community-rust
community-rust copied to clipboard
C4tWithShell
Cannot resolve path to source files when project is in a subdirectory
Problem
When the Rust project is located in a subdirectory, the sonar-scanner is unable to resolve the path, resulting in the warning:
WARN Failed to resolve 1 file path(s) in Clippy report. No issues imported related to file(s): src/main.rs
Related
The PR 66 introduced the FileAdjustor, which takes the path from the message:spans:file_name property of the generated clippy report. The file name seems to be relative to the project root.
Possible solution
One could consider the target:src_path property from the report as well, to find the source files.
An example report that shows the properties:
{
"reason": "compiler-message",
"package_id": "path+file:///builds/user/ci-test/rust/quality_test#0.1.0",
"manifest_path": "/builds/user/ci-test/rust/quality_test/Cargo.toml",
"target": {
"kind": [
"bin"
],
"crate_types": [
"bin"
],
"name": "quality_test",
"src_path": "/builds/user/ci-test/rust/quality_test/src/main.rs",
"edition": "2021",
"doc": true,
"doctest": false,
"test": true
},
"message": {
"rendered": "warning: unneeded `return` statement\n --> src/main.rs:2:5\n |\n2 | return &val[12..50];\n | ^^^^^^^^^^^^^^^^^^^\n |\n = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_return\n = note: `#[warn(clippy::needless_return)]` on by default\nhelp: remove `return`\n |\n2 - return &val[12..50];\n2 + &val[12..50]\n |\n\n",
"$message_type": "diagnostic",
"children": [
{
"children": [],
"code": null,
"level": "help",
"message": "for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#needless_return",
"rendered": null,
"spans": []
},
{
"children": [],
"code": null,
"level": "note",
"message": "`#[warn(clippy::needless_return)]` on by default",
"rendered": null,
"spans": []
},
{
"children": [],
"code": null,
"level": "help",
"message": "remove `return`",
"rendered": null,
"spans": [
{
"byte_end": 53,
"byte_start": 34,
"column_end": 24,
"column_start": 5,
"expansion": null,
"file_name": "src/main.rs",
"is_primary": true,
"label": null,
"line_end": 2,
"line_start": 2,
"suggested_replacement": "&val[12..50]",
"suggestion_applicability": "MachineApplicable",
"text": [
{
"highlight_end": 24,
"highlight_start": 5,
"text": " return &val[12..50];"
}
]
},
{
"byte_end": 54,
"byte_start": 53,
"column_end": 25,
"column_start": 24,
"expansion": null,
"file_name": "src/main.rs",
"is_primary": true,
"label": null,
"line_end": 2,
"line_start": 2,
"suggested_replacement": "",
"suggestion_applicability": "MachineApplicable",
"text": [
{
"highlight_end": 25,
"highlight_start": 24,
"text": " return &val[12..50];"
}
]
}
]
}
],
"code": {
"code": "clippy::needless_return",
"explanation": null
},
"level": "warning",
"message": "unneeded `return` statement",
"spans": [
{
"byte_end": 53,
"byte_start": 34,
"column_end": 24,
"column_start": 5,
"expansion": null,
"file_name": "src/main.rs",
"is_primary": true,
"label": null,
"line_end": 2,
"line_start": 2,
"suggested_replacement": null,
"suggestion_applicability": null,
"text": [
{
"highlight_end": 24,
"highlight_start": 5,
"text": " return &val[12..50];"
}
]
}
]
}
}
Hi! Right, I think target:src_path will allow to drop the FileAdjustor as it is using the absolute path
I prepared the new version with fix, I am still testing it but you can check if it fixes your issue https://github.com/C4tWithShell/community-rust/releases/tag/v0.2.5
Hi. Thanks for the quick fix!
