objects.
We have passed around these permission objects as function arguments to the schema building blocks since we first started dealing with permissions during the PDV refactor - see hasura/graphql-engine@5168b99e463199b1934d8645bd6cd37eddb64ae1 in hasura/graphql-engine#4111. This means that, for instance, `tableSelectionSet` has as its type:
```haskell
tableSelectionSet ::
forall b r m n.
MonadBuildSchema b r m n =>
SourceName ->
TableInfo b ->
SelPermInfo b ->
m (Parser 'Output n (AnnotatedFields b))
```
There are three reasons to change this.
1. We often pass a `Maybe (xPermInfo b)` instead of a proper `xPermInfo b`, and it's not clear what the intended semantics of this is. Some potential improvements on the data types involved are discussed in issue hasura/graphql-engine-mono#3125.
2. In most cases we also already pass a `TableInfo b`, and together with the `MonadRole` that is usually also in scope, this means that we could look up the required permissions regardless: so passing the permissions explicitly undermines the "single source of truth" principle. Breaking this principle also makes the code more difficult to read.
3. We are working towards role-based parsers (see hasura/graphql-engine-mono#2711), where the `{InputFields,Field,}Parser` objects are constructed in a role-invariant way, so that we have a single object that can be used for all roles. In particular, this means that the schema building blocks _need_ to be constructed in a role-invariant way. While this PR doesn't accomplish that, it does reduce the amount of role-specific arguments being passed, thus fixing hasura/graphql-engine-mono#3068.
Concretely, this PR simply drops the `xPermInfo b` argument from almost all schema building blocks. Instead these objects are looked up from the `TableInfo b` as-needed. The resulting code is considerably simpler and shorter.
One way to interpret this change is as follows. Before this PR, we figured out permissions at the top-level in `Hasura.GraphQL.Schema`, passing down the obtained `xPermInfo` objects as required. After this PR, we have a bottom-up approach where the schema building blocks themselves decide whether they want to be included for a particular role.
So this moves some permission logic out of `Hasura.GraphQL.Schema`, which is very complex.
PR-URL: https://github.com/hasura/graphql-engine-mono/pull/3608
GitOrigin-RevId: 51a744f34ec7d57bc8077667ae7f9cb9c4f6c962
2022-02-17 11:16:20 +03:00
|
|
|
|
updateOperators tableInfo updatePermissions = do
|
2021-11-26 16:47:12 +03:00
|
|
|
|
SU.buildUpdateOperators
|
|
|
|
|
(PGIR.UpdateSet <$> SU.presetColumns updatePermissions)
|
|
|
|
|
[ PGIR.UpdateSet <$> SU.setOp,
|
|
|
|
|
PGIR.UpdateInc <$> SU.incOp,
|
2021-11-25 00:39:42 +03:00
|
|
|
|
PGIR.UpdatePrepend <$> prependOp,
|
|
|
|
|
PGIR.UpdateAppend <$> appendOp,
|
|
|
|
|
PGIR.UpdateDeleteKey <$> deleteKeyOp,
|
|
|
|
|
PGIR.UpdateDeleteElem <$> deleteElemOp,
|
|
|
|
|
PGIR.UpdateDeleteAtPath <$> deleteAtPathOp
|
2021-11-18 21:02:58 +03:00
|
|
|
|
]
|
|
|
|
|
tableInfo
|