Closed miracle2k closed 6 years ago
Thanks for raising the issue!
I thought quite a lot about this and decided to totally drop built-in parsing for chunked responses in v0.2.0, so this issue doesn't apply anymore.
With the new iteration mechanism (iter_resp
), users can customize point parsing as they prefer (for both chunked and non-chunked responses). It should allow for querying with multiple statements/series/measurements as long as you write an appropriate parser.
Things should now be more consistent (across chunked/non-chunked responses), more flexible (with the use of custom parsers) and easier to use (simple iteration via iter_resp
).
Currently, when querying multiple series with
chunked=True
, there is no way to know which series a data point belongs to; the interface only provides the point and does not pass along the name of the measurement.I fixed this for myself here: https://github.com/plugaai/aioinflux/compare/master...miracle2k:multiseries?expand=1 but it's not a clean pull request because it depends on my changes to make the library run on Python 3.5 (for PyPy).