Frame out the env spec cep #50
Conversation
| dependencies: | ||
| - python=3.7 | ||
| - codecov | ||
| prefix: /Users/username/anaconda3/envs/env-name |
There was a problem hiding this comment.
@jezdez I've been poking at this name vs prefix locally with conda env create and mamba env create. Neither seem to do anything with this prefix key, even though it shows up in conda env export. It seems as though this is a valid key for the spec that conda env will accept, but conda env doesn't actually do anything with this key. can you confirm that I'm understanding this behavior correctly?
$ conda env export
name: base
channels:
- conda-forge
dependencies:
- _libgcc_mutex=0.1=conda_forge
- _openmp_mutex=4.5=2_gnu
- brotlipy=0.7.0=py310h5764c6d_1005
- bzip2=1.0.8=h7f98852_4
- c-ares=1.18.1=h7f98852_0
- ca-certificates=2022.12.7=ha878542_0
- certifi=2022.12.7=pyhd8ed1ab_0
- cffi=1.15.1=py310h255011f_3
- charset-normalizer=3.1.0=pyhd8ed1ab_0
- colorama=0.4.6=pyhd8ed1ab_0
- conda=23.1.0=py310hff52083_0
- conda-package-handling=2.0.2=pyh38be061_0
- conda-package-streaming=0.7.0=pyhd8ed1ab_1
- cryptography=40.0.1=py310h34c0648_0
- fmt=9.1.0=h924138e_0
- icu=72.1=hcb278e6_0
- idna=3.4=pyhd8ed1ab_0
- keyutils=1.6.1=h166bdaf_0
- krb5=1.20.1=h81ceb04_0
- ld_impl_linux-64=2.40=h41732ed_0
- libarchive=3.6.2=h3d51595_0
- libcurl=7.88.1=hdc1c0ab_1
- libedit=3.1.20191231=he28a2e2_2
- libev=4.33=h516909a_1
- libffi=3.4.2=h7f98852_5
- libgcc-ng=12.2.0=h65d4601_19
- libgomp=12.2.0=h65d4601_19
- libiconv=1.17=h166bdaf_0
- libmamba=1.4.1=hcea66bb_0
- libmambapy=1.4.1=py310h1428755_0
- libnghttp2=1.52.0=h61bc06f_0
- libnsl=2.0.0=h7f98852_0
- libsolv=0.7.23=h3eb15da_0
- libsqlite=3.40.0=h753d276_0
- libssh2=1.10.0=hf14f497_3
- libstdcxx-ng=12.2.0=h46fd767_19
- libuuid=2.38.1=h0b41bf4_0
- libxml2=2.10.3=hfdac1af_6
- libzlib=1.2.13=h166bdaf_4
- lz4-c=1.9.4=hcb278e6_0
- lzo=2.10=h516909a_1000
- mamba=1.4.1=py310h51d5547_0
- ncurses=6.3=h27087fc_1
- openssl=3.1.0=h0b41bf4_0
- pip=23.0.1=pyhd8ed1ab_0
- pluggy=1.0.0=pyhd8ed1ab_5
- pybind11-abi=4=hd8ed1ab_3
- pycosat=0.6.4=py310h5764c6d_1
- pycparser=2.21=pyhd8ed1ab_0
- pyopenssl=23.1.1=pyhd8ed1ab_0
- pysocks=1.7.1=pyha2e5f31_6
- python=3.10.10=he550d4f_0_cpython
- python_abi=3.10=3_cp310
- readline=8.2=h8228510_1
- reproc=14.2.4=h0b41bf4_0
- reproc-cpp=14.2.4=hcb278e6_0
- requests=2.28.2=pyhd8ed1ab_1
- ruamel.yaml=0.17.21=py310h1fa729e_3
- ruamel.yaml.clib=0.2.7=py310h1fa729e_1
- setuptools=65.6.3=pyhd8ed1ab_0
- tk=8.6.12=h27826a3_0
- toolz=0.12.0=pyhd8ed1ab_0
- tqdm=4.65.0=pyhd8ed1ab_1
- tzdata=2023c=h71feb2d_0
- urllib3=1.26.15=pyhd8ed1ab_0
- wheel=0.40.0=pyhd8ed1ab_0
- xz=5.2.6=h166bdaf_0
- yaml-cpp=0.7.0=h27087fc_2
- zstandard=0.19.0=py310hdeb6495_1
- zstd=1.5.2=h3eb15da_6
prefix: /home/ericdill/mambaforge
There was a problem hiding this comment.
Interesting. I've been confused by it since it started appearing.
| "type": "array", | ||
| "items": { | ||
| "anyOf": [ | ||
| {"type": "string"}, |
There was a problem hiding this comment.
How far down the MatchSpec rabbit hole do we want to go with this jsonschema? jsonschema does support various formatting validators like regex, but I'm a little hesitant to try and capture MatchSpec in a regex... Thoughts?
There was a problem hiding this comment.
It feels we need a JSONSchema for the MatchSpec first, and then we can refer to it if anything. Can we leave a "type hint" for now? e.g. MatchSpecStr?
| <tr><td> Author(s) </td> | ||
| <td> | ||
| Eric Dill | ||
| Marius van Niekerk |
There was a problem hiding this comment.
@mariusvniekerk you interested in being a co-author of this?
|
Should this CEP also mention which of the keys are strictly required and which are optional? I've found that only |
|
For reference(s), this epic lists a few issues where similar conversations happened: conda/conda#11341 |
|
Btw, thanks to @AlbertDeFusco I added a https://github.com/jaimergp/schemas/blob/revamp/conda_models/environment_file.py |
discussed with @mariusvniekerk, @AlbertDeFusco , @jezdez and @jaimergp.
opening this PR as a draft PR for further conversation. it's not really ready for full review yet.
there's a notebook that validates the env yamls that are part of this PR. those env yamls (and the validation code) should probably move over to the conda specs repo once this is finished and accepted.