refactor: share .records file discovery (ask u3)

Extract ListNonEmptyFiles and HostFromFileName into internal/recordsdir
for aggregate and storage ImportFromDir. Behavior unchanged.

Made-with: Cursor

4 files changed, 107 insertions, 33 deletions

diff --git a/internal/goprecords/aggregate.go b/internal/goprecords/aggregate.go
index cefb545..31a0e1c 100644
--- a/internal/goprecords/aggregate.go
+++ b/internal/goprecords/aggregate.go
@@ -5,10 +5,9 @@ import (
 	"context"
 	"fmt"
 	"os"
-	"path/filepath"
-	"strings"
 
 	"codeberg.org/snonux/goprecords/internal/recordline"
+	"codeberg.org/snonux/goprecords/internal/recordsdir"
 )
 
 // Aggregates holds all category maps. Host uses HostAggregate; others use Aggregate.
@@ -37,23 +36,13 @@ func (ag *Aggregator) Aggregate(ctx context.Context) (*Aggregates, error) {
 		KernelMajor: make(map[string]*Aggregate),
 		KernelName:  make(map[string]*Aggregate),
 	}
-	entries, err := os.ReadDir(ag.statsDir)
+	files, err := recordsdir.ListNonEmptyFiles(ag.statsDir)
 	if err != nil {
 		return nil, fmt.Errorf("read stats dir: %w", err)
 	}
-	for _, e := range entries {
-		if e.IsDir() || !strings.HasSuffix(e.Name(), ".records") {
-			continue
-		}
-		path := filepath.Join(ag.statsDir, e.Name())
-		info, err := os.Stat(path)
-		if err != nil || info.Size() == 0 {
-			continue
-		}
-		host := strings.TrimSuffix(e.Name(), filepath.Ext(e.Name()))
-		if idx := strings.Index(host, "."); idx > 0 {
-			host = host[:idx]
-		}
+	for _, f := range files {
+		host := f.Host
+		path := f.Path
 		if _, exists := out.Host[host]; exists {
 			return nil, fmt.Errorf("record file for %s already processed - duplicate inputs?", host)
 		}
diff --git a/internal/recordsdir/recordsdir.go b/internal/recordsdir/recordsdir.go
new file mode 100644
index 0000000..9f3ce5b
--- /dev/null
+++ b/internal/recordsdir/recordsdir.go
@@ -0,0 +1,40 @@
+package recordsdir
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+)
+
+type Entry struct {
+	Path string
+	Host string
+}
+
+func HostFromFileName(name string) string {
+	host := strings.TrimSuffix(name, filepath.Ext(name))
+	if idx := strings.Index(host, "."); idx > 0 {
+		host = host[:idx]
+	}
+	return host
+}
+
+func ListNonEmptyFiles(dir string) ([]Entry, error) {
+	entries, err := os.ReadDir(dir)
+	if err != nil {
+		return nil, err
+	}
+	var out []Entry
+	for _, e := range entries {
+		if e.IsDir() || !strings.HasSuffix(e.Name(), ".records") {
+			continue
+		}
+		path := filepath.Join(dir, e.Name())
+		info, err := os.Stat(path)
+		if err != nil || info.Size() == 0 {
+			continue
+		}
+		out = append(out, Entry{Path: path, Host: HostFromFileName(e.Name())})
+	}
+	return out, nil
+}
diff --git a/internal/recordsdir/recordsdir_test.go b/internal/recordsdir/recordsdir_test.go
new file mode 100644
index 0000000..bee8d67
--- /dev/null
+++ b/internal/recordsdir/recordsdir_test.go
@@ -0,0 +1,58 @@
+package recordsdir
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+)
+
+func TestHostFromFileName(t *testing.T) {
+	tests := []struct {
+		file string
+		want string
+	}{
+		{"earth.records", "earth"},
+		{"myhost.example.com.records", "myhost"},
+		{"single.records", "single"},
+	}
+	for _, tt := range tests {
+		t.Run(tt.file, func(t *testing.T) {
+			if got := HostFromFileName(tt.file); got != tt.want {
+				t.Fatalf("HostFromFileName(%q) = %q, want %q", tt.file, got, tt.want)
+			}
+		})
+	}
+}
+
+func TestListNonEmptyFiles(t *testing.T) {
+	dir := t.TempDir()
+	if err := os.WriteFile(filepath.Join(dir, "skip.txt"), []byte("x"), 0o644); err != nil {
+		t.Fatal(err)
+	}
+	if err := os.WriteFile(filepath.Join(dir, "empty.records"), nil, 0o644); err != nil {
+		t.Fatal(err)
+	}
+	if err := os.WriteFile(filepath.Join(dir, "h1.records"), []byte("line\n"), 0o644); err != nil {
+		t.Fatal(err)
+	}
+	if err := os.Mkdir(filepath.Join(dir, "nested.records"), 0o755); err != nil {
+		t.Fatal(err)
+	}
+	entries, err := ListNonEmptyFiles(dir)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(entries) != 1 {
+		t.Fatalf("len(entries) = %d, want 1: %#v", len(entries), entries)
+	}
+	if entries[0].Host != "h1" || filepath.Base(entries[0].Path) != "h1.records" {
+		t.Fatalf("unexpected entry: %#v", entries[0])
+	}
+}
+
+func TestListNonEmptyFiles_ReadError(t *testing.T) {
+	_, err := ListNonEmptyFiles(filepath.Join(t.TempDir(), "nonexistent-subdir-nope"))
+	if err == nil {
+		t.Fatal("expected error")
+	}
+}
diff --git a/internal/storage/db.go b/internal/storage/db.go
index 7b577e7..fe789d6 100644
--- a/internal/storage/db.go
+++ b/internal/storage/db.go
@@ -6,10 +6,9 @@ import (
 	"database/sql"
 	"fmt"
 	"os"
-	"path/filepath"
-	"strings"
 
 	"codeberg.org/snonux/goprecords/internal/recordline"
+	"codeberg.org/snonux/goprecords/internal/recordsdir"
 	_ "modernc.org/sqlite"
 )
 
@@ -67,7 +66,7 @@ func ImportFromDir(ctx context.Context, db *sql.DB, statsDir string) error {
 	if err := ResetRecords(ctx, db); err != nil {
 		return fmt.Errorf("reset records: %w", err)
 	}
-	entries, err := os.ReadDir(statsDir)
+	files, err := recordsdir.ListNonEmptyFiles(statsDir)
 	if err != nil {
 		return fmt.Errorf("read dir: %w", err)
 	}
@@ -81,20 +80,8 @@ func ImportFromDir(ctx context.Context, db *sql.DB, statsDir string) error {
 		return fmt.Errorf("prepare insert: %w", err)
 	}
 	defer insert.Close()
-	for _, e := range entries {
-		if e.IsDir() || !strings.HasSuffix(e.Name(), ".records") {
-			continue
-		}
-		path := filepath.Join(statsDir, e.Name())
-		info, err := os.Stat(path)
-		if err != nil || info.Size() == 0 {
-			continue
-		}
-		host := strings.TrimSuffix(e.Name(), filepath.Ext(e.Name()))
-		if idx := strings.Index(host, "."); idx > 0 {
-			host = host[:idx]
-		}
-		if err := importFile(ctx, insert, path, host); err != nil {
+	for _, f := range files {
+		if err := importFile(ctx, insert, f.Path, f.Host); err != nil {
 			return err
 		}
 	}