graphql-engine/server/src-lib/Hasura/Backends/MSSQL/FromIr.hs

1116 lines
45 KiB
Haskell
Raw Normal View History

{-# LANGUAGE ViewPatterns #-}
-- | Translate from the DML to the TSql dialect.
module Hasura.Backends.MSSQL.FromIr
( fromSelectRows
, mkSQLSelect
, fromRootField
, fromSelectAggregate
, fromAnnBoolExp
, Error(..)
, runFromIr
, FromIr
, jsonFieldName
, fromDelete
) where
import Hasura.Prelude
import qualified Data.HashMap.Strict as HM
import qualified Data.Map.Strict as M
import qualified Data.Text as T
import qualified Database.ODBC.SQLServer as ODBC
import Control.Monad.Validate
import Data.Map.Strict (Map)
import Data.Proxy
server: support remote relationships on SQL Server and BigQuery (#1497) Remote relationships are now supported on SQL Server and BigQuery. The major change though is the re-architecture of remote join execution logic. Prior to this PR, each backend is responsible for processing the remote relationships that are part of their AST. This is not ideal as there is nothing specific about a remote join's execution that ties it to a backend. The only backend specific part is whether or not the specification of the remote relationship is valid (i.e, we'll need to validate whether the scalars are compatible). The approach now changes to this: 1. Before delegating the AST to the backend, we traverse the AST, collect all the remote joins while modifying the AST to add necessary join fields where needed. 1. Once the remote joins are collected from the AST, the database call is made to fetch the response. The necessary data for the remote join(s) is collected from the database's response and one or more remote schema calls are constructed as necessary. 1. The remote schema calls are then executed and the data from the database and from the remote schemas is joined to produce the final response. ### Known issues 1. Ideally the traversal of the IR to collect remote joins should return an AST which does not include remote join fields. This operation can be type safe but isn't taken up as part of the PR. 1. There is a lot of code duplication between `Transport/HTTP.hs` and `Transport/Websocket.hs` which needs to be fixed ASAP. This too hasn't been taken up by this PR. 1. The type which represents the execution plan is only modified to handle our current remote joins and as such it will have to be changed to accommodate general remote joins. 1. Use of lenses would have reduced the boilerplate code to collect remote joins from the base AST. 1. The current remote join logic assumes that the join columns of a remote relationship appear with their names in the database response. This however is incorrect as they could be aliased. This can be taken up by anyone, I've left a comment in the code. ### Notes to the reviewers I think it is best reviewed commit by commit. 1. The first one is very straight forward. 1. The second one refactors the remote join execution logic but other than moving things around, it doesn't change the user facing functionality. This moves Postgres specific parts to `Backends/Postgres` module from `Execute`. Some IR related code to `Hasura.RQL.IR` module. Simplifies various type class function signatures as a backend doesn't have to handle remote joins anymore 1. The third one fixes partial case matches that for some weird reason weren't shown as warnings before this refactor 1. The fourth one generalizes the validation logic of remote relationships and implements `scalarTypeGraphQLName` function on SQL Server and BigQuery which is used by the validation logic. This enables remote relationships on BigQuery and SQL Server. https://github.com/hasura/graphql-engine-mono/pull/1497 GitOrigin-RevId: 77dd8eed326602b16e9a8496f52f46d22b795598
2021-06-11 06:26:50 +03:00
import qualified Hasura.RQL.IR as IR
import qualified Hasura.RQL.Types.Column as IR
import qualified Hasura.RQL.Types.Common as IR
import qualified Hasura.RQL.Types.Relationship as IR
import Hasura.Backends.MSSQL.Instances.Types ()
import Hasura.Backends.MSSQL.Types as TSQL
import Hasura.SQL.Backend
--------------------------------------------------------------------------------
-- Types
-- | Most of these errors should be checked for legitimacy.
data Error
= UnsupportedOpExpG (IR.OpExpG 'MSSQL Expression)
| FunctionNotSupported
| NodesUnsupportedForNow
server: support remote relationships on SQL Server and BigQuery (#1497) Remote relationships are now supported on SQL Server and BigQuery. The major change though is the re-architecture of remote join execution logic. Prior to this PR, each backend is responsible for processing the remote relationships that are part of their AST. This is not ideal as there is nothing specific about a remote join's execution that ties it to a backend. The only backend specific part is whether or not the specification of the remote relationship is valid (i.e, we'll need to validate whether the scalars are compatible). The approach now changes to this: 1. Before delegating the AST to the backend, we traverse the AST, collect all the remote joins while modifying the AST to add necessary join fields where needed. 1. Once the remote joins are collected from the AST, the database call is made to fetch the response. The necessary data for the remote join(s) is collected from the database's response and one or more remote schema calls are constructed as necessary. 1. The remote schema calls are then executed and the data from the database and from the remote schemas is joined to produce the final response. ### Known issues 1. Ideally the traversal of the IR to collect remote joins should return an AST which does not include remote join fields. This operation can be type safe but isn't taken up as part of the PR. 1. There is a lot of code duplication between `Transport/HTTP.hs` and `Transport/Websocket.hs` which needs to be fixed ASAP. This too hasn't been taken up by this PR. 1. The type which represents the execution plan is only modified to handle our current remote joins and as such it will have to be changed to accommodate general remote joins. 1. Use of lenses would have reduced the boilerplate code to collect remote joins from the base AST. 1. The current remote join logic assumes that the join columns of a remote relationship appear with their names in the database response. This however is incorrect as they could be aliased. This can be taken up by anyone, I've left a comment in the code. ### Notes to the reviewers I think it is best reviewed commit by commit. 1. The first one is very straight forward. 1. The second one refactors the remote join execution logic but other than moving things around, it doesn't change the user facing functionality. This moves Postgres specific parts to `Backends/Postgres` module from `Execute`. Some IR related code to `Hasura.RQL.IR` module. Simplifies various type class function signatures as a backend doesn't have to handle remote joins anymore 1. The third one fixes partial case matches that for some weird reason weren't shown as warnings before this refactor 1. The fourth one generalizes the validation logic of remote relationships and implements `scalarTypeGraphQLName` function on SQL Server and BigQuery which is used by the validation logic. This enables remote relationships on BigQuery and SQL Server. https://github.com/hasura/graphql-engine-mono/pull/1497 GitOrigin-RevId: 77dd8eed326602b16e9a8496f52f46d22b795598
2021-06-11 06:26:50 +03:00
| ConnectionsNotSupported
deriving (Show, Eq)
-- | The base monad used throughout this module for all conversion
-- functions.
--
-- It's a Validate, so it'll continue going when it encounters errors
-- to accumulate as many as possible.
--
-- It also contains a mapping from entity prefixes to counters. So if
-- my prefix is "table" then there'll be a counter that lets me
-- generate table1, table2, etc. Same for any other prefix needed
-- (e.g. names for joins).
--
-- A ReaderT is used around this in most of the module too, for
-- setting the current entity that a given field name refers to. See
-- @fromPGCol@.
newtype FromIr a = FromIr
{ unFromIr :: StateT (Map Text Int) (Validate (NonEmpty Error)) a
} deriving (Functor, Applicative, Monad, MonadValidate (NonEmpty Error))
data StringifyNumbers
= StringifyNumbers
| LeaveNumbersAlone
deriving (Eq)
--------------------------------------------------------------------------------
-- Runners
runFromIr :: FromIr a -> Validate (NonEmpty Error) a
runFromIr fromIr = evalStateT (unFromIr fromIr) mempty
--------------------------------------------------------------------------------
-- Similar rendition of old API
mkSQLSelect
:: IR.JsonAggSelect
-> IR.AnnSelectG 'MSSQL (Const Void) (IR.AnnFieldG 'MSSQL (Const Void)) Expression
-> FromIr TSQL.Select
mkSQLSelect jsonAggSelect annSimpleSel =
case jsonAggSelect of
IR.JASMultipleRows -> fromSelectRows annSimpleSel
IR.JASSingleObject ->
fromSelectRows annSimpleSel <&> \sel ->
sel
{ selectFor =
JsonFor
ForJson {jsonCardinality = JsonSingleton, jsonRoot = NoRoot}
, selectTop = Top 1
}
-- | Convert from the IR database query into a select.
