What causes "table out of scope" in pathwaycom/pathway?

If user code: report upstream with a minimal example; the library should guarantee scope consistency

pathwaycom/pathway · error · OutOfScopeError

table out of scope

Error message

table out of scope

What it means

GraphRunnerScope.get_table raises OutOfScopeError('table out of scope') when asked for the api.Table registered under a given universe.Universe key but no table was ever set for that universe (key not in self.tables). This is internal engine bookkeeping: extraction order or cross-scope table usage is broken.

Source

Thrown at python/pathway/internals/graph_runner/state.py:142

    def get_universe(self, key: universe.Universe):
        if key not in self.universes:
            return self.extract_universe(key)
        return self.universes[key]

    def has_universe(self, key: universe.Universe) -> bool:
        return key in self.universes

    def add_computer_logic(self, computer_callback: Callable) -> int:
        id = len(self.computers)
        self.computers.append(computer_callback)
        return id

    def get_computer_logic(self, id: int) -> Callable:
        return self.computers[id]

    def get_table(self, key: universe.Universe) -> api.Table:
        if key not in self.tables:
            raise OutOfScopeError("table out of scope")
        return self.tables[key]

    def get_tables(self, keys: Iterable[universe.Universe]) -> list[api.Table]:
        return [self.get_table(key) for key in keys]

    def set_table(self, storage: Storage, table: api.Table) -> None:
        self.tables[storage._universe] = table
        self.storages[storage._universe] = storage

    def get_storage(self, key: universe.Universe) -> Storage:
        if key not in self.storages:
            raise OutOfScopeError("path storage out of scope")
        return self.storages[key]

    def get_storages(self, keys: Iterable[universe.Universe]) -> list[Storage]:
        return [self.get_storage(key) for key in keys]

    def set_error_log(self, table: table.Table, error_log: api.ErrorLog) -> None:

View on GitHub (pinned to fa2f74a464)

Solutions

If user code: report upstream with a minimal example; the library should guarantee scope consistency
If custom operator: register the table for the universe via set_table before get_table
Verify all operator dependencies are declared so the graph runner processes producers before consumers

Defensive patterns

Strategy: try-catch

Try / catch

from pathway.internals.graph_runner.state import OutOfScopeError

try:
    ...  # build/run pipeline
except OutOfScopeError as e:
    raise RuntimeError("internal graph-runner state error; report upstream") from e

Prevention

Report occurrences upstream with a minimal reproducer; user code cannot fix scope registration
Avoid mixing Column/Table objects captured across different graph builds in the same process

When it happens

Trigger: An operator handler calls get_table for a universe whose storage was never passed through set_table; reusing universes from a different GraphRunnerScope; incorrect dependency ordering in a custom operator handler.

Common situations: Developing custom operators against pathway.internals; pathway version mismatches between captured graph state and runner state; almost never produced by pure user-level pipeline code.

Related errors

AI-assisted analysis of pathwaycom/pathway@fa2f74a464 (2026-08-15). Data as JSON: /api/errors/6788898e18d0d0f6. Report an issue: GitHub.