graphql-engine/server/src-lib/Hasura/RQL/DDL/Schema/Cache/Permission.hs

332 lines
15 KiB
Haskell
Raw Normal View History

[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
{-# LANGUAGE Arrows #-}
{-# LANGUAGE ViewPatterns #-}
module Hasura.RQL.DDL.Schema.Cache.Permission
( buildTablePermissions
, mkPermissionMetadataObject
, mkRemoteSchemaPermissionMetadataObject
, orderRoles
) where
import Hasura.Prelude
import qualified Data.Graph as G
import qualified Data.HashMap.Strict as M
import qualified Data.HashMap.Strict.Extended as M
import qualified Data.HashSet as Set
import qualified Data.List.NonEmpty as NE
import qualified Data.Sequence as Seq
[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
import Control.Arrow.Extended
import Data.Aeson
import Data.Proxy
import Data.Text.Extended
import qualified Hasura.Incremental as Inc
import qualified Hasura.SQL.AnyBackend as AB
import Hasura.Base.Error
import Hasura.RQL.DDL.Permission
import Hasura.RQL.DDL.Schema.Cache.Common
import Hasura.RQL.Types
[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
import Hasura.Server.Types
import Hasura.Session
[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
{- Note: [Inherited roles architecture for postgres read queries]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1. Schema generation
--------------------
Schema generation for inherited roles is similar to the schema
generation of non-inherited roles. In the case of inherited roles,
we combine the `SelectPermInfo`s (see `combineSelectPermInfos`) of the
inherited role's role set and a new `SelectPermInfo` will be generated
which will be the select permission of the inherited role.
2. SQL generation
-----------------
See note [SQL generation for inherited roles]
3. Introspection
----------------
The columns accessible to an inherited role are explicitly set to
nullable irrespective of the nullability of the DB column to accomodate
cell value nullification.
-}
-- | This type is only used in the `combineSelectPermInfos` for
-- combining select permissions efficiently
data CombinedSelPermInfo (b :: BackendType)
= CombinedSelPermInfo
{ cspiCols :: ![(M.HashMap (Column b) (Maybe (AnnColumnCaseBoolExpPartialSQL b)))]
, cspiScalarComputedFields :: ![(M.HashMap ComputedFieldName (Maybe (AnnColumnCaseBoolExpPartialSQL b)))]
, cspiFilter :: ![(AnnBoolExpPartialSQL b)]
, cspiLimit :: !(Maybe Int)
, cspiAllowAgg :: !Bool
, cspiRequiredHeaders :: !(Set.HashSet Text)
}
[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
-- | combineSelectPermInfos combines multiple `SelPermInfo`s
-- into one `SelPermInfo`. Two `SelPermInfo` will
-- be combined in the following manner:
--
-- 1. Columns - The `SelPermInfo` contains a hashset of the columns that are
-- accessible to the role. To combine two `SelPermInfo`s, every column of the
-- hashset is coupled with the boolean expression (filter) of the `SelPermInfo`
-- and a hash map of all the columns is created out of it, this hashmap is
-- generated for the `SelPermInfo`s that are going to be combined. These hashmaps
-- are then unioned and the values of these hashmaps are `OR`ed. When a column
-- is accessible to all the select permissions then the nullability of the column
-- is inferred from the DB column otherwise the column is explicitly marked as
-- nullable to accomodate cell-value nullification.
-- 2. Scalar computed fields - Scalar computed fields work the same as Columns (#1)
-- 3. Filter / Boolean expression - The filters are combined using a `BoolOr`
-- 4. Limit - Limits are combined by taking the maximum of the two limits
[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
-- 5. Allow Aggregation - Aggregation is allowed, if any of the permissions allow it.