fromRootField :: IR.QueryDB 'MSSQL (Const Void) Expression -> FromIr Select
fromRootField =
\case
server: support remote relationships on SQL Server and BigQuery (#1497) Remote relationships are now supported on SQL Server and BigQuery. The major change though is the re-architecture of remote join execution logic. Prior to this PR, each backend is responsible for processing the remote relationships that are part of their AST. This is not ideal as there is nothing specific about a remote join's execution that ties it to a backend. The only backend specific part is whether or not the specification of the remote relationship is valid (i.e, we'll need to validate whether the scalars are compatible). The approach now changes to this: 1. Before delegating the AST to the backend, we traverse the AST, collect all the remote joins while modifying the AST to add necessary join fields where needed. 1. Once the remote joins are collected from the AST, the database call is made to fetch the response. The necessary data for the remote join(s) is collected from the database's response and one or more remote schema calls are constructed as necessary. 1. The remote schema calls are then executed and the data from the database and from the remote schemas is joined to produce the final response. ### Known issues 1. Ideally the traversal of the IR to collect remote joins should return an AST which does not include remote join fields. This operation can be type safe but isn't taken up as part of the PR. 1. There is a lot of code duplication between `Transport/HTTP.hs` and `Transport/Websocket.hs` which needs to be fixed ASAP. This too hasn't been taken up by this PR. 1. The type which represents the execution plan is only modified to handle our current remote joins and as such it will have to be changed to accommodate general remote joins. 1. Use of lenses would have reduced the boilerplate code to collect remote joins from the base AST. 1. The current remote join logic assumes that the join columns of a remote relationship appear with their names in the database response. This however is incorrect as they could be aliased. This can be taken up by anyone, I've left a comment in the code. ### Notes to the reviewers I think it is best reviewed commit by commit. 1. The first one is very straight forward. 1. The second one refactors the remote join execution logic but other than moving things around, it doesn't change the user facing functionality. This moves Postgres specific parts to `Backends/Postgres` module from `Execute`. Some IR related code to `Hasura.RQL.IR` module. Simplifies various type class function signatures as a backend doesn't have to handle remote joins anymore 1. The third one fixes partial case matches that for some weird reason weren't shown as warnings before this refactor 1. The fourth one generalizes the validation logic of remote relationships and implements `scalarTypeGraphQLName` function on SQL Server and BigQuery which is used by the validation logic. This enables remote relationships on BigQuery and SQL Server. https://github.com/hasura/graphql-engine-mono/pull/1497 GitOrigin-RevId: 77dd8eed326602b16e9a8496f52f46d22b795598
2021-06-11 06:26:50 +03:00
(IR.QDBSingleRow s) -> mkSQLSelect IR.JASSingleObject s
(IR.QDBMultipleRows s) -> mkSQLSelect IR.JASMultipleRows s
(IR.QDBAggregation s) -> fromSelectAggregate Nothing s
server: support remote relationships on SQL Server and BigQuery (#1497) Remote relationships are now supported on SQL Server and BigQuery. The major change though is the re-architecture of remote join execution logic. Prior to this PR, each backend is responsible for processing the remote relationships that are part of their AST. This is not ideal as there is nothing specific about a remote join's execution that ties it to a backend. The only backend specific part is whether or not the specification of the remote relationship is valid (i.e, we'll need to validate whether the scalars are compatible). The approach now changes to this: 1. Before delegating the AST to the backend, we traverse the AST, collect all the remote joins while modifying the AST to add necessary join fields where needed. 1. Once the remote joins are collected from the AST, the database call is made to fetch the response. The necessary data for the remote join(s) is collected from the database's response and one or more remote schema calls are constructed as necessary. 1. The remote schema calls are then executed and the data from the database and from the remote schemas is joined to produce the final response. ### Known issues 1. Ideally the traversal of the IR to collect remote joins should return an AST which does not include remote join fields. This operation can be type safe but isn't taken up as part of the PR. 1. There is a lot of code duplication between `Transport/HTTP.hs` and `Transport/Websocket.hs` which needs to be fixed ASAP. This too hasn't been taken up by this PR. 1. The type which represents the execution plan is only modified to handle our current remote joins and as such it will have to be changed to accommodate general remote joins. 1. Use of lenses would have reduced the boilerplate code to collect remote joins from the base AST. 1. The current remote join logic assumes that the join columns of a remote relationship appear with their names in the database response. This however is incorrect as they could be aliased. This can be taken up by anyone, I've left a comment in the code. ### Notes to the reviewers I think it is best reviewed commit by commit. 1. The first one is very straight forward. 1. The second one refactors the remote join execution logic but other than moving things around, it doesn't change the user facing functionality. This moves Postgres specific parts to `Backends/Postgres` module from `Execute`. Some IR related code to `Hasura.RQL.IR` module. Simplifies various type class function signatures as a backend doesn't have to handle remote joins anymore 1. The third one fixes partial case matches that for some weird reason weren't shown as warnings before this refactor 1. The fourth one generalizes the validation logic of remote relationships and implements `scalarTypeGraphQLName` function on SQL Server and BigQuery which is used by the validation logic. This enables remote relationships on BigQuery and SQL Server. https://github.com/hasura/graphql-engine-mono/pull/1497 GitOrigin-RevId: 77dd8eed326602b16e9a8496f52f46d22b795598
2021-06-11 06:26:50 +03:00
(IR.QDBConnection _) -> refute $ pure ConnectionsNotSupported
--------------------------------------------------------------------------------
-- Top-level exported functions
fromSelectRows :: IR.AnnSelectG 'MSSQL (Const Void) (IR.AnnFieldG 'MSSQL (Const Void)) Expression -> FromIr TSQL.Select
fromSelectRows annSelectG = do
selectFrom <-
case from of
IR.FromTable qualifiedObject -> fromQualifiedTable qualifiedObject
[Preview] Inherited roles for postgres read queries fixes #3868 docker image - `hasura/graphql-engine:inherited-roles-preview-48b73a2de` Note: To be able to use the inherited roles feature, the graphql-engine should be started with the env variable `HASURA_GRAPHQL_EXPERIMENTAL_FEATURES` set to `inherited_roles`. Introduction ------------ This PR implements the idea of multiple roles as presented in this [paper](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/FGALanguageICDE07.pdf). The multiple roles feature in this PR can be used via inherited roles. An inherited role is a role which can be created by combining multiple singular roles. For example, if there are two roles `author` and `editor` configured in the graphql-engine, then we can create a inherited role with the name of `combined_author_editor` role which will combine the select permissions of the `author` and `editor` roles and then make GraphQL queries using the `combined_author_editor`. How are select permissions of different roles are combined? ------------------------------------------------------------ A select permission includes 5 things: 1. Columns accessible to the role 2. Row selection filter 3. Limit 4. Allow aggregation 5. Scalar computed fields accessible to the role Suppose there are two roles, `role1` gives access to the `address` column with row filter `P1` and `role2` gives access to both the `address` and the `phone` column with row filter `P2` and we create a new role `combined_roles` which combines `role1` and `role2`. Let's say the following GraphQL query is queried with the `combined_roles` role. ```graphql query { employees { address phone } } ``` This will translate to the following SQL query: ```sql select (case when (P1 or P2) then address else null end) as address, (case when P2 then phone else null end) as phone from employee where (P1 or P2) ``` The other parameters of the select permission will be combined in the following manner: 1. Limit - Minimum of the limits will be the limit of the inherited role 2. Allow aggregations - If any of the role allows aggregation, then the inherited role will allow aggregation 3. Scalar computed fields - same as table column fields, as in the above example APIs for inherited roles: ---------------------- 1. `add_inherited_role` `add_inherited_role` is the [metadata API](https://hasura.io/docs/1.0/graphql/core/api-reference/index.html#schema-metadata-api) to create a new inherited role. It accepts two arguments `role_name`: the name of the inherited role to be added (String) `role_set`: list of roles that need to be combined (Array of Strings) Example: ```json { "type": "add_inherited_role", "args": { "role_name":"combined_user", "role_set":[ "user", "user1" ] } } ``` After adding the inherited role, the inherited role can be used like single roles like earlier Note: An inherited role can only be created with non-inherited/singular roles. 