Implement multiple label #2082

rafsun42 · 2024-08-29T19:39:59Z

Add support for multiple labels in CREATE, MERGE and MATCH clause.

Important Notes:

Before merging, please rerun the workflow again. Any tests that were added after this branch was last updated will fail because this PR makes minor changes in how vertex and edge objects are output.

Update:

~~Base of this PR is changed to a new remote branch multiple_label_16, which is based off master.~~

It represents label expression of different type: empty, single or multiple.

Previously, label field was char* type. The change affected the type cypher_node, cypher_relationship and cypher_target_node. As well as, any places where these types are used.

Supports queries like- MATCH (v:A|B|C) RETURN v MATCH ()-[e:A|B|C]->() RETURN v

Some examples of supported multiple label queries: CREATE (:a:b) MERGE (:a:b) MATCH (:a:b) MATCH (:a|b) See regress/sql/multiple_label.sql for more details on what kind of queries are supported. Change summary: --------------- * A new column `allrelations` is added to ag_label catalog * Change in creating AGE relations logic * Change in MATCH's transformation logic (related to building parse namespace item)

The logic for building vertex objects is updated. Agtype vertex objects can be built from either a single label (as a cstring) or multiple labels (as an agtype array). The following functions are updated to reflect this- agtype_typecast_vertex, agtype_in and _agtype_build_vertex. if _agtype_build_vertex is called from SQL, its label argument must be explicitly cast to avoid ambiguity in function overload. The `_label_names` function is added to extract label names from a vertex ID as a list of string. It is used as a helper function to build vertex objects. A new cache called `allrelations` is also added. This is used by _label_names to search for all labels that are related to a given relation. Multiple helper functions are added to extract label infromation from an entity ID. For example, entity's relation ID, relation name, label names. These are used by CREATE, DELETE, MERGE, VLE and SET executors for building a vertex's object or updating its relation. All test files are updated to show the label field as an array in the output. In all test SQLs, _agtype_build_vertex's label argument is explicity cast.

It updates the function filter_vertices_on_label_id(). Additional changes: ------------------- - Add internal function _label_ids

Cache issues fixed: ------------------- - Use of wrong data type for cache entry in label relation cache (pre-existing) - Use of wrong update function for catalog table (related to multiple label) Other changes: -------------- - The function _label_name() is unsupported for vertices

Changes: -------- - Update create_label_expr_relations() to return RangeVar. It removes redundant call to label_expr_relname() in the code that also calls this function. - Use deconstruct_array() to convert ArrayType* to List* - Update test files after rebase

This fixes some compile-time errors that occur if PostgreSQL is configured with the --with-llvm option.

Changes: ------- * Include missing header files * Update newly added tests * Other minor changes

Following PRs are reapplied: 1465, 1509, 1514, and 1518.

* Regression tests * Comments * Minor overlook during rebasing

MuhammadTahaNaveed · 2024-08-30T14:52:34Z

@rafsun42 The second CREATE command has AB and C as labels but the relation for AB and _agr_AB_C was not created.

multi-label-PR-review=# CREATE EXTENSION age;
CREATE EXTENSION
multi-label-PR-review=# SET search_path=ag_catalog;
SET
multi-label-PR-review=# SELECT create_graph('graph');
NOTICE:  graph "graph" has been created
 create_graph 
--------------
 
