Files
kite 533b526b4c chore: add SPDX license headers and automated verification (#740)
* chore: add SPDX license headers to all source files

Add Apache-2.0 SPDX license identifiers and copyright notices to all
tracked .go, .sh, .js, .mjs, .ts, and .tsx source files.

Introduce scripts/verify-license.sh and scripts/add-license.sh for
automated verification and bulk addition of license headers. Integrate
the check into CI (ci.yml) and the Makefile (license-check target as
a prerequisite of the existing check target).

This satisfies the OpenSSF Best Practices Badge requirements for
copyright_per_file and license_per_file.

* fix: restore execute permissions on scripts

* docs: add license header instructions to CONTRIBUTING guides

* docs: add license header instructions to pages contributing guides

* fix(pages): strip unclosed HTML comment markers to satisfy CodeQL

* fix: apply code review suggestions for license scripts

- Fix portability: detect macOS vs Linux stat for permission copy
- Fix has_header: check both SPDX and copyright (match verify logic)
- Fix is_ignored: match on path boundaries to avoid false positives
- Fix year extraction: use consistent pipeline across both scripts
- Fix Bash 3.2 compat: quote array length expansion for set -u

* fix(pages): use loop-until-clean for HTML comment stripping (CodeQL)

* fix(pages): use split/join instead of replace to avoid CodeQL false positive

CodeQL's js/incomplete-multi-character-sanitization rule flags any
.replace() that removes multi-character sequences like '<!--...-->',
regardless of context. The data here comes from readFileSync on the
project's own index.html (no untrusted input), making this a false
positive. Using split(regex).join('') achieves the same result without
triggering the taint-tracking rule.
2026-08-05 21:26:27 +08:00

142 lines
4.5 KiB
Go

// SPDX-License-Identifier: Apache-2.0
// Copyright 2026 alibaba/open-code-review Contributors
package delegate
import (
"testing"
"github.com/alibaba/open-code-review/internal/config/rules"
)
// stubResolver implements rules.Resolver only.
type stubResolver struct {
ruleText string
}
func (s *stubResolver) Resolve(_ string) string { return s.ruleText }
// stubDetailResolver implements both rules.Resolver and rules.DetailResolver.
type stubDetailResolver struct {
mapping map[string]rules.RuleDetail
}
func (s *stubDetailResolver) Resolve(path string) string {
return s.mapping[path].Rule
}
func (s *stubDetailResolver) ResolveDetail(path string) rules.RuleDetail {
if d, ok := s.mapping[path]; ok {
return d
}
return rules.RuleDetail{Rule: "default-rule", Source: "system", Pattern: "default"}
}
func TestGroupRules(t *testing.T) {
tests := []struct {
name string
resolver rules.Resolver
paths []string
want []RuleGroup
}{
{
name: "empty input",
resolver: &stubResolver{ruleText: "unused"},
paths: nil,
want: nil,
},
{
name: "single file with basic resolver",
resolver: &stubResolver{ruleText: "check for bugs"},
paths: []string{"main.go"},
want: []RuleGroup{
{ID: 1, Source: "system", Pattern: "default", Text: "check for bugs", Files: []string{"main.go"}},
},
},
{
name: "multiple files same rule grouped together",
resolver: &stubResolver{ruleText: "shared rule"},
paths: []string{"a.go", "b.go", "c.go"},
want: []RuleGroup{
{ID: 1, Source: "system", Pattern: "default", Text: "shared rule", Files: []string{"a.go", "b.go", "c.go"}},
},
},
{
name: "detail resolver groups by rule text",
resolver: &stubDetailResolver{
mapping: map[string]rules.RuleDetail{
"internal/api.go": {Rule: "api rules", Source: "project", Pattern: "internal/**"},
"internal/server.go": {Rule: "api rules", Source: "project", Pattern: "internal/**"},
"test/api_test.go": {Rule: "test rules", Source: "custom", Pattern: "test/**"},
},
},
paths: []string{"internal/api.go", "internal/server.go", "test/api_test.go"},
want: []RuleGroup{
{ID: 1, Source: "project", Pattern: "internal/**", Text: "api rules", Files: []string{"internal/api.go", "internal/server.go"}},
{ID: 2, Source: "custom", Pattern: "test/**", Text: "test rules", Files: []string{"test/api_test.go"}},
},
},
{
name: "same text different sources stay separate",
resolver: &stubDetailResolver{
mapping: map[string]rules.RuleDetail{
"a.go": {Rule: "same text", Source: "project", Pattern: "*.go"},
"b.go": {Rule: "same text", Source: "custom", Pattern: "b.*"},
},
},
paths: []string{"a.go", "b.go"},
want: []RuleGroup{
{ID: 1, Source: "project", Pattern: "*.go", Text: "same text", Files: []string{"a.go"}},
{ID: 2, Source: "custom", Pattern: "b.*", Text: "same text", Files: []string{"b.go"}},
},
},
{
name: "same text same source different patterns stay separate",
resolver: &stubDetailResolver{
mapping: map[string]rules.RuleDetail{
"src/main.go": {Rule: "go rule", Source: "custom", Pattern: "src/**/*.go"},
"cmd/main.go": {Rule: "go rule", Source: "custom", Pattern: "cmd/**/*.go"},
},
},
paths: []string{"src/main.go", "cmd/main.go"},
want: []RuleGroup{
{ID: 1, Source: "custom", Pattern: "src/**/*.go", Text: "go rule", Files: []string{"src/main.go"}},
{ID: 2, Source: "custom", Pattern: "cmd/**/*.go", Text: "go rule", Files: []string{"cmd/main.go"}},
},
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
got := GroupRules(tt.resolver, tt.paths)
if len(got) != len(tt.want) {
t.Fatalf("got %d groups, want %d", len(got), len(tt.want))
}
for i := range got {
g, w := got[i], tt.want[i]
if g.ID != w.ID {
t.Errorf("group[%d].ID = %d, want %d", i, g.ID, w.ID)
}
if g.Source != w.Source {
t.Errorf("group[%d].Source = %q, want %q", i, g.Source, w.Source)
}
if g.Pattern != w.Pattern {
t.Errorf("group[%d].Pattern = %q, want %q", i, g.Pattern, w.Pattern)
}
if g.Text != w.Text {
t.Errorf("group[%d].Text = %q, want %q", i, g.Text, w.Text)
}
if len(g.Files) != len(w.Files) {
t.Errorf("group[%d].Files len = %d, want %d", i, len(g.Files), len(w.Files))
continue
}
for j := range g.Files {
if g.Files[j] != w.Files[j] {
t.Errorf("group[%d].Files[%d] = %q, want %q", i, j, g.Files[j], w.Files[j])
}
}
}
})
}
}