2. `drop_inherited_role` The `drop_inherited_role` API accepts the name of the inherited role and drops it from the metadata. It accepts a single argument: `role_name`: name of the inherited role to be dropped Example: ```json { "type": "drop_inherited_role", "args": { "role_name":"combined_user" } } ``` Metadata --------- The derived roles metadata will be included under the `experimental_features` key while exporting the metadata. ```json { "experimental_features": { "derived_roles": [ { "role_name": "manager_is_employee_too", "role_set": [ "employee", "manager" ] } ] } } ``` Scope ------ Only postgres queries and subscriptions are supported in this PR. Important points: ----------------- 1. All columns exposed to an inherited role will be marked as `nullable`, this is done so that cell value nullification can be done. TODOs ------- - [ ] Tests - [ ] Test a GraphQL query running with a inherited role without enabling inherited roles in experimental features - [] Tests for aggregate queries, limit, computed fields, functions, subscriptions (?) - [ ] Introspection test with a inherited role (nullability changes in a inherited role) - [ ] Docs - [ ] Changelog Co-authored-by: Vamshi Surabhi <6562944+0x777@users.noreply.github.com> GitOrigin-RevId: 3b8ee1e11f5ceca80fe294f8c074d42fbccfec63
2021-03-08 14:14:13 +03:00
IR.FromFunction {} -> refute $ pure FunctionNotSupported
Args { argsOrderBy
, argsWhere
, argsJoins
, argsTop
, argsDistinct = Proxy
, argsOffset
, argsExistingJoins
} <- runReaderT (fromSelectArgsG args) (fromAlias selectFrom)
fieldSources <-
runReaderT
(traverse (fromAnnFieldsG argsExistingJoins stringifyNumbers) fields)
(fromAlias selectFrom)
filterExpression <-
runReaderT (fromAnnBoolExp permFilter) (fromAlias selectFrom)
let selectProjections =
concatMap (toList . fieldSourceProjections) fieldSources
pure
Select
{ selectOrderBy = argsOrderBy
, selectTop = permissionBasedTop <> argsTop
, selectProjections
, selectFrom = Just selectFrom
, selectJoins = argsJoins <> mapMaybe fieldSourceJoin fieldSources
, selectWhere = argsWhere <> Where [filterExpression]
, selectFor =
JsonFor ForJson {jsonCardinality = JsonArray, jsonRoot = NoRoot}
, selectOffset = argsOffset
}
where
IR.AnnSelectG { _asnFields = fields
, _asnFrom = from
, _asnPerm = perm
, _asnArgs = args
, _asnStrfyNum = num
} = annSelectG
IR.TablePerm {_tpLimit = mPermLimit, _tpFilter = permFilter} = perm
permissionBasedTop =
[Preview] Inherited roles for postgres read queries fixes #3868 docker image - `hasura/graphql-engine:inherited-roles-preview-48b73a2de` Note: To be able to use the inherited roles feature, the graphql-engine should be started with the env variable `HASURA_GRAPHQL_EXPERIMENTAL_FEATURES` set to `inherited_roles`. Introduction ------------ This PR implements the idea of multiple roles as presented in this [paper](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/FGALanguageICDE07.pdf). The multiple roles feature in this PR can be used via inherited roles. An inherited role is a role which can be created by combining multiple singular roles. For example, if there are two roles `author` and `editor` configured in the graphql-engine, then we can create a inherited role with the name of `combined_author_editor` role which will combine the select permissions of the `author` and `editor` roles and then make GraphQL queries using the `combined_author_editor`. How are select permissions of different roles are combined? ------------------------------------------------------------ A select permission includes 5 things: 1. Columns accessible to the role 2. Row selection filter 3. Limit 4. Allow aggregation 5. Scalar computed fields accessible to the role Suppose there are two roles, `role1` gives access to the `address` column with row filter `P1` and `role2` gives access to both the `address` and the `phone` column with row filter `P2` and we create a new role `combined_roles` which combines `role1` and `role2`. Let's say the following GraphQL query is queried with the `combined_roles` role. ```graphql query { employees { address phone } } ``` This will translate to the following SQL query: ```sql select (case when (P1 or P2) then address else null end) as address, (case when P2 then phone else null end) as phone from employee where (P1 or P2) ``` The other parameters of the select permission will be combined in the following manner: 1. Limit - Minimum of the limits will be the limit of the inherited role 2. Allow aggregations - If any of the role allows aggregation, then the inherited role will allow aggregation 3. Scalar computed fields - same as table column fields, as in the above example APIs for inherited roles: ---------------------- 1. `add_inherited_role` `add_inherited_role` is the [metadata API](https://hasura.io/docs/1.0/graphql/core/api-reference/index.html#schema-metadata-api) to create a new inherited role. It accepts two arguments `role_name`: the name of the inherited role to be added (String) `role_set`: list of roles that need to be combined (Array of Strings) Example: ```json { "type": "add_inherited_role", "args": { "role_name":"combined_user", "role_set":[ "user", "user1" ] } } ``` After adding the inherited role, the inherited role can be used like single roles like earlier Note: An inherited role can only be created with non-inherited/singular roles. 2. `drop_inherited_role` The `drop_inherited_role` API accepts the name of the inherited role and drops it from the metadata. It accepts a single argument: `role_name`: name of the inherited role to be dropped Example: ```json { "type": "drop_inherited_role", "args": { "role_name":"combined_user" } } ``` Metadata --------- The derived roles metadata will be included under the `experimental_features` key while exporting the metadata. ```json { "experimental_features": { "derived_roles": [ { "role_name": "manager_is_employee_too", "role_set": [ "employee", "manager" ] } ] } } ``` Scope ------ Only postgres queries and subscriptions are supported in this PR. Important points: ----------------- 1. All columns exposed to an inherited role will be marked as `nullable`, this is done so that cell value nullification can be done. TODOs ------- - [ ] Tests - [ ] Test a GraphQL query running with a inherited role without enabling inherited roles in experimental features - [] Tests for aggregate queries, limit, computed fields, functions, subscriptions (?) - [ ] Introspection test with a inherited role (nullability changes in a inherited role) - [ ] Docs - [ ] Changelog Co-authored-by: Vamshi Surabhi <6562944+0x777@users.noreply.github.com> GitOrigin-RevId: 3b8ee1e11f5ceca80fe294f8c074d42fbccfec63
2021-03-08 14:14:13 +03:00
maybe NoTop Top mPermLimit
stringifyNumbers =
if num
then StringifyNumbers
else LeaveNumbersAlone
mkNodesSelect :: Args -> Where -> Expression -> Top -> From -> [(Int, (IR.FieldName, [Projection]))] -> [(Int, [Projection])]
mkNodesSelect Args{..} foreignKeyConditions filterExpression permissionBasedTop selectFrom nodes =
[ (index,
[ ExpressionProjection $ Aliased
{ aliasedThing = SelectExpression $ Select
{ selectProjections = projections
, selectTop = permissionBasedTop <> argsTop
, selectFrom = pure selectFrom
, selectJoins = argsJoins
, selectWhere = argsWhere <> Where [filterExpression] <> foreignKeyConditions
, selectFor =
JsonFor ForJson {jsonCardinality = JsonArray, jsonRoot = NoRoot}
, selectOrderBy = argsOrderBy
, selectOffset = argsOffset
}
, aliasedAlias = IR.getFieldNameTxt fieldName
}
] -- singleton
)
| (index, (fieldName, projections)) <- nodes ]
--
-- The idea here is that LIMIT/OFFSET and aggregates don't mix
-- well. Therefore we have a nested query:
--
-- select sum(*), .. FROM (select * from x offset o limit l) p
--
-- That's why @projections@ appears on the outer, and is a
-- @StarProjection@ for the inner. But the joins, conditions, top,
-- offset are on the inner.
