feat: html output (v1)

README.md

@@ -21,6 +21,12 @@ Find structurally duplicate functions in Go code.
 godedup ./...
 ```
 
+**HTML output**
+
+![HTML](https://raw.githubusercontent.com/hashmap-kz/assets/main/godedup/godedup-html-v1.png)
+
+**Table output with clickable `file.go:line` locations in supported terminals and editors**
+
 ```
 $ godedup --output=table --exclude '_test\.go$'
 
@@ -83,14 +89,15 @@ Examples:
   godedup --exclude '_test\.go$' --exclude '\.pb\.go$' ./...
   godedup --exclude '(_test|[.]pb|[.]deepcopy)[.]go$' ./...
   godedup --output table ./...
+  godedup --output html ./... > godedup.html
   godedup --output json ./... | jq .
 
 Flags:
   --min-similarity  float    minimum similarity threshold (default: 0.85)
   --min-stmts       int      minimum statements to analyze (default: 3)
   --exact                    report only exact structural clones
   --exclude                  exclude files matching regexp (may be repeated)
-  --output          string   output format: text, table, json (default: text)
+  --output          string   output format: text, table, html, json (default: text)
   --version                  print version
 ```
 

internal/hash/hash.go

@@ -23,6 +23,7 @@ type FuncInfo struct {
 	StmtSeq  []uint64 // per-statement hashes for similarity comparison
 	NumStmts int      // total statement count (excluding blank lines)
 	NumLines int      // line span of the function body
+	Source   string   // original function source, used by rich reports
 }
 
 // Hasher computes structural hashes of AST nodes.

internal/load/load.go

@@ -1,6 +1,7 @@
 package load
 
 import (
+	"bytes"
 	"go/ast"
 	"go/parser"
 	"go/token"
@@ -79,7 +80,12 @@ func parseFile(path string, fset *token.FileSet, hasher *hash.Hasher, inp *cmd.L
 		return nil
 	}
 
-	f, err := parser.ParseFile(fset, path, nil, 0)
+	src, err := os.ReadFile(path)
+	if err != nil {
+		return err
+	}
+
+	f, err := parser.ParseFile(fset, path, src, 0)
 	if err != nil {
 		// skip unparseable files
 		return nil
@@ -99,6 +105,7 @@ func parseFile(path string, fset *token.FileSet, hasher *hash.Hasher, inp *cmd.L
 		}
 
 		info := hasher.HashFunc(pkg, path, fn)
+		info.Source = sourceSpan(src, fset, fn)
 		if info.Name == "" {
 			continue
 		}
@@ -107,3 +114,12 @@ func parseFile(path string, fset *token.FileSet, hasher *hash.Hasher, inp *cmd.L
 
 	return nil
 }
+
+func sourceSpan(src []byte, fset *token.FileSet, fn *ast.FuncDecl) string {
+	start := fset.Position(fn.Pos()).Offset
+	end := fset.Position(fn.End()).Offset
+	if start < 0 || end < start || end > len(src) {
+		return ""
+	}
+	return string(bytes.TrimRight(src[start:end], "\n"))
+}

internal/load/load_test.go

@@ -4,6 +4,7 @@ import (
 	"os"
 	"path/filepath"
 	"regexp"
+	"strings"
 	"testing"
 
 	"github.com/hashmap-kz/godedup/internal/cmd"
@@ -164,6 +165,29 @@ func Vendored() int {
 	}
 }
 
+func TestLoadStoresFunctionSource(t *testing.T) {
+	dir := t.TempDir()
+	file := filepath.Join(dir, "source.go")
+	writeFile(t, file, `package sample
+func Source() int {
+	a := 1
+	b := 2
+	return a + b
+}
+`)
+
+	result, err := Load([]string{file}, emptyLoadInput())
+	if err != nil {
+		t.Fatalf("Load() error = %v", err)
+	}
+	if len(result.Funcs) != 1 {
+		t.Fatalf("len(Funcs) = %d, want 1", len(result.Funcs))
+	}
+	if got := result.Funcs[0].Source; !strings.Contains(got, "func Source() int") || !strings.Contains(got, "return a + b") {
+		t.Fatalf("Source was not captured correctly:\n%s", got)
+	}
+}
+
 func TestLoadSingleFile(t *testing.T) {
 	dir := t.TempDir()
 	file := filepath.Join(dir, "single.go")

internal/report/consts.go

@@ -0,0 +1,7 @@
+package report
+
+const (
+	kindExact     = "EXACT"
+	kindNear      = "NEAR"
+	sim100Percent = "100%"
+)