Reactive API (ipyflow.api)

These are the user-facing helpers re-exported at the top level of ipyflow (from ipyflow import deps, users, code, ...). Each symbol-introspection helper takes the value of a notebook symbol; a tracer hook resolves it to the corresponding Symbol. See Introspecting the dataflow graph for worked examples.

Symbol introspection

ipyflow.api.lift.lift(sym: Any) → Symbol

Given the programmatic usage of some symbol, look up the corresponding Symbol metadata.

ipyflow.api.lift.code(sym: Any, **kwargs: Any) → HTML | str

Given the programmatic usage of some symbol, look up the corresponding code for that symbol.

ipyflow.api.lift.timestamp(sym: Any) → Timestamp

Given the programmatic usage of some symbol, look up the corresponding timestamp for that symbol.

ipyflow.api.lift.deps(sym: Any) → List[Symbol]

Given the programmatic usage of some symbol, look up the corresponding dependencies for that symbol.

ipyflow.api.lift.users(sym: Any) → List[Symbol]

Given the programmatic usage of some symbol, look up the corresponding users of that symbol.

ipyflow.api.lift.mutate(sym: Any) → None

Force mutation for a particular symbol.

ipyflow.api.lift.rdeps(sym: Any) → List[Symbol]

Given the programmatic usage of some symbol, look up the corresponding recursive dependencies for that symbol.

ipyflow.api.lift.rusers(sym: Any) → List[Symbol]

Given the programmatic usage of some symbol, look up the corresponding users of that symbol.

ipyflow.api.lift.watchpoints(sym: Any) → Watchpoints

Given the programmatic usage of some symbol, look up the corresponding watchpoints for that symbol.

ipyflow.api.lift.set_tag(sym: Any, tag_value: str) → None

Add the tag value to the symbol.

ipyflow.api.lift.unset_tag(sym: Any, tag_value: str) → None

Remove the tag value from the symbol.

ipyflow.api.lift.has_tag(sym: Any, tag_value: str) → bool

Test whether the symbol has the value tag.

Cell output recovery

ipyflow.api.cells.stdout(ts_or_cell_num: int | Timestamp) → str | None

Return the stdout captured during a cell’s execution.

Parameters:

ts_or_cell_num – a cell execution counter, or a Timestamp (whose cell_num is used).

Returns:

the captured stdout as a string, or None if the cell produced no captured output.

Raises:

ValueError – if no cell with that counter has executed yet.

ipyflow.api.cells.stderr(ts_or_cell_num: int | Timestamp) → str | None

Return the stderr captured during a cell’s execution.

Parameters:

ts_or_cell_num – a cell execution counter, or a Timestamp (whose cell_num is used).

Returns:

the captured stderr as a string, or None if the cell produced no captured output.

Raises:

ValueError – if no cell with that counter has executed yet.

ipyflow.api.cells.reproduce_cell(ctr: int, show_input: bool = True, show_output: bool = True, lookback: int = 0)

Re-render the input and/or output of a previous cell execution.

Because ipyflow captures each cell’s output, this can recover results that autosave (for example, a reactive re-run) has since overwritten – within the current kernel session.

Parameters:

• ctr – the execution counter of the cell to reproduce.

• show_input – whether to render the cell’s input source.

• show_output – whether to render the cell’s captured output.

• lookback – how many executions to step back for this cell. 0 (the default) is the latest execution; 1 is the one before it, and so on – useful for recovering a result a later re-execution replaced.