The first CI analysis of main failed its gate: 16 Rust bugs and new-code coverage 78.3 % against 80 %, over all of 1.4.0 (sending a version did not move the previous-version baseline). - rust:S6677 x3: a failure's message is logged as `error`, not as a second `message` field (CLI, update flow, app startup); golden.rs reads the CLI's log to hold it, and failed first. - rust:S7493 x13: std::fs inside async fns goes through tokio::fs (tools venv sync, deptry's report, export/requirements/report commands). - Tests for 1.4.0 code nothing drove: rollback end to end, forcing a latest this Python cannot run, PEP 691 parsing (L55), user_file_error (U25), a plan verified in another environment, PyPiReleases' cache, and the plan/env stores' success and late-error paths. InstallSet::only had no caller and is removed. - ROADMAP, TESTING and CLAUDE.md record what the analysis found. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
84 lines
3.9 KiB
Properties
84 lines
3.9 KiB
Properties
# SonarQube Cloud - CI-based analysis, run by .github/workflows/sonar.yml.
|
|
#
|
|
# Automatic analysis (the SonarCloud GitHub App scanning on its own) was switched off for this:
|
|
# it cannot run a build, so it never analysed a line of Rust - two thirds of the code - and it
|
|
# could import no coverage at all. It also sent no project version, so the "previous version"
|
|
# new-code period stayed anchored to the first analysis (2026-09-27) and "new code" only ever grew.
|
|
# The workflow passes `sonar.projectVersion` from package.json, which is one of the five places the
|
|
# version lives (CLAUDE.md) and what the previous-version period keys on. The first analysis to
|
|
# send one kept the old baseline; whether a bump moves it is to be read from the gate, not assumed.
|
|
|
|
sonar.organization=poli0981
|
|
sonar.projectKey=poli0981_pipdock
|
|
sonar.projectName=pipdock
|
|
sonar.sourceEncoding=UTF-8
|
|
|
|
# --- Scope ------------------------------------------------------------------------------------
|
|
# Everything tracked, as automatic analysis saw it, with tests split out. Untracked build output
|
|
# (target/, node_modules/, dist/, coverage/) is already out: files .gitignore names are never
|
|
# indexed. Test patterns appear twice, as sources' exclusions and as tests' inclusions, because
|
|
# the two sets must be disjoint or the scanner refuses to index a file twice.
|
|
sonar.sources=.
|
|
sonar.tests=.
|
|
sonar.test.inclusions=\
|
|
ui/src/**/*.test.ts,\
|
|
ui/src/**/*.test.tsx,\
|
|
ui/src/test/**,\
|
|
crates/*/tests/**
|
|
# Images and fonts are tracked too, and the secrets sensor read every one as UTF-8 text and warned
|
|
# about each of them; nothing in them is code.
|
|
sonar.exclusions=\
|
|
ui/src/**/*.test.ts,\
|
|
ui/src/**/*.test.tsx,\
|
|
ui/src/test/**,\
|
|
crates/*/tests/**,\
|
|
**/*.png,\
|
|
**/*.ico,\
|
|
**/*.icns,\
|
|
**/*.woff2
|
|
# Fixtures are data, not code: captured engine output, and sample Python projects that are broken
|
|
# on purpose so that Code Health has something to find.
|
|
sonar.test.exclusions=\
|
|
crates/*/tests/fixtures/**,\
|
|
crates/*/tests/snapshots/**
|
|
|
|
# --- Coverage -----------------------------------------------------------------------------------
|
|
# Rust from cargo-llvm-cov and TypeScript from vitest's v8 provider, both LCOV. Clippy's JSON is
|
|
# imported as external issues; `-D warnings` in ci-rust.yml keeps it empty on any green build, so
|
|
# what it adds is the finding on the line, in the PR, when that build is red.
|
|
sonar.rust.lcov.reportPaths=target/lcov-rust.info
|
|
sonar.rust.clippyReport.reportPaths=target/clippy.json
|
|
sonar.javascript.lcov.reportPaths=coverage/lcov.info
|
|
|
|
# What no test in this repository can execute, so a 0% here would measure nothing:
|
|
# * xtask, scripts/ and the root *.config files are developer tooling, run by hand;
|
|
# * spikes/ are the research scripts the adapters were built from (spikes/README.md);
|
|
# * probe.py runs inside the user's interpreter - the Rust suite parses canned output, and only
|
|
# ci-integration.yml runs it for real;
|
|
# * generated.ts is `cargo run -p xtask -- bindings` output, and the two entry points mount what
|
|
# the tests render directly.
|
|
sonar.coverage.exclusions=\
|
|
xtask/**,\
|
|
scripts/**,\
|
|
spikes/**,\
|
|
*.config.ts,\
|
|
*.config.js,\
|
|
crates/pipdock-core/examples/**,\
|
|
crates/pipdock-core/src/probe.py,\
|
|
src-tauri/build.rs,\
|
|
src-tauri/src/main.rs,\
|
|
ui/src/main.tsx,\
|
|
ui/src/ipc/generated.ts
|
|
|
|
# Regenerating the bindings rewrites dozens of near-identical type declarations at once, which
|
|
# would read as new duplication on the next analysis without anyone having copied anything. Test
|
|
# files never count toward duplication, but the Rust sensor computed theirs anyway and logged a
|
|
# warning per file saying it would be dropped.
|
|
sonar.cpd.exclusions=\
|
|
ui/src/ipc/generated.ts,\
|
|
crates/*/tests/**
|
|
|
|
# --- Languages ----------------------------------------------------------------------------------
|
|
# probe.py's floor and ceiling (ARCHITECTURE section 10: managed Pythons 3.10-3.14).
|
|
sonar.python.version=3.10,3.11,3.12,3.13,3.14
|