-- 6. Request Headers - Request headers are concatenated
--
-- To maintain backwards compatibility, we handle the case of single select permission
-- differently i.e. we don't want the case statements that always evaluate to true with
-- the columns
--
combineSelectPermInfos
:: forall b
. (Backend b)
=> NE.NonEmpty (SelPermInfo b)
-> SelPermInfo b
combineSelectPermInfos (headSelPermInfo NE.:| []) = headSelPermInfo
combineSelectPermInfos selPermInfos@(headSelPermInfo NE.:| restSelPermInfos) =
let CombinedSelPermInfo {..}
= foldr combine (modifySingleSelectPerm headSelPermInfo) restSelPermInfos
mergeColumnsWithBoolExp xs
| length selPermInfos == length xs = Nothing
| otherwise = foldr combineCaseBoolExps Nothing xs
in SelPermInfo (mergeColumnsWithBoolExp <$> M.unionsAll cspiCols)
(mergeColumnsWithBoolExp <$> M.unionsAll cspiScalarComputedFields)
(BoolOr cspiFilter)
cspiLimit
cspiAllowAgg
(toList cspiRequiredHeaders)
where
modifySingleSelectPerm :: SelPermInfo b -> CombinedSelPermInfo b
modifySingleSelectPerm SelPermInfo {..} =
let columnCaseBoolExp = fmap AnnColumnCaseBoolExpField spiFilter
colsWithColCaseBoolExp = spiCols $> Just columnCaseBoolExp
scalarCompFieldsWithColCaseBoolExp = spiScalarComputedFields $> Just columnCaseBoolExp
in
CombinedSelPermInfo [colsWithColCaseBoolExp]
[scalarCompFieldsWithColCaseBoolExp]
[spiFilter]
spiLimit
spiAllowAgg
(Set.fromList spiRequiredHeaders)
combine :: SelPermInfo b -> CombinedSelPermInfo b -> CombinedSelPermInfo b
combine (modifySingleSelectPerm -> lSelPermInfo) accSelPermInfo =
CombinedSelPermInfo
{ cspiCols = (cspiCols lSelPermInfo) <> (cspiCols accSelPermInfo)
, cspiScalarComputedFields =
(cspiScalarComputedFields lSelPermInfo) <> (cspiScalarComputedFields accSelPermInfo)
, cspiFilter = (cspiFilter lSelPermInfo) <> (cspiFilter accSelPermInfo)
, cspiLimit =
case (cspiLimit lSelPermInfo, cspiLimit accSelPermInfo) of
(Nothing, Nothing) -> Nothing
(Just l, Nothing) -> Just l
(Nothing, Just r) -> Just r
(Just l , Just r) -> Just $ max l r
[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
, cspiAllowAgg = cspiAllowAgg lSelPermInfo || cspiAllowAgg accSelPermInfo
, cspiRequiredHeaders = (cspiRequiredHeaders lSelPermInfo) <> (cspiRequiredHeaders accSelPermInfo)
}
combineCaseBoolExps l r =
case (l, r) of
(Nothing, Nothing) -> Nothing
(Just caseBoolExp, Nothing) -> Just caseBoolExp
(Nothing, Just caseBoolExp) -> Just caseBoolExp
(Just caseBoolExpL, Just caseBoolExpR) -> Just $ BoolOr [caseBoolExpL, caseBoolExpR]
-- | 'orderRoles' is used to order the roles, in such a way that given
-- a role R with n parent roles - PR1, PR2 .. PRn, then the 'orderRoles'
-- function will order the roles in such a way that all the parent roles
-- precede the role R. Note that the order of the parent roles itself doesn't
-- matter as long as they precede the roles on which they are dependent on.
--
-- For example, the orderRoles may return `[PR1, PR3, PR2, ... PRn, R]`
-- or `[PR5, PR3, PR1 ... R]`, both of them are correct because all
-- the parent roles precede the inherited role R, assuming the parent roles
-- themselves don't have any parents for the sake of this example.