--
mkAggregateSelect :: Args -> Where -> From -> [(Int, (IR.FieldName, [Projection]))] -> [(Int, [Projection])]
mkAggregateSelect Args {..} foreignKeyConditions selectFrom aggregates =
[ ( index
, [ ExpressionProjection $
Aliased
{ aliasedThing =
JsonQueryExpression $
SelectExpression $
Select
{ selectProjections = reproject aggSubselectName <$> projections
, selectTop = NoTop
, selectFrom = pure $
FromSelect
Aliased
{ aliasedAlias = aggSubselectName
, aliasedThing =
Select
{ selectProjections = pure StarProjection
, selectTop = argsTop
, selectFrom = pure selectFrom
, selectJoins = argsJoins
, selectWhere = argsWhere <> foreignKeyConditions
, selectFor = NoFor
, selectOrderBy = mempty
, selectOffset = argsOffset
}
}
, selectJoins = mempty
, selectWhere = mempty
, selectFor =
JsonFor
ForJson
{jsonCardinality = JsonSingleton, jsonRoot = NoRoot}
, selectOrderBy = mempty
, selectOffset = Nothing
}
, aliasedAlias = IR.getFieldNameTxt fieldName
}
] -- singleton
)
| (index, (fieldName, projections)) <- aggregates
]
-- | Re-project projections in the aggSubselectName scope
--
-- For example,
--
-- [ AggregateProjection
-- (Aliased {aliasedThing = CountAggregate StarCountable, aliasedAlias = "count"})
-- , AggregateProjection
-- (Aliased
-- { aliasedThing = OpAggregate "sum"
-- [ ColumnExpression
-- (FieldName
-- { fieldName = "id"
-- , fieldNameEntity = "t_person1" -- <<<<< This needs to be `aggSubselectName`
-- })
-- ]
-- , aliasedAlias = "sum"
-- })
--
reproject :: Text -> Projection -> Projection
reproject label = \case
AggregateProjection (Aliased {aliasedThing = OpAggregate aggName expressions, ..}) ->
AggregateProjection (Aliased {aliasedThing = OpAggregate aggName (fixColumnEntity label <$> expressions), ..})
AggregateProjection (Aliased {aliasedThing = CountAggregate countableFieldnames, ..}) ->
AggregateProjection (Aliased {aliasedThing = CountAggregate $ fixEntity label <$> countableFieldnames, ..})
AggregateProjection (Aliased {aliasedThing = JsonQueryOpAggregate aggName expressions, ..}) ->
AggregateProjection (Aliased {aliasedThing = JsonQueryOpAggregate aggName (fixColumnEntity label <$> expressions), ..})
x -> x
where
fixColumnEntity :: Text -> Expression -> Expression
fixColumnEntity entity = \case
ColumnExpression fName ->
ColumnExpression $ fixEntity entity fName
JsonQueryExpression expression ->
JsonQueryExpression $ fixColumnEntity entity expression
SelectExpression Select{..} ->
SelectExpression $ Select {selectProjections = reproject entity <$> selectProjections, ..}
x -> x
fixEntity :: Text -> FieldName -> FieldName
fixEntity entity FieldName{..} = FieldName {fieldNameEntity = entity, ..}
fromSelectAggregate
:: Maybe (EntityAlias, HashMap ColumnName ColumnName)
-> IR.AnnSelectG 'MSSQL (Const Void) (IR.TableAggregateFieldG 'MSSQL (Const Void)) Expression
-> FromIr TSQL.Select
fromSelectAggregate
mparentRelationship
IR.AnnSelectG
{ _asnFields = (zip [0..] -> fields)
, _asnFrom = from
, _asnPerm = IR.TablePerm {_tpLimit = (maybe NoTop Top -> permissionBasedTop), _tpFilter = permFilter}
, _asnArgs = args
, _asnStrfyNum = (bool LeaveNumbersAlone StringifyNumbers -> stringifyNumbers)
}
= do
selectFrom <- case from of
IR.FromTable qualifiedObject -> fromQualifiedTable qualifiedObject
IR.FromFunction {} -> refute $ pure FunctionNotSupported
-- Below: When we're actually a RHS of a query (of CROSS APPLY),
-- then we'll have a LHS table that we're joining on. So we get the
-- conditions expressions from the field mappings. The LHS table is
-- the entityAlias, and the RHS table is selectFrom.
mforeignKeyConditions <- fmap (Where . fromMaybe []) $ for mparentRelationship $
\(entityAlias, mapping) ->
runReaderT (fromMapping selectFrom mapping) entityAlias
filterExpression <- runReaderT (fromAnnBoolExp permFilter) (fromAlias selectFrom)
args'@Args{argsExistingJoins} <-
runReaderT (fromSelectArgsG args) (fromAlias selectFrom)
-- Although aggregates, exps and nodes could be handled in one list,
-- we need to separately treat the subselect expressions
expss :: [(Int, [Projection])] <- flip runReaderT (fromAlias selectFrom) $ sequence $ mapMaybe fromTableExpFieldG fields
nodes :: [(Int, (IR.FieldName, [Projection]))] <-
flip runReaderT (fromAlias selectFrom) $ sequence $ mapMaybe (fromTableNodesFieldG argsExistingJoins stringifyNumbers) fields
aggregates :: [(Int, (IR.FieldName, [Projection]))] <-
flip runReaderT (fromAlias selectFrom) $ sequence $ mapMaybe fromTableAggFieldG fields
pure
Select
{ selectProjections =
concatMap snd $ sortBy (comparing fst) $
expss
<> mkNodesSelect args' mforeignKeyConditions filterExpression permissionBasedTop selectFrom nodes
<> mkAggregateSelect args' mforeignKeyConditions selectFrom aggregates
, selectTop = NoTop
, selectFrom = pure $ FromOpenJson $ Aliased
{ aliasedThing = OpenJson
{ openJsonExpression = ValueExpression $ ODBC.TextValue "[0]"
, openJsonWith = Nothing
}
, aliasedAlias = existsFieldName
}
, selectJoins = mempty -- JOINs and WHEREs are only relevant in subselects
, selectWhere = mempty
, selectFor = JsonFor ForJson {jsonCardinality = JsonSingleton, jsonRoot = NoRoot}
, selectOrderBy = Nothing
, selectOffset = Nothing
}
--------------------------------------------------------------------------------
-- GraphQL Args
data Args = Args
{ argsWhere :: Where
, argsOrderBy :: Maybe (NonEmpty OrderBy)
, argsJoins :: [Join]
, argsTop :: Top
, argsOffset :: Maybe Expression
, argsDistinct :: Proxy (Maybe (NonEmpty FieldName))
, argsExistingJoins :: Map TableName EntityAlias
} deriving (Show)
data UnfurledJoin = UnfurledJoin
{ unfurledJoin :: Join
, unfurledObjectTableAlias :: Maybe (TableName, EntityAlias)
-- ^ Recorded if we joined onto an object relation.
} deriving (Show)
fromSelectArgsG :: IR.SelectArgsG 'MSSQL Expression -> ReaderT EntityAlias FromIr Args
fromSelectArgsG selectArgsG = do
let argsOffset = ValueExpression . ODBC.IntValue . fromIntegral <$> moffset
argsWhere <-
maybe (pure mempty) (fmap (Where . pure) . fromAnnBoolExp) mannBoolExp
argsTop <-
maybe (pure mempty) (pure . Top) mlimit
-- Not supported presently, per Vamshi:
--
-- > It is hardly used and we don't have to go to great lengths to support it.
--
-- But placeholdering the code so that when it's ready to be used,
-- you can just drop the Proxy wrapper.
let argsDistinct = Proxy
(argsOrderBy, joins) <-
runWriterT (traverse fromAnnotatedOrderByItemG (maybe [] toList orders))
-- Any object-relation joins that we generated, we record their
-- generated names into a mapping.
let argsExistingJoins =
M.fromList (mapMaybe unfurledObjectTableAlias (toList joins))
pure
Args
{ argsJoins = toList (fmap unfurledJoin joins)
, argsOrderBy = nonEmpty argsOrderBy
, ..
}
where
IR.SelectArgs { _saWhere = mannBoolExp
, _saLimit = mlimit
, _saOffset = moffset
, _saOrderBy = orders
} = selectArgsG
-- | Produce a valid ORDER BY construct, telling about any joins
-- needed on the side.
fromAnnotatedOrderByItemG
:: IR.AnnotatedOrderByItemG 'MSSQL Expression
-> WriterT (Seq UnfurledJoin) (ReaderT EntityAlias FromIr) OrderBy
fromAnnotatedOrderByItemG IR.OrderByItemG {obiType, obiColumn = obiColumn, obiNulls} = do
(orderByFieldName, orderByType) <- unfurlAnnotatedOrderByElement obiColumn
let orderByNullsOrder = fromMaybe NullsAnyOrder obiNulls
orderByOrder = fromMaybe AscOrder obiType
pure OrderBy {..}
-- | Unfurl the nested set of object relations (tell'd in the writer)
-- that are terminated by field name (IR.AOCColumn and
-- IR.AOCArrayAggregation).
unfurlAnnotatedOrderByElement
:: IR.AnnotatedOrderByElement 'MSSQL Expression
-> WriterT (Seq UnfurledJoin) (ReaderT EntityAlias FromIr) (FieldName, Maybe TSQL.ScalarType)
unfurlAnnotatedOrderByElement =
\case
IR.AOCColumn pgColumnInfo -> do
fieldName <- lift (fromPGColumnInfo pgColumnInfo)
pure
( fieldName
, case (IR.pgiType pgColumnInfo) of
IR.ColumnScalar t -> Just t
-- Above: It is of interest to us whether the type is
-- text/ntext/image. See ToQuery for more explanation.