I would've loved to try it myself. Unfortunately, I wasn't able to start SonarQube:10.7-community with the plugin, since the Sonar Plugin API version requirement seems to be not fulfilled, according to the error message:
2024.10.21 07:23:54 WARN web[][o.s.c.a.AnnotationConfigApplicationContext] Exception encountered during context initialization - cancelling refresh attempt: org.springframework.beans.factory.BeanCreationException: Error creating bean with name 'jdk.internal.loader.ClassLoaders$AppClassLoader@71529963-org.sonar.server.plugins.ServerPluginManager': Initialization of bean failed; nested exception is Plugin Rust language analyzer [communityrust] requires at least Sonar Plugin API version 10.12.0.2522 (current: 10.11.0.2468)
org.springframework.beans.factory.BeanCreationException: Error creating bean with name 'jdk.internal.loader.ClassLoaders$AppClassLoader@71529963-org.sonar.server.plugins.ServerPluginManager': Initialization of bean failed; nested exception is Plugin Rust language analyzer [communityrust] requires at least Sonar Plugin API version 10.12.0.2522 (current: 10.11.0.2468)
Which version of SonarQube are you using and for which version is the 0.2.5 release intended?
Sorry, my bad I uploaded wrong file, 10.12.0.2522 is the newest api version, it is not supported yet I replaced the file with the correct version 10.11.0.2468
I will work on workflow for plugin to avoid this problem in future
Thanks, the version incompatibility seems to be fixed. A new error arose during the start of SonarQube though. I now get the issue that a rule name seems to be too long:
2024.10.22 08:03:37 WARN web[][o.s.c.a.AnnotationConfigApplicationContext] Exception encountered during context initialization - cancelling refresh attempt: org.springframework.beans.factory.BeanCreationException: Error creating bean with name 'jdk.internal.loader.ClassLoaders$AppClassLoader@71529963-org.sonar.server.rule.registration.RulesRegistrant': Initialization of bean failed; nested exception is java.lang.IllegalArgumentException: Rule name is too long: Checks for casts between numeric types that may truncate large values. This is expected behavior, so the cast is `Allow` by default. It suggests user either explicitly ignore the lint, or use `try_from()` and handle the truncation, default, or panic explicitly.
2024.10.22 08:03:37 ERROR web[][o.s.s.p.Platform] Background initialization failed. Stopping SonarQube
org.springframework.beans.factory.BeanCreationException: Error creating bean with name 'jdk.internal.loader.ClassLoaders$AppClassLoader@71529963-org.sonar.server.rule.registration.RulesRegistrant': Initialization of bean failed; nested exception is java.lang.IllegalArgumentException: Rule name is too long: Checks for casts between numeric types that may truncate large values. This is expected behavior, so the cast is `Allow` by default. It suggests user either explicitly ignore the lint, or use `try_from()` and handle the truncation, default, or panic explicitly.
Is there a way to extend this length limitation?
I see, I was able to reproduce on clean deploy. I will make adjustments
I will test it and will reach you after that
UPD: Done, there should not be any errors @schaffenrath, can you check if it solved issue for you?
Just tested it and works great. Clippy report for a Rust project in a sub-folder is now also processed correctly. Thanks!
I believe this now reports on the wrong file in certain cases, causing the import to fail with an invalid location:
java.lang.IllegalArgumentException: 68 is not a valid line offset for pointer. File src/lib.rs has 36 character(s) at line 7
For example, the edited line below fails because it now incorrectly reports on lib.rs whereas the lint comes from ffi.rs.
{
"reason": "compiler-message",
"target": {
"src_path": "/Users/bas/bzhoek/id3rs/src/lib.rs"
},
"message": {
"rendered": "warning: unsafe function's docs are missing a `# Safety` section\n --> src/ffi.rs:42:1\n |\n42 | pub unsafe extern \"C\" fn id3_free(ptr: *mut ID3rs) {\n | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n |\n = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#missing_safety_doc\n\n",
"$message_type": "diagnostic",
"level": "warning",
"message": "unsafe function's docs are missing a `# Safety` section",
"spans": [
{
"byte_end": 1292,
"byte_start": 1242,
"column_end": 51,
"column_start": 1,
"file_name": "src/ffi.rs",
"text": [
{
"highlight_end": 51,
"highlight_start": 1,
"text": "pub unsafe extern \"C\" fn id3_free(ptr: *mut ID3rs) {"
}
]
}
]
}
}
