Fix lsp_servers[].file_patterns by wtfbbqhax · Pull Request #72 · bug-ops/mcpls

wtfbbqhax · 2026-03-19T21:34:24Z

Description

This change fixes language detection when files are matched via lsp_servers[].file_patterns (for example **/.h, **/.c). serve() now uses an effective extension map that merges workspace extension mappings with extensions inferred from each LSP server’s file_patterns, so C/C++ headers and sources can resolve to the configured server language instead of falling back to plaintext (or wrong defaults).

Type of Change

Bug fix (non-breaking change that fixes an issue)
New feature (non-breaking change that adds functionality)
Breaking change (fix or feature that would cause existing functionality to change)
Documentation update
Refactoring (no functional changes)

Related Issues

N/a

Checklist

I have read the CONTRIBUTING guide
My code follows the project's coding style
I have added tests that prove my fix/feature works
All new and existing tests pass
I have updated the documentation accordingly
I have updated the CHANGELOG.md (for user-facing changes)

Additional Notes

Added ServerConfig::build_effective_extension_map() and wired it into serve().
Added tests for:
- file-pattern-derived extension overrides (.c/.h -> cpp)
- ignoring complex patterns that do not provide a simple extension
Verified locally with: cargo test -p mcpls-core --lib (310 passed, 0 failed).

Copilot

Pull request overview

Fixes language detection when lsp_servers[].file_patterns are used by deriving/overlaying extension→language mappings from those patterns, ensuring files like **/*.c / **/*.h resolve to the configured server language (e.g., cpp) instead of falling back to defaults like c/plaintext.

Changes:

serve() now initializes the Translator with an effective extension map built from both workspace mappings and LSP server file_patterns.
Added conservative glob-pattern extension extraction and ServerConfig::build_effective_extension_map().
Added unit tests covering file-pattern-derived overrides and ignoring complex patterns that don’t yield a simple extension.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
crates/mcpls-core/src/lib.rs	Switches `serve()` to use the new effective extension map for translator initialization.
crates/mcpls-core/src/config/mod.rs	Adds pattern→extension inference, builds merged extension map on `ServerConfig`, and includes tests for the new behavior.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

GitHub's default CodeQL setup only runs on PRs from the same repository. Fork PRs (like #72) are skipped because pull_request events from forks don't have write access to the base repo's security events. Add an explicit codeql.yml workflow using pull_request_target so CodeQL analysis runs for both fork and non-fork PRs. The workflow checks out the PR head SHA explicitly while keeping permissions minimal (security-events: write, contents: read only).

* fix(ci): add CodeQL workflow for fork pull requests GitHub's default CodeQL setup only runs on PRs from the same repository. Fork PRs (like #72) are skipped because pull_request events from forks don't have write access to the base repo's security events. Add an explicit codeql.yml workflow using pull_request_target so CodeQL analysis runs for both fork and non-fork PRs. The workflow checks out the PR head SHA explicitly while keeping permissions minimal (security-events: write, contents: read only). * fix(ci): remove unsafe checkout ref in pull_request_target workflow Checking out fork head SHA in a pull_request_target context grants untrusted code access to the privileged workflow environment. Remove the explicit ref override so checkout defaults to the base branch, which is the correct and secure behavior for CodeQL analysis. Fixes CodeQL alert: Checkout of untrusted code in trusted context * fix(ci): restore fork head checkout with security justification Restore ref: pull_request.head.sha for pull_request_target so CodeQL scans the actual fork PR changes instead of the base branch. This pattern is safe because the job has no access to secrets and only holds security-events: write + contents: read permissions. Added inline comment to suppress the CodeQL false-positive alert and document the security reasoning.

