version and cache spec

[bendichter]

addresses #1330

Adds required version to namespace and caches spec in file. This should allow pynwb to read the file without needing to import from allensdk.

[claassistantio]

All committers have signed the CLA.

[codecov-io]

Codecov Report

Merging #1358 into rc/1.6.0 will decrease coverage by 0.72%.
The diff coverage is 0%.

@@             Coverage Diff              @@
##           rc/1.6.0    #1358      +/-   ##
============================================
- Coverage     34.35%   33.63%   -0.73%     
============================================
  Files           340      340              
  Lines         33004    33003       -1     
============================================
- Hits          11340    11099     -241     
- Misses        21664    21904     +240

Impacted Files	Coverage Δ
...dk/brain_observatory/ecephys/write_nwb/__main__.py	66.17% <0%> (+0.24%)	⬆️
allensdk/brain_observatory/nwb/__init__.py	57% <0%> (-17.59%)	⬇️
...avior/behavior_ophys_api/behavior_ophys_nwb_api.py	42.93% <0%> (-15.22%)	⬇️
...ory/ecephys/stimulus_analysis/drifting_gratings.py	63.84% <0%> (-13.47%)	⬇️
allensdk/ephys/ephys_extractor.py	34.65% <0%> (-12.82%)	⬇️
...n_observatory/ecephys/stimulus_analysis/flashes.py	68.08% <0%> (-9.58%)	⬇️
...bservatory/ecephys/stimulus_analysis/dot_motion.py	85.89% <0%> (-6.42%)	⬇️
allensdk/ephys/ephys_features.py	43.85% <0%> (-4.73%)	⬇️
...atory/ecephys/stimulus_analysis/static_gratings.py	66.66% <0%> (-2.95%)	⬇️
...ephys/stimulus_analysis/receptive_field_mapping.py	86.66% <0%> (-2.78%)	⬇️
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7611d50...391459e. Read the comment docs.

[kschelonka]

Does the update to cache_spec when writing the files require any change on our methods to read them back in?

[bendichter]

@kschelonka, No the update to cache_spec when writing the files should not require any changes on your methods to read them back in.

Right now, reading these files requires an import of the extension because the extension is not cached. With the extension cached, you will be able to read the file with

from pynwb import NWBHDF5IO
nwb = NWBHDF5IO(filepath, 'r', load_namespaces=True).read()

and it should not require any other imports. You will retain the ability to read them the old way, importing the extension and using that extension to read the file without loading the namespace.

[kschelonka]

I marked this as no longer a draft. Is that correct?

[bendichter]

Yes that's correct. It was in draft mode while I sorted out CI issues and it seems like you are on that.

…

On Fri, Feb 28, 2020, 4:08 PM Kat Schelonka ***@***.***> wrote: I marked this as no longer a draft. Is that correct? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1358?email_source=notifications&email_token=AAGOEEXSYYXHD5QL6Y3BKPDRFGKQVA5CNFSM4KSG6O32YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOENKPM7Q#issuecomment-592770686>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAGOEEXZWVKWRQADNZBVUFLRFGKQVANCNFSM4KSG6O3Q> .

[kschelonka]

Looks good to me, thanks for your contribution.