_ -> Nothing)
IR.AOCObjectRelation IR.RelInfo {riMapping = mapping, riRTable = table} annBoolExp annOrderByElementG -> do
selectFrom <- lift (lift (fromQualifiedTable table))
joinAliasEntity <-
lift (lift (generateEntityAlias (ForOrderAlias (tableNameText table))))
foreignKeyConditions <- lift (fromMapping selectFrom mapping)
-- TODO: Because these object relations are re-used by regular
-- object mapping queries, this WHERE may be unnecessarily
-- restrictive. But I actually don't know from where such an
-- expression arises in the source GraphQL syntax.
--
-- Worst case scenario, we could put the WHERE in the key of the
-- Map in 'argsExistingJoins'. That would guarantee only equal
-- selects are re-used.
whereExpression <-
lift (local (const (fromAlias selectFrom)) (fromAnnBoolExp annBoolExp))
tell
(pure
UnfurledJoin
{ unfurledJoin =
Join
{ joinSource =
JoinSelect
Select
{ selectTop = NoTop
, selectProjections = [StarProjection]
, selectFrom = Just selectFrom
, selectJoins = []
, selectWhere =
Where (foreignKeyConditions <> [whereExpression])
, selectFor = NoFor
, selectOrderBy = Nothing
, selectOffset = Nothing
}
, joinJoinAlias =
JoinAlias {joinAliasEntity, joinAliasField = Nothing}
}
, unfurledObjectTableAlias = Just (table, EntityAlias joinAliasEntity)
})
local
(const (EntityAlias joinAliasEntity))
(unfurlAnnotatedOrderByElement annOrderByElementG)
IR.AOCArrayAggregation IR.RelInfo {riMapping = mapping, riRTable = tableName} annBoolExp annAggregateOrderBy -> do
selectFrom <- lift (lift (fromQualifiedTable tableName))
let alias = aggFieldName
joinAliasEntity <-
lift (lift (generateEntityAlias (ForOrderAlias (tableNameText tableName))))
foreignKeyConditions <- lift (fromMapping selectFrom mapping)
whereExpression <-
lift (local (const (fromAlias selectFrom)) (fromAnnBoolExp annBoolExp))
aggregate <-
lift
(local
(const (fromAlias selectFrom))
(case annAggregateOrderBy of
IR.AAOCount -> pure (CountAggregate StarCountable)
IR.AAOOp text pgColumnInfo -> do
fieldName <- fromPGColumnInfo pgColumnInfo
pure (OpAggregate text (pure (ColumnExpression fieldName)))))
tell
(pure
(UnfurledJoin
{ unfurledJoin =
Join
{ joinSource =
JoinSelect
Select
{ selectTop = NoTop
, selectProjections =
[ AggregateProjection
Aliased
{ aliasedThing = aggregate
, aliasedAlias = alias
}
]
, selectFrom = Just selectFrom
, selectJoins = []
, selectWhere =
Where
(foreignKeyConditions <> [whereExpression])
, selectFor = NoFor
, selectOrderBy = Nothing
, selectOffset = Nothing
}
, joinJoinAlias =
JoinAlias {joinAliasEntity, joinAliasField = Nothing}
}
, unfurledObjectTableAlias = Nothing
}))
pure
( FieldName {fieldNameEntity = joinAliasEntity, fieldName = alias}
, Nothing)
--------------------------------------------------------------------------------
-- Conversion functions
tableNameText :: TableName -> Text
tableNameText (TableName {tableName}) = tableName
-- | This is really the start where you query the base table,
-- everything else is joins attached to it.
fromQualifiedTable :: TableName -> FromIr From
fromQualifiedTable schemadTableName@(TableName{tableName}) = do
alias <- generateEntityAlias (TableTemplate tableName)
pure
(FromQualifiedTable
(Aliased
{ aliasedThing = schemadTableName
, aliasedAlias = alias
}))
fromTableName :: TableName -> FromIr EntityAlias
fromTableName TableName{tableName} = do
alias <- generateEntityAlias (TableTemplate tableName)
pure (EntityAlias alias)
fromAnnBoolExp
:: IR.GBoolExp 'MSSQL (IR.AnnBoolExpFld 'MSSQL Expression)
-> ReaderT EntityAlias FromIr Expression
fromAnnBoolExp = traverse fromAnnBoolExpFld >=> fromGBoolExp
fromAnnBoolExpFld
:: IR.AnnBoolExpFld 'MSSQL Expression
-> ReaderT EntityAlias FromIr Expression
fromAnnBoolExpFld =
\case
IR.AVColumn pgColumnInfo opExpGs -> do
expression <- fromColumnInfoForBoolExp pgColumnInfo
expressions <- traverse (lift . fromOpExpG expression) opExpGs
pure (AndExpression expressions)
IR.AVRelationship IR.RelInfo {riMapping = mapping, riRTable = table} annBoolExp -> do
selectFrom <- lift (fromQualifiedTable table)
foreignKeyConditions <- fromMapping selectFrom mapping
whereExpression <-
local (const (fromAlias selectFrom)) (fromAnnBoolExp annBoolExp)
pure
(ExistsExpression
Select
{ selectOrderBy = Nothing
, selectProjections =
[ ExpressionProjection
(Aliased
{ aliasedThing = trueExpression
, aliasedAlias = existsFieldName
})
]
, selectFrom = Just selectFrom
, selectJoins = mempty
, selectWhere = Where (foreignKeyConditions <> [whereExpression])
, selectTop = NoTop
, selectFor = NoFor
, selectOffset = Nothing
})
-- | For boolean operators, various comparison operators used need
-- special handling to ensure that SQL Server won't outright reject
-- the comparison. See also 'shouldCastToVarcharMax'.
fromColumnInfoForBoolExp :: IR.ColumnInfo 'MSSQL -> ReaderT EntityAlias FromIr Expression
fromColumnInfoForBoolExp IR.ColumnInfo {pgiColumn = pgCol, pgiType} = do
fieldName <- columnNameToFieldName pgCol <$> ask
if shouldCastToVarcharMax pgiType -- See function commentary.
then pure (CastExpression (ColumnExpression fieldName) "VARCHAR(MAX)")
else pure (ColumnExpression fieldName)
-- | There's a problem of comparing text fields with =, <, etc. that
-- SQL Server completely refuses to do so. So one way to workaround
-- this restriction is to automatically cast such text fields to
-- varchar(max).
shouldCastToVarcharMax :: IR.ColumnType 'MSSQL -> Bool
shouldCastToVarcharMax typ =
typ == IR.ColumnScalar TextType || typ == IR.ColumnScalar WtextType
fromPGColumnInfo :: IR.ColumnInfo 'MSSQL -> ReaderT EntityAlias FromIr FieldName
fromPGColumnInfo IR.ColumnInfo {pgiColumn = pgCol} =
columnNameToFieldName pgCol <$> ask
-- entityAlias <- ask
-- pure
-- (columnNameToFieldName pgCol entityAlias
-- FieldName
-- {fieldName = PG.getPGColTxt pgCol, fieldNameEntity = entityAliasText})
fromGExists :: IR.GExists 'MSSQL Expression -> ReaderT EntityAlias FromIr Select
fromGExists IR.GExists {_geTable, _geWhere} = do
selectFrom <- lift (fromQualifiedTable _geTable)
whereExpression <-
local (const (fromAlias selectFrom)) (fromGBoolExp _geWhere)
pure
Select
{ selectOrderBy = Nothing
, selectProjections =
[ ExpressionProjection
(Aliased
{ aliasedThing = trueExpression
, aliasedAlias = existsFieldName
})
]
, selectFrom = Just selectFrom
, selectJoins = mempty
, selectWhere = Where [whereExpression]
, selectTop = NoTop
, selectFor = NoFor
, selectOffset = Nothing
}
--------------------------------------------------------------------------------
-- Sources of projected fields
--
-- Because in the IR, a field projected can be a foreign object, we
-- have to both generate a projection AND on the side generate a join.
