fix(eslint-plugin): handle output() and input() functions in various rules
#2098
Conversation
output() and input() functions in various rulesoutput() and input() functions in various rules
☁️ Nx Cloud ReportCI is running/has finished running commands for commit 25e0fb0. As they complete they will appear below. Click to see the status, the terminal output, and the build insights. 📂 See all runs for this CI Pipeline Execution ✅ Successfully ran 7 targetsSent with 💌 from NxCloud. |
| export const INPUT_ALIAS = `:matches(PropertyDefinition, MethodDefinition[kind='set']) ${INPUT_DECORATOR} ${LITERAL_OR_TEMPLATE_ELEMENT}`; | ||
| export const INPUT_ALIAS = [ | ||
| `:matches(PropertyDefinition, MethodDefinition[kind='set']) ${INPUT_DECORATOR} > CallExpression > Literal`, | ||
| `:matches(PropertyDefinition, MethodDefinition[kind='set']) ${INPUT_DECORATOR} > CallExpression > TemplateLiteral > TemplateElement`, |
There was a problem hiding this comment.
ℹ️ I didn't use LITERAL_OR_TEMPLATE_ELEMENT here. Using that selector means the no-input-rename rule would need to check if the string that matched is the value of an alias property to avoid false positives. By using three different selectors, it guarantees that it matches to either @Output('foo') or @Output(`foo`) or @Output({ alias: 'foo' }), and won't match to something like @Output({ test: 'foo' }).
You'll see in the no-input-rename rule that I've been able to remove the two checks where it was looking for the parent property or object expression to determine if the matched literal was the value of an alias property.
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #2098 +/- ##
==========================================
+ Coverage 90.99% 91.05% +0.06%
==========================================
Files 181 181
Lines 3475 3521 +46
Branches 580 587 +7
==========================================
+ Hits 3162 3206 +44
+ Misses 165 164 -1
- Partials 148 151 +3
Flags with carried forward coverage won't be shown. Click here to find out more.
|
|
Thanks so much! |
|
I know this is old and closed, but can this be updated with an "allowPrefix" option like some other rules have? My current workaround is to just disable this rule. |
Fixes #2004
Rules affected:
no-output-nativeno-output-on-prefixno-output-renameno-input-renameThis change makes
no-output-nativecheck properties that are initialized to the result of theoutput()function. This was achieved by changing the selector to also look for properties that are assigned the result of a call expression for anoutputfunction that contains an object expression with analiasproperty. Theno-output-nativerule itself did not need to change.As mentioned by @5im0n in #2004 (comment), the
no-output-on-prefixrule was also not checking properties that are initialized to the result of theoutput()function. Because it uses the same selector, that rule now also checks those properties.The
no-output-renamerule also wasn't checking foroutput()properties, and due to the changes to the selectors, it now does. The rule itself did have to change to account for the alias coming from a property in an object expression instead of just being the single parameter to the@Output()decorator.The
no-input-renamerule wasn't affected by the changes to the selectors for the rules mentioned above, but I thought it was a good idea to update it at the same time because the changes that needed to be made to the rule were almost identical to the changes made tono-output-rename- it just had to account for thealiasproperty being on an object expression that is the second argument toinput()instead of the first argument like it would be foroutput().