refactor: split processor and matcher

Author: ldez

pkg/commands/flagsets.go

@@ -102,7 +102,7 @@ func setupIssuesFlagSet(v *viper.Viper, fs *pflag.FlagSet) {
 	internal.AddFlagAndBind(v, fs, fs.Bool, "exclude-dirs-use-default", "issues.exclude-dirs-use-default", true,
 		formatList("Use or not use default excluded directories:", processors.StdExcludeDirRegexps))
 
-	internal.AddFlagAndBind(v, fs, fs.String, "exclude-generated", "issues.exclude-generated", processors.AutogeneratedModeLax,
+	internal.AddFlagAndBind(v, fs, fs.String, "exclude-generated", "issues.exclude-generated", config.GeneratedModeLax,
 		color.GreenString("Mode of the generated files analysis"))
 
 	const newDesc = "Show only new issues: if there are unstaged changes or untracked files, only those changes " +

pkg/config/linters_exclusions.go

@@ -5,6 +5,12 @@ import (
 	"slices"
 )
 
+const (
+	GeneratedModeLax     = "lax"
+	GeneratedModeStrict  = "strict"
+	GeneratedModeDisable = "disable"
+)
+
 const (
 	ExclusionPresetComments             = "comments"
 	ExclusionPresetStdErrorHandling     = "stdErrorHandling"

pkg/config/loader.go

@@ -70,7 +70,7 @@ func (l *Loader) Load(opts LoadOptions) error {
 
 	if l.cfg.Linters.LinterExclusions.Generated == "" {
 		// `l.cfg.Issues.ExcludeGenerated` is always non-empty because of the flag default value.
-		l.cfg.Linters.LinterExclusions.Generated = cmp.Or(l.cfg.Issues.ExcludeGenerated, "strict")
+		l.cfg.Linters.LinterExclusions.Generated = cmp.Or(l.cfg.Issues.ExcludeGenerated, GeneratedModeStrict)
 	}
 
 	// Compatibility layer with v1.

pkg/result/processors/exclusion_generated_file_filter.go

@@ -2,61 +2,36 @@ package processors
 
 import (
 	"fmt"
-	"go/parser"
-	"go/token"
 	"path/filepath"
-	"regexp"
-	"strings"
 
+	"github.com/golangci/golangci-lint/pkg/config"
 	"github.com/golangci/golangci-lint/pkg/logutils"
 	"github.com/golangci/golangci-lint/pkg/result"
 )
 
-const (
-	AutogeneratedModeLax     = "lax"
-	AutogeneratedModeStrict  = "strict"
-	AutogeneratedModeDisable = "disable"
-)
-
-// The values must be in lowercase.
-const (
-	genCodeGenerated = "code generated"
-	genDoNotEdit     = "do not edit"
-
-	// Related to easyjson.
-	genAutoFile = "autogenerated file"
-
-	//nolint:lll // Long URL
-	// Related to Swagger Codegen.
-	// https://github.com/swagger-api/swagger-codegen/blob/61cfeac3b9d855b4eb8bffa0d118bece117bcb7d/modules/swagger-codegen/src/main/resources/go/partial_header.mustache#L16
-	// https://github.com/swagger-api/swagger-codegen/issues/12358
-	genSwaggerCodegen = "* generated by: swagger codegen "
-)
-
 var _ Processor = (*GeneratedFileFilter)(nil)
 
 type fileSummary struct {
 	generated bool
 }
 
 // GeneratedFileFilter filters generated files.
-//   - mode "lax": see `isGeneratedFileLax` documentation.
-//   - mode "strict": see `isGeneratedFileStrict` documentation.
-//   - mode "disable": skips this processor.
 type GeneratedFileFilter struct {
 	debugf logutils.DebugFunc
 
