t.rast.extract: Handle fully qualified map names and semantic labels #6300
Conversation
| for k, v in expected_info.items(): | ||
| assert ( | ||
| strds_info[k] == v | ||
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" |
There was a problem hiding this comment.
Was the default behavior of pytest not enough for dictionaries when developing your tests?
(Like for a bad value, missing, and extra keys)
There was a problem hiding this comment.
I wanted to check only a subset of the metadata keys. And unfortunately, it did not print out the key when I compared values are as expected. Are you aware of any build-in methods in pytest that would allow to test if a dictionary is covered by another (regardless of the order of keys and the like...)? I would be happy to adjust and drop the loop...
python/grass/temporal/extract.py
Outdated
| # Make sure STRDS is in the expression referenced with fully qualified name | ||
| if sp.base.get_map_id() not in expression: | ||
| expression = expression.replace(sp.base.get_name(), sp.base.get_map_id()) |
There was a problem hiding this comment.
You could confuse this by including one full name (id) and one name only like "abc@test1 + abc". I'm not sure how much it makes sense with how the inputs are handled. Based on the tests, it seems it does.
Related to that, you can have two "abc" in different mapsets. I don't know how that would be handled here, but maybe that's a separate issue from the one at hand.
There was a problem hiding this comment.
Right. Good catch. In 193207a I tried a different method using re.
| verbose=True, | ||
| overwrite=True, | ||
| ) | ||
| assert gisenv["MAPSET"] == "perc" |
There was a problem hiding this comment.
I lack explanation why this is tested. It is okay to test pre- and post-conditions. In this case, it looks wrong that this is tested at the end, while the values are acquired at the beginning (and not modified by anything).
| verbose=True, | ||
| overwrite=True, | ||
| ) | ||
| strds_info = gs.parse_key_val(tools.t_info(input=result, flags="g").text) |
There was a problem hiding this comment.
Does the following work for you or not?
| strds_info = gs.parse_key_val(tools.t_info(input=result, flags="g").text) | |
| strds_info = tools.t_info(input=result, flags="g").keyval |
It will make a mix of strings and numbers instead of strings only.
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
| assert ( | ||
| strds_info[k] == v | ||
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" |
There was a problem hiding this comment.
[ruff] reported by reviewdog 🐶
| assert ( | |
| strds_info[k] == v | |
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" | |
| assert strds_info[k] == v, ( | |
| f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" | |
| ) |
| assert ( | ||
| strds_info[k] == v | ||
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" |
There was a problem hiding this comment.
[ruff] reported by reviewdog 🐶
| assert ( | |
| strds_info[k] == v | |
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" | |
| assert strds_info[k] == v, ( | |
| f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" | |
| ) |
| assert ( | ||
| strds_info[k] == v | ||
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" |
There was a problem hiding this comment.
[ruff] reported by reviewdog 🐶
| assert ( | |
| strds_info[k] == v | |
| ), f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" | |
| assert strds_info[k] == v, ( | |
| f"Expected value for key '{k}' is {v}. Got: {strds_info[k]}" | |
| ) |
3 participants
This PR appends semantic labels to the basename when expression is given for extraction of raster maps and handles fully qualified STRDS names more reliably. It also adds a pytest based testcase.
Fixes #6254