--
-- So a @FieldSource@ couples the idea of the projected thing and the
-- source of it (via 'Aliased').
data FieldSource
= ExpressionFieldSource (Aliased Expression)
| JoinFieldSource (Aliased Join)
| AggregateFieldSource [Aliased Aggregate]
deriving (Eq, Show)
-- | Get FieldSource from a TAFExp type table aggregate field
fromTableExpFieldG :: -- TODO: Convert function to be similar to Nodes function
(Int, (IR.FieldName, IR.TableAggregateFieldG 'MSSQL (Const Void) Expression)) ->
Maybe (ReaderT EntityAlias FromIr (Int, [Projection]))
fromTableExpFieldG = \case
(index, (IR.FieldName name, IR.TAFExp text)) -> Just $
pure $
(index, fieldSourceProjections $
ExpressionFieldSource
Aliased
{ aliasedThing = TSQL.ValueExpression (ODBC.TextValue text)
, aliasedAlias = name
})
_ -> Nothing
fromTableAggFieldG ::
(Int, (IR.FieldName, IR.TableAggregateFieldG 'MSSQL (Const Void) Expression)) ->
Maybe (ReaderT EntityAlias FromIr (Int, (IR.FieldName, [Projection])))
fromTableAggFieldG = \case
(index, (fieldName, IR.TAFAgg (aggregateFields :: [(IR.FieldName, IR.AggregateField 'MSSQL)]))) -> Just do
aggregates <-
for aggregateFields \(fieldName', aggregateField) ->
fromAggregateField aggregateField <&> \aliasedThing ->
Aliased {aliasedAlias = IR.getFieldNameTxt fieldName', ..}
pure (index, (fieldName, fieldSourceProjections $ AggregateFieldSource aggregates))
_ -> Nothing
fromTableNodesFieldG ::
Map TableName EntityAlias ->
StringifyNumbers ->
(Int, (IR.FieldName, IR.TableAggregateFieldG 'MSSQL (Const Void) Expression)) ->
Maybe (ReaderT EntityAlias FromIr (Int, (IR.FieldName, [Projection])))
fromTableNodesFieldG argsExistingJoins stringifyNumbers = \case
(index, (fieldName, IR.TAFNodes () (annFieldsG :: [(IR.FieldName, IR.AnnFieldG 'MSSQL (Const Void) Expression)]))) -> Just do
fieldSources' <- fromAnnFieldsG argsExistingJoins stringifyNumbers `traverse` annFieldsG
let nodesProjections' :: [Projection] = concatMap fieldSourceProjections fieldSources'
pure (index, (fieldName, nodesProjections'))
_ -> Nothing
fromAggregateField :: IR.AggregateField 'MSSQL -> ReaderT EntityAlias FromIr Aggregate
fromAggregateField aggregateField =
case aggregateField of
IR.AFExp text -> pure (TextAggregate text)
IR.AFCount countType -> CountAggregate <$> case countType of
StarCountable -> pure StarCountable
NonNullFieldCountable names -> NonNullFieldCountable <$> traverse fromPGCol names
DistinctCountable names -> DistinctCountable <$> traverse fromPGCol names
IR.AFOp IR.AggregateOp {_aoOp = op, _aoFields = fields} -> do
projections :: [Projection] <- for fields \(fieldName, pgColFld) ->
case pgColFld of
IR.CFCol pgCol _pgType -> do
fname <- fromPGCol pgCol
pure $ AggregateProjection $ Aliased (JsonQueryOpAggregate op [ColumnExpression fname]) (IR.getFieldNameTxt fieldName)
IR.CFExp text -> do
pure $ ExpressionProjection $ Aliased (ValueExpression (ODBC.TextValue text)) (IR.getFieldNameTxt fieldName)
pure $ OpAggregate op $
[ JsonQueryExpression $ SelectExpression $
emptySelect
{ selectProjections = projections
, selectFor = JsonFor $ ForJson JsonSingleton NoRoot
}
]
-- | The main sources of fields, either constants, fields or via joins.
fromAnnFieldsG ::
Map TableName EntityAlias
-> StringifyNumbers
-> (IR.FieldName, IR.AnnFieldG 'MSSQL (Const Void) Expression)
-> ReaderT EntityAlias FromIr FieldSource
fromAnnFieldsG existingJoins stringifyNumbers (IR.FieldName name, field) =
case field of
IR.AFColumn annColumnField -> do
expression <- fromAnnColumnField stringifyNumbers annColumnField
pure
(ExpressionFieldSource
Aliased {aliasedThing = expression, aliasedAlias = name})
IR.AFExpression text ->
pure
(ExpressionFieldSource
Aliased
{ aliasedThing = TSQL.ValueExpression (ODBC.TextValue text)
, aliasedAlias = name
})
IR.AFObjectRelation objectRelationSelectG ->
fmap
(\aliasedThing ->
JoinFieldSource (Aliased {aliasedThing, aliasedAlias = name}))
(fromObjectRelationSelectG existingJoins objectRelationSelectG)
IR.AFArrayRelation arraySelectG ->
fmap
(\aliasedThing ->
JoinFieldSource (Aliased {aliasedThing, aliasedAlias = name}))
(fromArraySelectG arraySelectG)
-- | Here is where we project a field as a column expression. If
-- number stringification is on, then we wrap it in a
-- 'ToStringExpression' so that it's casted when being projected.
fromAnnColumnField
:: StringifyNumbers
[Preview] Inherited roles for postgres read queries fixes #3868 docker image - `hasura/graphql-engine:inherited-roles-preview-48b73a2de` Note: To be able to use the inherited roles feature, the graphql-engine should be started with the env variable `HASURA_GRAPHQL_EXPERIMENTAL_FEATURES` set to `inherited_roles`. Introduction ------------ This PR implements the idea of multiple roles as presented in this [paper](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/FGALanguageICDE07.pdf). The multiple roles feature in this PR can be used via inherited roles. An inherited role is a role which can be created by combining multiple singular roles. For example, if there are two roles `author` and `editor` configured in the graphql-engine, then we can create a inherited role with the name of `combined_author_editor` role which will combine the select permissions of the `author` and `editor` roles and then make GraphQL queries using the `combined_author_editor`. How are select permissions of different roles are combined? ------------------------------------------------------------ A select permission includes 5 things: 1. Columns accessible to the role 2. Row selection filter 3. Limit 4. Allow aggregation 5. Scalar computed fields accessible to the role Suppose there are two roles, `role1` gives access to the `address` column with row filter `P1` and `role2` gives access to both the `address` and the `phone` column with row filter `P2` and we create a new role `combined_roles` which combines `role1` and `role2`. Let's say the following GraphQL query is queried with the `combined_roles` role. ```graphql query { employees { address phone } } ``` This will translate to the following SQL query: ```sql select (case when (P1 or P2) then address else null end) as address, (case when P2 then phone else null end) as phone from employee where (P1 or P2) ``` The other parameters of the select permission will be combined in the following manner: 1. Limit - Minimum of the limits will be the limit of the inherited role 2. Allow aggregations - If any of the role allows aggregation, then the inherited role will allow aggregation 3. Scalar computed fields - same as table column fields, as in the above example APIs for inherited roles: ---------------------- 1. `add_inherited_role` `add_inherited_role` is the [metadata API](https://hasura.io/docs/1.0/graphql/core/api-reference/index.html#schema-metadata-api) to create a new inherited role. It accepts two arguments `role_name`: the name of the inherited role to be added (String) `role_set`: list of roles that need to be combined (Array of Strings) Example: ```json { "type": "add_inherited_role", "args": { "role_name":"combined_user", "role_set":[ "user", "user1" ] } } ``` After adding the inherited role, the inherited role can be used like single roles like earlier Note: An inherited role can only be created with non-inherited/singular roles. 2. `drop_inherited_role` The `drop_inherited_role` API accepts the name of the inherited role and drops it from the metadata. It accepts a single argument: `role_name`: name of the inherited role to be dropped Example: ```json { "type": "drop_inherited_role", "args": { "role_name":"combined_user" } } ``` Metadata --------- The derived roles metadata will be included under the `experimental_features` key while exporting the metadata. ```json { "experimental_features": { "derived_roles": [ { "role_name": "manager_is_employee_too", "role_set": [ "employee", "manager" ] } ] } } ``` Scope ------ Only postgres queries and subscriptions are supported in this PR. Important points: ----------------- 1. All columns exposed to an inherited role will be marked as `nullable`, this is done so that cell value nullification can be done. TODOs ------- - [ ] Tests - [ ] Test a GraphQL query running with a inherited role without enabling inherited roles in experimental features - [] Tests for aggregate queries, limit, computed fields, functions, subscriptions (?) - [ ] Introspection test with a inherited role (nullability changes in a inherited role) - [ ] Docs - [ ] Changelog Co-authored-by: Vamshi Surabhi <6562944+0x777@users.noreply.github.com> GitOrigin-RevId: 3b8ee1e11f5ceca80fe294f8c074d42fbccfec63
2021-03-08 14:14:13 +03:00
-> IR.AnnColumnField 'MSSQL Expression
-> ReaderT EntityAlias FromIr Expression
fromAnnColumnField _stringifyNumbers annColumnField = do
fieldName <- fromPGCol pgCol
-- TODO: Handle stringifying large numbers
{-(IR.isScalarColumnWhere PG.isBigNum typ && stringifyNumbers == StringifyNumbers)-}
-- for geometry and geography values, the automatic json encoding on sql
-- server would fail. So we need to convert it to a format the json encoding
-- handles. Ideally we want this representation to be GeoJSON but sql server
-- doesn't have any functions to convert to GeoJSON format. So we return it in
-- WKT format
if typ == (IR.ColumnScalar GeometryType) || typ == (IR.ColumnScalar GeographyType)
then pure $ MethodExpression (ColumnExpression fieldName) "STAsText" []
else case caseBoolExpMaybe of
Nothing -> pure (ColumnExpression fieldName)
Just ex -> do
ex' <- (traverse fromAnnBoolExpFld >=> fromGBoolExp) (coerce ex)
pure (ConditionalProjection ex' fieldName)
where
IR.AnnColumnField { _acfInfo = IR.ColumnInfo{pgiColumn=pgCol,pgiType=typ}
, _acfAsText = _asText :: Bool
, _acfOp = _ :: Maybe (IR.ColumnOp 'MSSQL) -- TODO: What's this?