(1 row)

multi-label-PR-review=# SELECT * FROM cypher('graph', $$CREATE (x:A:B:C)$$) as (a agtype);
 a 
---
(0 rows)

multi-label-PR-review=# \dt+ graph.*
                                            List of relations
 Schema |       Name       | Type  |   Owner    | Persistence | Access method |    Size    | Description 
--------+------------------+-------+------------+-------------+---------------+------------+-------------
 graph  | A                | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | B                | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | C                | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | _ag_label_edge   | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | _ag_label_vertex | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | _agr_ABC         | table | taha-linux | permanent   | heap          | 16 kB      | 
(6 rows)

multi-label-PR-review=# SELECT * FROM cypher('graph', $$CREATE (x:AB:C)$$) as (a agtype);
 a 
---
(0 rows)

multi-label-PR-review=# \dt+ graph.*
                                            List of relations
 Schema |       Name       | Type  |   Owner    | Persistence | Access method |    Size    | Description 
--------+------------------+-------+------------+-------------+---------------+------------+-------------
 graph  | A                | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | B                | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | C                | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | _ag_label_edge   | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | _ag_label_vertex | table | taha-linux | permanent   | heap          | 8192 bytes | 
 graph  | _agr_ABC         | table | taha-linux | permanent   | heap          | 16 kB      | 
(6 rows)

MuhammadTahaNaveed · 2024-08-30T15:01:28Z

@rafsun42 The second CREATE command has AB and C as labels but the relation for AB and _agr_AB_C was not created.

With merge it caused an assertion failure

multi-label-PR-review=# CREATE EXTENSION age;
CREATE EXTENSION
multi-label-PR-review=# SET search_path=ag_catalog;
SET
multi-label-PR-review=# SELECT create_graph('graph');
NOTICE:  graph "graph" has been created
 create_graph 
--------------
 
(1 row)

multi-label-PR-review=# SELECT * FROM cypher('graph', $$CREATE (x:A:B:C)$$) as (a agtype);
 a 
---
(0 rows)

multi-label-PR-review=# SELECT * FROM cypher('graph', $$MERGE (x:AB:C)$$) as (a agtype);
TRAP: failed Assert("IsA(sss, SubqueryScanState)"), File: "src/backend/executor/cypher_merge.c", Line: 312, PID: 42605
postgres: taha-linux multi-label-PR-review [local] SELECT(ExceptionalCondition+0x6e)[0x56187d565e11]
../postgres-16/pgsql/lib/age.so(+0x17d16)[0x7fb668ebdd16]
../postgres-16/pgsql/lib/age.so(+0x17fc7)[0x7fb668ebdfc7]

rafsun42 · 2024-09-11T02:22:29Z

@MuhammadTahaNaveed Thanks for reporting this. The logic for generating intersection relation was shortsighted. It appends the individual label names together without any separator. So, both CREATE (:A:B:C) and CREATE (:AB:C) generates the same relation name _agr_ABC. The expectation is they should generate different relation name.

One solution can be using a special separator character and disallowing it in label names. For example, the % character (it is already disallowed I believe). So, the intersection relations will be _agr_A%B%C and _agr_AB%C respectively.

What's your thoughts?

MuhammadTahaNaveed · 2024-09-11T11:53:33Z

@MuhammadTahaNaveed Thanks for reporting this. The logic for generating intersection relation was shortsighted. It appends the individual label names together without any separator. So, both CREATE (:A:B:C) and CREATE (:AB:C) generates the same relation name _agr_ABC. The expectation is they should generate different relation name.

One solution can be using a special separator character and disallowing it in label names. For example, the % character (it is already disallowed I believe). So, the intersection relations will be _agr_A%B%C and _agr_AB%C respectively.

What's your thoughts?

@rafsun42 Yeah, I agree that there should be a seperator. Maybe hyphen(-) will be more appropriate than (%)? Hyphen is also not allowed in label names afaik.

This will create a distinction in intersection relation names between the following combinations: (MN:P) and (M:N:P). The credit for finding out this edge case goes to Muhammad Taha Naveed. Additionally, a parameter is added to the create_label function to skip checking valid label names. This is useful when creating intersection relations (the name can be invalid due to containing the separator symbol). Co-authored-by: Muhammad Taha Naveed <[email protected]>

rafsun42 · 2024-09-18T06:57:48Z

@MuhammadTahaNaveed Updated the PR to fix the issues.

jrgemignani · 2024-09-18T17:54:02Z

@rafsun42 The PR has conflicts as well as not running the 7 checks for some reason.