orderRoles
:: MonadError QErr m
=> [Role]
-> m [Role]
orderRoles allRoles = do
-- inherited roles can be created from other inherited and non-inherited roles
-- So, roles can be thought of as a graph where non-inherited roles don't have
-- any outgoing edges and inherited roles as nodes with edges to its parent roles
-- However, we can't allow cyclic roles since permissions built by a role is used
-- by the dependent roles to build their permissions and if cyclic roles were to be
-- allowed, the permissions building will be stuck in an infinite loop
let graphNodesList = [(role, _rRoleName role, toList (_unParentRoles . _rParentRoles $ role)) | role <- allRoles]
let orderedGraphNodes = G.stronglyConnComp graphNodesList -- topologically sort the nodes of the graph
cyclicRoles = filter checkCycle orderedGraphNodes
unless (null cyclicRoles) $ do
-- we're appending the first element of the list at the end, so that the error message will
-- contain the complete cycle of the roles
let roleCycles = map (tshow . map (roleNameToTxt . _rRoleName) . appendFirstElementAtEnd . G.flattenSCC) cyclicRoles
throw400 CyclicDependency $ "found cycle(s) in roles: " <> commaSeparated roleCycles
let allOrderedRoles = G.flattenSCCs orderedGraphNodes
pure allOrderedRoles
where
checkCycle = \case
G.AcyclicSCC _ -> False
G.CyclicSCC _ -> True
appendFirstElementAtEnd [] = []
appendFirstElementAtEnd (x:xs) = (x:xs) ++ [x]
buildTablePermissions
:: forall b arr m
. ( ArrowChoice arr, Inc.ArrowDistribute arr, Inc.ArrowCache m arr
, MonadError QErr m, ArrowWriter (Seq CollectedInfo) arr
[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
, HasServerConfigCtx m
, BackendMetadata b
, Inc.Cacheable (Proxy b)
)
=> ( Proxy b
, SourceName
, Inc.Dependency (TableCoreCache b)
, FieldInfoMap (FieldInfo b)
, TablePermissionInputs b
, [Role]
) `arr` (RolePermInfoMap b)
buildTablePermissions = Inc.cache proc (proxy, source, tableCache, tableFields, tablePermissions, orderedRoles) -> do
let alignedPermissions = alignPermissions tablePermissions
table = _tpiTable tablePermissions
[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
experimentalFeatures <- bindA -< _sccExperimentalFeatures <$> askServerConfigCtx
metadataRolePermissions <-
[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
(| Inc.keyed (\_ (insertPermission, selectPermission, updatePermission, deletePermission) -> do
insert <- buildPermission -< (proxy, tableCache, source, table, tableFields, listToMaybe insertPermission)
select <- buildPermission -< (proxy, tableCache, source, table, tableFields, listToMaybe selectPermission)
update <- buildPermission -< (proxy, tableCache, source, table, tableFields, listToMaybe updatePermission)
delete <- buildPermission -< (proxy, tableCache, source, table, tableFields, listToMaybe deletePermission)
returnA -< RolePermInfo insert select update delete)
|) alignedPermissions
[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
-- see [Inherited roles architecture for postgres read queries]
(| foldlA' (\accumulatedRolePermMap (Role roleName (ParentRoles parentRoles)) -> do
let metadataRoleSelectPermission = _permSel =<< M.lookup roleName accumulatedRolePermMap
roleSelectPermission <-
bindA -<
case metadataRoleSelectPermission of
-- There's an permission set for the role in the metadata, so we use that
Just metadataRoleSelectPermission' -> pure $ Just metadataRoleSelectPermission'
-- no permission found in the metadata, so we inherit permissions from the parent
-- roles
Nothing -> do
parentRoleTablePerms <-
for (toList parentRoles) $ \role ->
onNothing (M.lookup role accumulatedRolePermMap) $ throw500 $
-- this error will ideally never be thrown, but if it's thrown then
-- it's possible that the permissions for the role do exist, but it's