-	mode          string
-	strictPattern *regexp.Regexp
+	mode    string
+	matcher *GeneratedFileMatcher
 
 	fileSummaryCache map[string]*fileSummary
 }
 
 func NewGeneratedFileFilter(mode string) *GeneratedFileFilter {
 	return &GeneratedFileFilter{
-		debugf:           logutils.Debug(logutils.DebugKeyGeneratedFileFilter),
-		mode:             mode,
-		strictPattern:    regexp.MustCompile(`^// Code generated .* DO NOT EDIT\.$`),
+		debugf: logutils.Debug(logutils.DebugKeyGeneratedFileFilter),
+
+		mode:    mode,
+		matcher: NewGeneratedFileMatcher(mode),
+
 		fileSummaryCache: map[string]*fileSummary{},
 	}
 }
@@ -66,7 +41,7 @@ func (*GeneratedFileFilter) Name() string {
 }
 
 func (p *GeneratedFileFilter) Process(issues []result.Issue) ([]result.Issue, error) {
-	if p.mode == AutogeneratedModeDisable {
+	if p.mode == config.GeneratedModeDisable {
 		return issues, nil
 	}
 
@@ -89,93 +64,14 @@ func (p *GeneratedFileFilter) shouldPassIssue(issue *result.Issue) (bool, error)
 	fs = &fileSummary{}
 	p.fileSummaryCache[issue.FilePath()] = fs
 
-	if p.mode == AutogeneratedModeStrict {
-		var err error
-		fs.generated, err = p.isGeneratedFileStrict(issue.FilePath())
-		if err != nil {
-			return false, fmt.Errorf("failed to get doc (strict) of file %s: %w", issue.FilePath(), err)
-		}
-	} else {
-		doc, err := getComments(issue.FilePath())
-		if err != nil {
-			return false, fmt.Errorf("failed to get doc (lax) of file %s: %w", issue.FilePath(), err)
-		}
-
-		fs.generated = p.isGeneratedFileLax(doc)
+	var err error
+	fs.generated, err = p.matcher.IsGeneratedFile(issue.FilePath(), nil)
+	if err != nil {
+		return false, fmt.Errorf("failed to get doc (%s) of file %s: %w", p.mode, issue.FilePath(), err)
 	}
 
 	p.debugf("file %q is generated: %t", issue.FilePath(), fs.generated)
 
 	// don't report issues for autogenerated files
 	return !fs.generated, nil
 }
-
-// isGeneratedFileLax reports whether the source file is generated code.
-// The function uses a bit laxer rules than isGeneratedFileStrict to match more generated code.
-// See https://github.com/golangci/golangci-lint/issues/48 and https://github.com/golangci/golangci-lint/issues/72.
-func (p *GeneratedFileFilter) isGeneratedFileLax(doc string) bool {
-	markers := []string{genCodeGenerated, genDoNotEdit, genAutoFile, genSwaggerCodegen}
-
-	doc = strings.ToLower(doc)
-
-	for _, marker := range markers {
-		if strings.Contains(doc, marker) {
-			p.debugf("doc contains marker %q: file is generated", marker)
-
-			return true
-		}
-	}
-
-	p.debugf("doc of len %d doesn't contain any of markers: %s", len(doc), markers)
-
-	return false
-}
-
-// isGeneratedFileStrict returns true if the source file has a line that matches the regular expression:
-//
-//	^// Code generated .* DO NOT EDIT\.$
-//
-// This line must appear before the first non-comment, non-blank text in the file.
-// Based on https://go.dev/s/generatedcode.
-func (p *GeneratedFileFilter) isGeneratedFileStrict(filePath string) (bool, error) {
-	file, err := parser.ParseFile(token.NewFileSet(), filePath, nil, parser.PackageClauseOnly|parser.ParseComments)
-	if err != nil {
-		return false, fmt.Errorf("failed to parse file: %w", err)
-	}
-
-	if file == nil || len(file.Comments) == 0 {
-		return false, nil
-	}
-
-	for _, comment := range file.Comments {
-		if comment.Pos() > file.Package {
-			return false, nil
-		}
-
-		for _, line := range comment.List {
-			generated := p.strictPattern.MatchString(line.Text)
-			if generated {
-				p.debugf("doc contains ignore expression: file is generated")
-
-				return true, nil
-			}
-		}
-	}
-
-	return false, nil
-}
-
-func getComments(filePath string) (string, error) {
-	fset := token.NewFileSet()
-	syntax, err := parser.ParseFile(fset, filePath, nil, parser.PackageClauseOnly|parser.ParseComments)
-	if err != nil {
-		return "", fmt.Errorf("failed to parse file: %w", err)
-	}
-
-	var docLines []string
-	for _, c := range syntax.Comments {
-		docLines = append(docLines, strings.TrimSpace(c.Text()))
-	}
-
-	return strings.Join(docLines, "\n"), nil
-}

