Output raw call stack data

[nziebart]

It would be useful (and I suspect not too difficult) to be able to retrieve the raw call stack data for an invocation of py-spy. This would allow aggregating data from many machines, and performing ad-hoc aggregations over different time ranges.

[benfred]

This shouldn't be too hard to do.

I'm also thinking about adding a python api (using approach I wrote about here: http://benfrederickson.com/writing-python-extensions-in-rust-using-pyo3/) which would let you script operations like this.

[nziebart]

A python api would be awesome!

[nziebart]

It would also be nice to be able to get "idle" stack traces (i.e., not omit them). Sometimes it's useful to know where idle time is being spent (e.g., I might want to know what % of time I spend waiting on a particular db query)

[Norbo11]

Any progress on this?

I have a totally different use case than simply generating flame graphs or viewing the profiling information in real-time. I would love to be able to get the raw data used to generate the flame graph, for example.

[benfred]

I've added this in v0.2.0

@nziebart there is also now a ``--idle``` flag to show idle frames