, _acfCaseBoolExpression = caseBoolExpMaybe
} = annColumnField
-- | This is where a field name "foo" is resolved to a fully qualified
-- field name [table].[foo]. The table name comes from EntityAlias in
-- the ReaderT.
fromPGCol :: ColumnName -> ReaderT EntityAlias FromIr FieldName
fromPGCol pgCol = columnNameToFieldName pgCol <$> ask
-- entityAlias <- ask
-- pure (columnNameToFieldName pgCol entityAlias -- FieldName {fieldName = PG.getPGColTxt pgCol, fieldNameEntity = entityAliasText}
-- )
fieldSourceProjections :: FieldSource -> [Projection]
fieldSourceProjections =
\case
ExpressionFieldSource aliasedExpression ->
pure (ExpressionProjection aliasedExpression)
JoinFieldSource aliasedJoin ->
pure
(ExpressionProjection
(aliasedJoin
{ aliasedThing =
-- Basically a cast, to ensure that SQL Server won't
-- double-encode the JSON but will "pass it through"
-- untouched.
JsonQueryExpression
(ColumnExpression
(joinAliasToField
(joinJoinAlias (aliasedThing aliasedJoin))))
}))
AggregateFieldSource aggregates -> fmap AggregateProjection aggregates
joinAliasToField :: JoinAlias -> FieldName
joinAliasToField JoinAlias {..} =
FieldName
{ fieldNameEntity = joinAliasEntity
, fieldName = fromMaybe (error "TODO: Eliminate this case. joinAliasToField") joinAliasField
}
fieldSourceJoin :: FieldSource -> Maybe Join
fieldSourceJoin =
\case
JoinFieldSource aliasedJoin -> pure (aliasedThing aliasedJoin)
ExpressionFieldSource {} -> Nothing
AggregateFieldSource {} -> Nothing
--------------------------------------------------------------------------------
-- Joins
fromObjectRelationSelectG ::
Map TableName {-PG.QualifiedTable-} EntityAlias
-> IR.ObjectRelationSelectG 'MSSQL (Const Void) Expression
-> ReaderT EntityAlias FromIr Join
fromObjectRelationSelectG existingJoins annRelationSelectG = do
eitherAliasOrFrom <- lift (lookupTableFrom existingJoins tableFrom)
let entityAlias :: EntityAlias = either id fromAlias eitherAliasOrFrom
fieldSources <-
local
(const entityAlias)
(traverse (fromAnnFieldsG mempty LeaveNumbersAlone) fields)
let selectProjections =
concatMap (toList . fieldSourceProjections) fieldSources
joinJoinAlias <-
do fieldName <- lift (fromRelName aarRelationshipName)
alias <- lift (generateEntityAlias (ObjectRelationTemplate fieldName))
pure
JoinAlias
{joinAliasEntity = alias, joinAliasField = pure jsonFieldName}
let selectFor =
JsonFor ForJson {jsonCardinality = JsonSingleton, jsonRoot = NoRoot}
filterExpression <- local (const entityAlias) (fromAnnBoolExp tableFilter)
case eitherAliasOrFrom of
Right selectFrom -> do
foreignKeyConditions <- fromMapping selectFrom mapping
pure
Join
{ joinJoinAlias
, joinSource =
JoinSelect
Select
{ selectOrderBy = Nothing
, selectTop = NoTop
, selectProjections
, selectFrom = Just selectFrom
, selectJoins = mapMaybe fieldSourceJoin fieldSources
, selectWhere =
Where (foreignKeyConditions <> [filterExpression])
, selectFor
, selectOffset = Nothing
}
}
Left _entityAlias ->
pure
Join
{ joinJoinAlias
, joinSource =
JoinReselect
Reselect
{ reselectProjections = selectProjections
, reselectFor = selectFor
, reselectWhere = Where [filterExpression]
}
}
where
IR.AnnObjectSelectG { _aosFields = fields :: IR.AnnFieldsG 'MSSQL (Const Void) Expression
, _aosTableFrom = tableFrom :: TableName{-PG.QualifiedTable-}
, _aosTableFilter = tableFilter :: IR.AnnBoolExp 'MSSQL Expression
} = annObjectSelectG
IR.AnnRelationSelectG { aarRelationshipName
, aarColumnMapping = mapping :: HashMap ColumnName ColumnName -- PG.PGCol PG.PGCol
, aarAnnSelect = annObjectSelectG :: IR.AnnObjectSelectG 'MSSQL (Const Void) Expression
} = annRelationSelectG
lookupTableFrom ::
Map TableName {-PG.QualifiedTable-} EntityAlias
-> {-PG.QualifiedTable-}TableName
-> FromIr (Either EntityAlias From)
lookupTableFrom existingJoins tableFrom = do
case M.lookup tableFrom existingJoins of
Just entityAlias -> pure (Left entityAlias)
Nothing -> fmap Right (fromQualifiedTable tableFrom)
fromArraySelectG :: IR.ArraySelectG 'MSSQL (Const Void) Expression -> ReaderT EntityAlias FromIr Join
fromArraySelectG =
\case
IR.ASSimple arrayRelationSelectG ->
fromArrayRelationSelectG arrayRelationSelectG
IR.ASAggregate arrayAggregateSelectG ->
fromArrayAggregateSelectG arrayAggregateSelectG
fromArrayAggregateSelectG
:: IR.AnnRelationSelectG 'MSSQL (IR.AnnAggregateSelectG 'MSSQL (Const Void) Expression)
-> ReaderT EntityAlias FromIr Join
fromArrayAggregateSelectG annRelationSelectG = do
fieldName <- lift (fromRelName aarRelationshipName)
joinSelect <- do
lhsEntityAlias <- ask
-- With this, the foreign key relations are injected automatically
-- at the right place by fromSelectAggregate.
lift (fromSelectAggregate (pure (lhsEntityAlias, mapping)) annSelectG)
alias <- lift (generateEntityAlias (ArrayAggregateTemplate fieldName))
pure
Join
{ joinJoinAlias =
JoinAlias
{joinAliasEntity = alias, joinAliasField = pure jsonFieldName}
, joinSource = JoinSelect joinSelect
}
where
IR.AnnRelationSelectG { aarRelationshipName
, aarColumnMapping = mapping :: HashMap ColumnName ColumnName
, aarAnnSelect = annSelectG
} = annRelationSelectG
fromArrayRelationSelectG :: IR.ArrayRelationSelectG 'MSSQL (Const Void) Expression -> ReaderT EntityAlias FromIr Join
fromArrayRelationSelectG annRelationSelectG = do
fieldName <- lift (fromRelName aarRelationshipName)
sel <- lift (fromSelectRows annSelectG)
joinSelect <-
do foreignKeyConditions <- selectFromMapping sel mapping
pure
sel {selectWhere = Where foreignKeyConditions <> selectWhere sel}
alias <- lift (generateEntityAlias (ArrayRelationTemplate fieldName))
pure
Join
{ joinJoinAlias =
JoinAlias
{joinAliasEntity = alias, joinAliasField = pure jsonFieldName}
, joinSource = JoinSelect joinSelect
}
where
IR.AnnRelationSelectG { aarRelationshipName
, aarColumnMapping = mapping :: HashMap ColumnName ColumnName-- PG.PGCol PG.PGCol
, aarAnnSelect = annSelectG
} = annRelationSelectG
fromRelName :: IR.RelName -> FromIr Text
fromRelName relName =
pure (IR.relNameToTxt relName)
-- | The context given by the reader is of the previous/parent
-- "remote" table. The WHERE that we're generating goes in the child,
-- "local" query. The @From@ passed in as argument is the local table.
