sigurdm
commented
1 year ago The documentation logs can be seen here: https://pub.dev/documentation/tizen_interop/latest/log.txt
Closed swift-kim closed 1 year ago
sigurdm
commented
1 year ago The documentation logs can be seen here: https://pub.dev/documentation/tizen_interop/latest/log.txt
sigurdm
commented
1 year ago The documentation logs can be seen here: https://pub.dev/documentation/tizen_interop/latest/log.txt
Sorry, that was a red herring.
I ran a manual pana run, without the limit on output size, and here is the relevant section:
## ✗ Pass static analysis (20 / 30)
### [~] 20/30 points: code has no errors, warnings, lints, or formatting issues
Found 88745 issues. Showing the first 2:
<details>
<summary>
INFO: The variable name 'get_last_result' isn't a lowerCamelCase identifier.
</summary>
`lib/src/bindings/4.0/generated_bindings.dart:32:7`
╷ 32 │ int get_last_result() { │ ^^^^^^^^^^^^^^^ ╵
To reproduce make sure you are using the [lints_core](https://pub.dev/packages/lints) and run `dart analyze lib/src/bindings/4.0/generated_bindings.dart`
</details>
<details>
<summary>
INFO: The variable name '_get_last_resultPtr' isn't a lowerCamelCase identifier.
</summary>pub.dev analyzes the package with the core lints.
If you are certain that you don't want to follow the dart naming convention in the generated code, you can put: // ignore_for_file: type=lint at the top. See: https://dart.dev/guides/language/analysis-options#suppressing-rules-for-a-file
swift-kim
commented
1 year ago @sigurdm Thank you so much for the really quick response!
pub.dev analyzes the package with the core lints.
Does pana (pub.dev) always use the core lints even if there's a custom analysis_options.yaml in the package? I tried adding the following lines to my analysis_options.yaml since I wanted to apply the rules globally to all files, but it still didn't work.
include: package:lints/core.yaml
linter:
rules:
camel_case_types: false
non_constant_identifier_names: falseIf you are certain that you don't want to follow the dart naming convention in the generated code, you can put:
// ignore_for_file: type=lintat the top.
Maybe this is the only option I can use if pana works in that way.
sigurdm
commented
1 year ago Yeah - it always uses those and ignores the custom analysis_options.yaml.
swift-kim
commented
1 year ago Okay. I read the ffigen documentation again and noticed that adding the following preamble is the recommended way to suppress the warnings.
preamble: |
// ignore_for_file: camel_case_types, non_constant_identifier_namesThank you again for the help!
https://pub.dev/packages/tizen_interop/score
I'm a maintainer of the above package but I can't understand why the static analysis always fails with "Running
dart analyzeproduced too large output".The issue can be reproduced locally with pana.
But the error is not reproduced when I run
dart analyzedirectly with the exact same command used by pana.My Dart SDK version:
The package (tizen_interop) is obviously large in code size (about 1M lines) but this fact doesn't explain why pana fails. What would be the real cause of the issue and how should I deal with it?