MuhammadTahaNaveed

Left a single comment. Looks good to me otherwise.

MuhammadTahaNaveed · 2024-10-08T07:45:13Z

src/backend/utils/cache/ag_cache.c

+    ag_label = table_open(ag_label_relation_id(), AccessShareLock);
+
+    /* TODO: for now, do seq scan */
+    scan_desc = systable_beginscan(ag_label, InvalidOid, false, NULL, 0, NULL);


There is a small todo here. Scan key is not setup.

rafsun42 added 11 commits August 28, 2024 10:53

Add cypher_label_expr data type

4c8f639

It represents label expression of different type: empty, single or multiple.

Change label field's type to cypher_label_expr

c58521d

Previously, label field was char* type. The change affected the type cypher_node, cypher_relationship and cypher_target_node. As well as, any places where these types are used.

Implement label expression type OR

fdeadbc

Supports queries like- MATCH (v:A|B|C) RETURN v MATCH ()-[e:A|B|C]->() RETURN v

Add multiple label support to MATCH queries without variables

f954921

It updates the function filter_vertices_on_label_id(). Additional changes: ------------------- - Add internal function _label_ids

Add brackets to scope some case statements

f3495c3

This fixes some compile-time errors that occur if PostgreSQL is configured with the --with-llvm option.

Fix issues after rebasing

101eb94

Changes: ------- * Include missing header files * Update newly added tests * Other minor changes

Reapply clean up #include files PRs

c025fdf

Following PRs are reapplied: 1465, 1509, 1514, and 1518.

github-actions bot added the override-stale To keep issues/PRs untouched from stale action label Aug 29, 2024

rafsun42 changed the base branch from multiple_label_16 to master August 29, 2024 19:52

rafsun42 changed the base branch from master to multiple_label_16 August 29, 2024 19:53

rafsun42 changed the base branch from multiple_label_16 to master August 29, 2024 19:56

Update after rebase

23ce511

* Regression tests * Comments * Minor overlook during rebasing

rafsun42 force-pushed the multiple-label-updated branch from 4494c6e to 23ce511 Compare August 29, 2024 19:56

github-actions bot added the master label Aug 29, 2024

rafsun42 mentioned this pull request Aug 29, 2024

Implement multiple label #1785

Closed

rafsun42 requested review from jrgemignani and MuhammadTahaNaveed August 29, 2024 20:01

MuhammadTahaNaveed reviewed Oct 8, 2024

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implement multiple label #2082

Implement multiple label #2082

rafsun42 commented Aug 29, 2024 •

edited

Loading

MuhammadTahaNaveed commented Aug 30, 2024

MuhammadTahaNaveed commented Aug 30, 2024

rafsun42 commented Sep 11, 2024

MuhammadTahaNaveed commented Sep 11, 2024 •

edited

Loading

rafsun42 commented Sep 18, 2024

jrgemignani commented Sep 18, 2024

MuhammadTahaNaveed left a comment

MuhammadTahaNaveed Oct 8, 2024

Implement multiple label #2082

Are you sure you want to change the base?

Implement multiple label #2082

Conversation

rafsun42 commented Aug 29, 2024 • edited Loading

MuhammadTahaNaveed commented Aug 30, 2024

MuhammadTahaNaveed commented Aug 30, 2024

rafsun42 commented Sep 11, 2024

MuhammadTahaNaveed commented Sep 11, 2024 • edited Loading

rafsun42 commented Sep 18, 2024

jrgemignani commented Sep 18, 2024

MuhammadTahaNaveed left a comment

Choose a reason for hiding this comment

MuhammadTahaNaveed Oct 8, 2024

Choose a reason for hiding this comment

rafsun42 commented Aug 29, 2024 •

edited

Loading

MuhammadTahaNaveed commented Sep 11, 2024 •

edited

Loading