Problem While using `mcpls` in a C++ project I noticed that it always return the following error when calling get references from a '.h' header file: Error: tool call error: tool call failed for `mcpls/get_references` Caused by: tools/call failed: Mcp error: -32603: no LSP server configured for language: c After reviewing the documentation and creating an mcpls.toml with the following, it continued to not work [[lsp_servers]] language_id = "cpp" command = "clangd" args = ["--background-index", "--clang-tidy"] file_patterns = ["**/*.cpp", "**/*.cc", "**/*.cxx", "**/*.hpp", "**/*.c", "**/*.h"] # and/or with [[lsp_servers]] language_id = "c" command = "clangd" args = ["--background-index", "--clang-tidy"] file_patterns = ["**/*.c", "**/*.h"] With the above changing resulting in the language detection for matching file patterns to "plaintext" Error: tool call error: tool call failed for `mcpls/get_references` Caused by: tools/call failed: Mcp error: -32603: no LSP server configured for language: plaintext What changed - serve() now initializes the translator with an effective extension map built from both workspace mappings and LSP server file patterns: - crates/mcpls-core/src/lib.rs:114:114 - Added ServerConfig::build_effective_extension_map() to overlay extensions inferred from file_patterns: - crates/mcpls-core/src/config/mod.rs:287:287 - Added a small parser for simple glob extensions (e.g. **/*.h, *.c): - crates/mcpls-core/src/config/mod.rs:123:123 Tests added - Pattern-derived mapping overrides default extension mapping (.c/.h -> cpp): - crates/mcpls-core/src/config/mod.rs:700:700 - Complex non-simple patterns are ignored safely: - crates/mcpls-core/src/config/mod.rs:721:721 Verification - New tests passed. - Full mcpls-core unit/integration tests passed. - Existing unrelated doctest failure remains in lsp/types.rs (pre-existing visibility issue). Fix was implemented by Codex

bug-ops · 2026-03-19T23:43:56Z

crates/mcpls-core/src/config/mod.rs

+/// Extract a file extension from a glob-like file pattern.
+///
+/// Supports common patterns such as `**/*.rs` and `*.h`.
+/// Returns `None` for patterns without a simple trailing extension.
+fn extract_extension_from_pattern(pattern: &str) -> Option<String> {
+    let (_, ext) = pattern.rsplit_once('.')?;
+    if ext.is_empty() {
+        return None;
+    }
+
+    // Keep this conservative: only accept plain extension-like tokens.
+    if ext
+        .chars()
+        .all(|c| c.is_ascii_alphanumeric() || c == '_' || c == '-')
+    {
+        Some(ext.to_string())
+    } else {
+        None
+    }


Add direct unit tests for extract_extension_from_pattern

Currently this function is only tested indirectly through test_build_effective_extension_map_overrides_with_file_patterns and test_build_effective_extension_map_ignores_complex_patterns_without_extension.

Please add dedicated unit tests covering edge cases:

Empty string pattern: "" → None

Pattern without dot: "**/*" → None

Dotfile pattern: ".gitignore" → None

Multi-dot extensions: "foo.tar.gz" → Some("gz")

This ensures the conservative parsing logic is well-validated.

Optional: Consider returning &str instead of String since rsplit_once already returns slices of the input.

bug-ops

Thanks for tackling this! A few items to polish before merge:

test_serve_initializes_translator_with_extensions(): This test calls the old API (workspace.build_extension_map()). Update it to test the new API (config.build_effective_extension_map()) or clarify if it's testing workspace-level mappings in isolation.
CHANGELOG.md: Document the default behavior change for C/C++ files.

bug-ops requested review from Copilot March 19, 2026 22:52

Copilot started reviewing on behalf of bug-ops March 19, 2026 22:53 View session

Copilot AI reviewed Mar 19, 2026

View reviewed changes

bug-ops mentioned this pull request Mar 19, 2026

fix(labeler): use pull_request_target event for fork compatibility #73

Merged

bug-ops force-pushed the fix_file_patterns branch from 2b424f9 to 155e79c Compare March 19, 2026 23:08

github-actions bot added rust Rust code changes mcpls-core mcpls-core crate changes labels Mar 19, 2026

Copilot AI reviewed Mar 19, 2026

View reviewed changes

bug-ops mentioned this pull request Mar 19, 2026

fix(ci): add CodeQL workflow for fork pull requests #74

Merged

bug-ops force-pushed the fix_file_patterns branch from 155e79c to 1353ef9 Compare March 19, 2026 23:35

bug-ops reviewed Mar 19, 2026

View reviewed changes

bug-ops reviewed Mar 20, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix lsp_servers[].file_patterns#72

Fix lsp_servers[].file_patterns#72
wtfbbqhax wants to merge 1 commit intobug-ops:mainfrom
wtfbbqhax:fix_file_patterns

wtfbbqhax commented Mar 19, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

bug-ops Mar 19, 2026

Uh oh!

bug-ops left a comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

wtfbbqhax commented Mar 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Type of Change

Related Issues

Checklist

Additional Notes

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

bug-ops Mar 19, 2026

Choose a reason for hiding this comment

Uh oh!

bug-ops left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

wtfbbqhax commented Mar 19, 2026 •

edited

Loading