pkg/result/processors/exclusion_generated_file_filter_test.go

@@ -10,145 +10,10 @@ import (
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 
+	"github.com/golangci/golangci-lint/pkg/config"
 	"github.com/golangci/golangci-lint/pkg/result"
 )
 
-func TestGeneratedFileFilter_isGeneratedFileLax_generated(t *testing.T) {
-	p := NewGeneratedFileFilter(AutogeneratedModeLax)
-
-	comments := []string{
-		`	// generated by stringer -type Pill pill.go; DO NOT EDIT`,
-		`// Code generated by "stringer -type Pill pill.go"; DO NOT EDIT`,
-		`// Code generated by vfsgen; DO NOT EDIT`,
-		`// Created by cgo -godefs - DO NOT EDIT`,
-		`/* Created by cgo - DO NOT EDIT. */`,
-		`// Generated by stringer -i a.out.go -o anames.go -p ppc64
-// Do not edit.`,
-		`// DO NOT EDIT
-// generated by: x86map -fmt=decoder ../x86.csv`,
-		`// DO NOT EDIT.
-// Generate with: go run gen.go -full -output md5block.go`,
-		`// generated by "go run gen.go". DO NOT EDIT.`,
-		`// DO NOT EDIT. This file is generated by mksyntaxgo from the RE2 distribution.`,
-		`// GENERATED BY make_perl_groups.pl; DO NOT EDIT.`,
-		`// generated by mknacl.sh - do not edit`,
-		`// DO NOT EDIT ** This file was generated with the bake tool ** DO NOT EDIT //`,
-		`// Generated by running
-//  maketables --tables=all --data=http://www.unicode.org/Public/8.0.0/ucd/UnicodeData.txt
-// --casefolding=http://www.unicode.org/Public/8.0.0/ucd/CaseFolding.txt
-// DO NOT EDIT`,
-		`/*
-* CODE GENERATED AUTOMATICALLY WITH github.com/ernesto-jimenez/gogen/unmarshalmap
-* THIS FILE SHOULD NOT BE EDITED BY HAND
-  */`,
-		`// AUTOGENERATED FILE: easyjson file.go`,
-		` * Generated by: Swagger Codegen (https://github.com/swagger-api/swagger-codegen.git)`,
-	}
-
-	for _, comment := range comments {
-		t.Run(comment, func(t *testing.T) {
-			t.Parallel()
-
-			generated := p.isGeneratedFileLax(comment)
-			assert.True(t, generated)
-		})
-	}
-}
-
-func TestGeneratedFileFilter_isGeneratedFileLax_nonGenerated(t *testing.T) {
-	p := NewGeneratedFileFilter(AutogeneratedModeLax)
-
-	comments := []string{
-		"code not generated by",
-		"test",
-	}
-
-	for _, comment := range comments {
-		t.Run(comment, func(t *testing.T) {
-			t.Parallel()
-
-			generated := p.isGeneratedFileLax(comment)
-			assert.False(t, generated)
-		})
-	}
-}
-
-func TestGeneratedFileFilter_isGeneratedFileStrict(t *testing.T) {
-	p := NewGeneratedFileFilter(AutogeneratedModeStrict)
-
-	testCases := []struct {
-		desc     string
-		filepath string
-		assert   assert.BoolAssertionFunc
-	}{
-		{
-			desc:     "go strict",
-			filepath: filepath.FromSlash("testdata/exclusion_generated_file_filter/go_strict.go"),
-			assert:   assert.True,
-		},
-		{
-			desc:     "go strict invalid",