--
-- We should hope to see e.g. "post.category = category.id" for a
-- local table of post and a remote table of category.
--
-- The left/right columns in @HashMap PG.PGCol PG.PGCol@ corresponds
-- to the left/right of @select ... join ...@. Therefore left=remote,
-- right=local in this context.
fromMapping ::
From
-> HashMap ColumnName ColumnName-- PG.PGCol PG.PGCol
-> ReaderT EntityAlias FromIr [Expression]
fromMapping localFrom =
traverse
(\(remotePgCol, localPgCol) -> do
localFieldName <- local (const (fromAlias localFrom)) (fromPGCol localPgCol)
remoteFieldName <- fromPGCol remotePgCol
pure
(OpExpression TSQL.EQ'
(ColumnExpression localFieldName)
(ColumnExpression remoteFieldName))) .
HM.toList
selectFromMapping :: Select
-> HashMap ColumnName ColumnName
-> ReaderT EntityAlias FromIr [Expression]
selectFromMapping Select {selectFrom = Nothing } = const (pure [])
selectFromMapping Select {selectFrom = Just from} = fromMapping from
--------------------------------------------------------------------------------
-- Basic SQL expression types
fromOpExpG :: Expression -> IR.OpExpG 'MSSQL Expression -> FromIr Expression
fromOpExpG expression op =
case op of
IR.ANISNULL -> pure $ TSQL.IsNullExpression expression
IR.ANISNOTNULL -> pure $ TSQL.IsNotNullExpression expression
IR.AEQ False val -> pure $ nullableBoolEquality expression val
IR.AEQ True val -> pure $ OpExpression TSQL.EQ' expression val
IR.ANE False val -> pure $ nullableBoolInequality expression val
IR.ANE True val -> pure $ OpExpression TSQL.NEQ' expression val
IR.AGT val -> pure $ OpExpression TSQL.GT expression val
IR.ALT val -> pure $ OpExpression TSQL.LT expression val
IR.AGTE val -> pure $ OpExpression TSQL.GTE expression val
IR.ALTE val -> pure $ OpExpression TSQL.LTE expression val
IR.AIN val -> pure $ OpExpression TSQL.IN expression val
IR.ANIN val -> pure $ OpExpression TSQL.NIN expression val
IR.ALIKE val -> pure $ OpExpression TSQL.LIKE expression val
IR.ANLIKE val -> pure $ OpExpression TSQL.NLIKE expression val
IR.ABackendSpecific o -> case o of
ASTContains val -> pure $ TSQL.STOpExpression TSQL.STContains expression val
ASTCrosses val -> pure $ TSQL.STOpExpression TSQL.STCrosses expression val
ASTEquals val -> pure $ TSQL.STOpExpression TSQL.STEquals expression val
ASTIntersects val -> pure $ TSQL.STOpExpression TSQL.STIntersects expression val
ASTOverlaps val -> pure $ TSQL.STOpExpression TSQL.STOverlaps expression val
ASTTouches val -> pure $ TSQL.STOpExpression TSQL.STTouches expression val
ASTWithin val -> pure $ TSQL.STOpExpression TSQL.STWithin expression val
-- As of March 2021, only geometry/geography casts are supported
IR.ACast _casts -> refute (pure (UnsupportedOpExpG op)) -- mkCastsExp casts
-- We do not yet support column names in permissions
IR.CEQ _rhsCol -> refute (pure (UnsupportedOpExpG op)) -- S.BECompare S.SEQ lhs $ mkQCol rhsCol
IR.CNE _rhsCol -> refute (pure (UnsupportedOpExpG op)) -- S.BECompare S.SNE lhs $ mkQCol rhsCol
IR.CGT _rhsCol -> refute (pure (UnsupportedOpExpG op)) -- S.BECompare S.SGT lhs $ mkQCol rhsCol
IR.CLT _rhsCol -> refute (pure (UnsupportedOpExpG op)) -- S.BECompare S.SLT lhs $ mkQCol rhsCol
IR.CGTE _rhsCol -> refute (pure (UnsupportedOpExpG op)) -- S.BECompare S.SGTE lhs $ mkQCol rhsCol
IR.CLTE _rhsCol -> refute (pure (UnsupportedOpExpG op)) -- S.BECompare S.SLTE lhs $ mkQCol rhsCol
nullableBoolEquality :: Expression -> Expression -> Expression
nullableBoolEquality x y =
OrExpression
[ OpExpression TSQL.EQ' x y
, AndExpression [IsNullExpression x, IsNullExpression y]
]
nullableBoolInequality :: Expression -> Expression -> Expression
nullableBoolInequality x y =
OrExpression
[ OpExpression TSQL.NEQ' x y
, AndExpression [IsNotNullExpression x, IsNullExpression y]
]
fromGBoolExp :: IR.GBoolExp 'MSSQL Expression -> ReaderT EntityAlias FromIr Expression
fromGBoolExp =
\case
IR.BoolAnd expressions ->
fmap AndExpression (traverse fromGBoolExp expressions)
IR.BoolOr expressions ->
fmap OrExpression (traverse fromGBoolExp expressions)
IR.BoolNot expression -> fmap NotExpression (fromGBoolExp expression)
IR.BoolExists gExists -> fmap ExistsExpression (fromGExists gExists)
IR.BoolFld expression -> pure expression
--------------------------------------------------------------------------------
-- Delete
fromDelete :: IR.AnnDel 'MSSQL -> FromIr Delete
fromDelete (IR.AnnDel tableName (permFilter, whereClause) _ _) = do
tableAlias <- fromTableName tableName
runReaderT
(do permissionsFilter <- fromAnnBoolExp permFilter
whereExpression <- fromAnnBoolExp whereClause
pure
Delete
{ deleteTable =
Aliased
{ aliasedAlias = entityAliasText tableAlias
, aliasedThing = tableName
}
, deleteWhere = Where [permissionsFilter, whereExpression]
})
tableAlias
--------------------------------------------------------------------------------
-- Misc combinators
trueExpression :: Expression
trueExpression = ValueExpression (ODBC.BoolValue True)
--------------------------------------------------------------------------------
-- Constants
jsonFieldName :: Text
jsonFieldName = "json"
aggFieldName :: Text
aggFieldName = "agg"
aggSubselectName :: Text
aggSubselectName = "agg_sub"
existsFieldName :: Text
existsFieldName = "exists_placeholder"
--------------------------------------------------------------------------------
-- Name generation
data NameTemplate
= ArrayRelationTemplate Text
| ArrayAggregateTemplate Text
| ObjectRelationTemplate Text
| TableTemplate Text
| ForOrderAlias Text
generateEntityAlias :: NameTemplate -> FromIr Text
generateEntityAlias template = do
FromIr (modify' (M.insertWith (+) prefix start))
i <- FromIr get
pure (prefix <> tshow (fromMaybe start (M.lookup prefix i)))
where
start = 1
prefix = T.take 20 rendered
rendered =
case template of
ArrayRelationTemplate sample -> "ar_" <> sample
ArrayAggregateTemplate sample -> "aa_" <> sample
ObjectRelationTemplate sample -> "or_" <> sample
TableTemplate sample -> "t_" <> sample
ForOrderAlias sample -> "order_" <> sample
fromAlias :: From -> EntityAlias
fromAlias (FromQualifiedTable Aliased {aliasedAlias}) = EntityAlias aliasedAlias
fromAlias (FromOpenJson Aliased {aliasedAlias}) = EntityAlias aliasedAlias
fromAlias (FromSelect Aliased {aliasedAlias}) = EntityAlias aliasedAlias
columnNameToFieldName :: ColumnName -> EntityAlias -> FieldName
columnNameToFieldName (ColumnName fieldName) EntityAlias {entityAliasText = fieldNameEntity} =
FieldName {fieldName, fieldNameEntity}