Feature: bidirectional AWG/mm2 unit conversion #41
Conversation
Add an inverted dictionary and a lookup function from awg -> mm2. Also do some minor refactoring. Both sides of the conversion table were converted to strings, since '0000' and '2/0' are perfectly valid AWG values.
Show conversions for ex02, and make sure it displays conversions in both directions. Rebuild the example files.
|
This looks great. Can we have AWG be output in upper case, and have it detect upper and lower case in the YAML input? It's the only nit-pick I can come up with ;) |
|
Hmm the converted value is (hardcoded) upper case, the lowercase you see is from the original YAML; that's passed directly into the gauge_unit variable and used as is. To clarify, do you mean that 1) all AWG strings should be converted to upper case for rendering, or that 2), all AWG strings should respect and use the casing from the YAML file? The first i think should be separately refactored, but it would yield more consistent output. The second would probably be a bit of a pain, because the converted values won't have a YAML original AWG string to base casing on, and if we inspect the rest of the YAML input we might very well find multiple different cases. In fact, the input isn't actually sanity checked, so it's quite possible to specify anything other than mm2 as e.g. 'MuricanWireGauge' or 'bananas' in the YAML, currently. |
|
Your edit beat me to the punch. I think AWG should always be rendered as upper case, but we do need to take into account where unknown units ('bananas', square inches, ...) are used, these should simply be passed through, with no equivalent conversion and no case conversion either. |
|
Okay, so for the purposes of this PR:
sound about right? |
The parsing allows arbitrary units to be used for cable dimensions -- this might be valid units, e.g. square inches, or invalid, e.g. bananas. We only allow conversion between mm2 and AWG, so check that the gauge_unit is either of those before conversion. If not, pass through as is.
|
Sounds good. If there's a separate PR coming for the last point, is this one ready for merging? |
|
Actually that was a quick fix, pushing a commit with that as well in a sec seeing as it's related |
Convert e.g. 'awg, 'AwG' to upper case for consistent rendering. Leave any other input gauge units as they were.
src/wireviz/Harness.py
Outdated
| if cable.show_equiv: | ||
| if cable.gauge_unit =='mm\u00B2': | ||
| awg_fmt = f' ({awg_equiv(cable.gauge)} AWG)' | ||
| else: |
There was a problem hiding this comment.
Should this be elif cable.gauge_unit.upper() == 'AWG' or cable.gauge_unit == '' or something similar to avoid converting 'bananas' to AWG by assuming 'bananas' equals mm2?
There was a problem hiding this comment.
Yes I think that is a good call
There was a problem hiding this comment.
The 'AWG' and 'mm' constant strings should also be defined only once in the helper file and then use identifiers in the rest of the code.
There was a problem hiding this comment.
👍 especially the Unicode escaping is a pain. Will you include the fix in this PR? should be easy enough
There was a problem hiding this comment.
yeah, that was fixed in 625b21e. And i agree about the multiple definitions, but that's a separate refactor job IMO. Plus I need to go do some actual paid work for a while, so it won't get done for a couple of hours, and I'd rather get this closed out in the meanwhile.
There was a problem hiding this comment.
Up to you. Let me know when you're ready to merge this one.
There was a problem hiding this comment.
If you're fine with the string casing fix for now i'd say go for merge
There was a problem hiding this comment.
@n42 I'm sorry for pointing out the issue you already had fixed, but I could not see your fix at that time. Github have been slow and somewhat unstable today, so I guess my view was not fully updated.
|
Figured that might as well be done early-on, so put it alongside the mm2 -> mm^2 conversion in the dataclass |
Broke out this feature from PR #9 and resubmitting separately with minor changes after rebase:
Feature: Add automatic AWG->mm2 conversion as well, when using show_equiv. Some minor refactoring there as well, modified ex02 to show it the results. Note that the lookup table has been changed to strings on both sides, this is because e.g. '000', '2/0' and the like are perfectly valid AWG values.
Fixes [Feature] Bidirectional awg<->mm2 unit conversion #40