-			filepath: filepath.FromSlash("testdata/exclusion_generated_file_filter/go_strict_invalid.go"),
-			assert:   assert.False,
-		},
-	}
-
-	for _, test := range testCases {
-		t.Run(test.desc, func(t *testing.T) {
-			t.Parallel()
-
-			generated, err := p.isGeneratedFileStrict(test.filepath)
-			require.NoError(t, err)
-
-			test.assert(t, generated)
-		})
-	}
-}
-
-func Test_getComments(t *testing.T) {
-	testCases := []struct {
-		fpath string
-		doc   string
-	}{
-		{
-			fpath: filepath.FromSlash("testdata/exclusion_generated_file_filter/exclude.go"),
-			doc: `first line
-second line
-third line
-this text also
-and this text also`,
-		},
-		{
-			fpath: filepath.FromSlash("testdata/exclusion_generated_file_filter/exclude_doc.go"),
-			doc:   `DO NOT EDIT`,
-		},
-		{
-			fpath: filepath.FromSlash("testdata/exclusion_generated_file_filter/exclude_block_comment.go"),
-			doc: `* first line
- *
- * second line
- * third line
-and this text also
-this type of block comment also
-this one line comment also`,
-		},
-	}
-
-	for _, tc := range testCases {
-		doc, err := getComments(tc.fpath)
-		require.NoError(t, err)
-		assert.Equal(t, tc.doc, doc)
-	}
-}
-
-// Issue 954: Some lines can be very long, e.g. auto-generated
-// embedded resources. Reported on file of 86.2KB.
-func Test_getComments_fileWithLongLine(t *testing.T) {
-	fpath := filepath.FromSlash("testdata/exclusion_generated_file_filter/exclude_long_line.go")
-	_, err := getComments(fpath)
-	assert.NoError(t, err)
-}
-
 func TestGeneratedFileFilter_shouldPassIssue(t *testing.T) {
 	testCases := []struct {
 		desc   string
@@ -158,7 +23,7 @@ func TestGeneratedFileFilter_shouldPassIssue(t *testing.T) {
 	}{
 		{
 			desc: "lax ",
-			mode: AutogeneratedModeLax,
+			mode: config.GeneratedModeLax,
 			issue: &result.Issue{
 				FromLinter: "example",
 				Pos: token.Position{
@@ -169,7 +34,7 @@ func TestGeneratedFileFilter_shouldPassIssue(t *testing.T) {
 		},
 		{
 			desc: "strict ",
-			mode: AutogeneratedModeStrict,
+			mode: config.GeneratedModeStrict,
 			issue: &result.Issue{
 				FromLinter: "example",
 				Pos: token.Position{
@@ -208,7 +73,7 @@ func TestGeneratedFileFilter_shouldPassIssue_error(t *testing.T) {
 	}{
 		{
 			desc: "non-existing file (lax)",
-			mode: AutogeneratedModeLax,
+			mode: config.GeneratedModeLax,
 			issue: &result.Issue{
 				FromLinter: "example",
 				Pos: token.Position{
@@ -220,7 +85,7 @@ func TestGeneratedFileFilter_shouldPassIssue_error(t *testing.T) {
 		},
 		{
 			desc: "non-existing file (strict)",
-			mode: AutogeneratedModeStrict,
+			mode: config.GeneratedModeStrict,
 			issue: &result.Issue{
 				FromLinter: "example",
 				Pos: token.Position{