evanh/feat/use composite queries metrics formulas #5717
Conversation
| @@ -0,0 +1,71 @@ | |||
| pytest tests/query/parser/test_formula_mql_query.py -k "test_simple_formula" | |||
There was a problem hiding this comment.
This file was just to help me track my progress, so I didn't have to context switch as much. Please ignore.
| @@ -160,15 +160,6 @@ def get_alias_node_map(self) -> Mapping[str, IndividualNode[TSimpleDataSource]]: | |||
| **self.right_node.get_alias_node_map(), | |||
| } | |||
|
|
|||
| def __post_init__(self) -> None: | |||
There was a problem hiding this comment.
This was removed because of resolution: this check happens before resolving happens, so there are bunch of columns in the query that "don't exist".
| @@ -0,0 +1,185 @@ | |||
| from __future__ import annotations | |||
There was a problem hiding this comment.
Moved all of this logic to this file to simplify the parser code a bit. Also made it able to work for simple and formula queries.
| @@ -0,0 +1,260 @@ | |||
| from __future__ import annotations | |||
| @@ -1,3 +1,5 @@ | |||
| from __future__ import annotations | |||
There was a problem hiding this comment.
I know the pipeline isn't suppose to be functional yet, but something I came across when re-organizing the query pipeline is adding the storages set keys to JOINABLE_STORAGE_SETS in order for composite query pipeline to work properly:
snuba/snuba/clusters/storage_sets.py
Line 86 in 8261937
Initial draft of changes to support multi-type formulas.
This PR only has changes to the MQL parser, the rest of the pipeline doesn't work yet.
To see an example of a finished query, look at
tests/query/parser/test_formula_mql_query.py.