-- not yet built due to wrong ordering of the roles, check `orderRoles`
"buildTablePermissions: table role permissions for role: " <> role <<> " not found"
pure $ combineSelectPermInfos <$> NE.nonEmpty (catMaybes $ _permSel <$> parentRoleTablePerms)
let rolePermInfo = RolePermInfo Nothing roleSelectPermission Nothing Nothing
returnA -< M.insert roleName rolePermInfo accumulatedRolePermMap)
|) metadataRolePermissions (bool mempty orderedRoles $ EFInheritedRoles `elem` experimentalFeatures)
-- build permissions for inherited roles only when inherited roles is enabled
where
mkMap :: [PermDef a] -> HashMap RoleName (PermDef a)
mkMap = mapFromL _pdRole
alignPermissions TablePermissionInputs{..} =
let insertsMap = M.map (\a -> ([a], [], [], [])) (mkMap _tpiInsert)
selectsMap = M.map (\a -> ([], [a], [], [])) (mkMap _tpiSelect)
updatesMap = M.map (\a -> ([], [], [a], [])) (mkMap _tpiUpdate)
deletesMap = M.map (\a -> ([], [], [], [a])) (mkMap _tpiDelete)
unionMap = M.unionWith (<>)
in insertsMap `unionMap` selectsMap `unionMap` updatesMap `unionMap` deletesMap
mkPermissionMetadataObject
:: forall b a
. (ToJSON (a b), BackendMetadata b, IsPerm a)
=> SourceName
-> TableName b
-> PermDef (a b)
-> MetadataObject
mkPermissionMetadataObject source table permDef =
let permType = permAccToType (permAccessor :: PermAccessor b (PermInfo a b))
objectId = MOSourceObjId source
$ AB.mkAnyBackend
$ SMOTableObj @b table
$ MTOPerm (_pdRole permDef) permType
definition = toJSON $ WithTable @b source table permDef
in MetadataObject objectId definition
mkRemoteSchemaPermissionMetadataObject
:: AddRemoteSchemaPermissions
-> MetadataObject
mkRemoteSchemaPermissionMetadataObject (AddRemoteSchemaPermissions rsName roleName defn _) =
let objectId = MORemoteSchemaPermissions rsName roleName
in MetadataObject objectId $ toJSON defn
withPermission
:: forall bknd a b c s arr
. (ArrowChoice arr, ArrowWriter (Seq CollectedInfo) arr, BackendMetadata bknd, ToJSON (c bknd), IsPerm c)
=> WriterA (Seq SchemaDependency) (ErrorA QErr arr) (a, s) b
-> ( a
, ((SourceName, TableName bknd, PermDef (c bknd), Proxy bknd), s)
) `arr` (Maybe b)
withPermission f = proc (e, ((source, table, permission, _proxy), s)) -> do
let metadataObject = mkPermissionMetadataObject @bknd source table permission
permType = permAccToType (permAccessor :: PermAccessor bknd (PermInfo c bknd))
roleName = _pdRole permission
schemaObject = SOSourceObj source
$ AB.mkAnyBackend
$ SOITableObj @bknd table
$ TOPerm roleName permType
addPermContext err = "in permission for role " <> roleName <<> ": " <> err
(| withRecordInconsistency (
(| withRecordDependencies (
(| modifyErrA (f -< (e, s))
|) (addTableContext @bknd table . addPermContext))
|) metadataObject schemaObject)
|) metadataObject
buildPermission
:: forall b a arr m
. ( ArrowChoice arr
, ArrowWriter (Seq CollectedInfo) arr
, Inc.ArrowCache m arr
, Inc.Cacheable (a b)
, Inc.Cacheable (Proxy b)
, MonadError QErr m
, BackendMetadata b
, ToJSON (a b)
, IsPerm a
)
=> ( Proxy b
, Inc.Dependency (TableCoreCache b)
, SourceName
, TableName b
, FieldInfoMap (FieldInfo b)
, Maybe (PermDef (a b))
) `arr` Maybe (PermInfo a b)
buildPermission = Inc.cache proc (proxy, tableCache, source, table, tableFields, maybePermission) -> do
(| traverseA ( \permission ->
(| withPermission (do
bindErrorA -< when (_pdRole permission == adminRoleName) $
throw400 ConstraintViolation "cannot define permission for admin role"
(info, dependencies) <- liftEitherA <<< Inc.bindDepend -< runExceptT $
runTableCoreCacheRT (buildPermInfo source table tableFields permission) (source, tableCache)
tellA -< Seq.fromList dependencies
returnA -< info)
|) (source, table, permission, proxy))
|) maybePermission
>-> (\info -> join info >- returnA)