2 files changed, 51 insertions, 2 deletions
diff --git a/mountain_pass/extract.go b/mountain_pass/extract.go
index 705dbd3..9643603 100644
--- a/mountain_pass/extract.go
+++ b/mountain_pass/extract.go
@@ -24,14 +24,28 @@ const insertStatement = `
 		longitude = excluded.longitude
 `
 
-func ExtractMountainPasses(ctx context.Context, pbfPath string, db *sql.DB) (int, error) {
+// ExtractMountainPasses scans pbfPath for mountain pass nodes and upserts them
+// into the database. An optional reporter is invoked as scanning progresses
+// with bytes scanned and the total file size, so callers (e.g. the CLI) can
+// render progress.
+func ExtractMountainPasses(ctx context.Context, pbfPath string, db *sql.DB, reporters ...func(current, total int64)) (int, error) {
 	file, err := os.Open(pbfPath)
 	if err != nil {
 		return 0, fmt.Errorf("failed to open pbf file: %w", err)
 	}
 	defer file.Close()
 
-	scanner := osmpbf.New(ctx, file, 4)
+	info, err := file.Stat()
+	if err != nil {
+		return 0, fmt.Errorf("failed to stat pbf file: %w", err)
+	}
+	total := info.Size()
+	report := func(int64, int64) {}
+	if len(reporters) > 0 {
+		report = reporters[0]
+	}
+	reader := &countingReader{r: file}
+	scanner := osmpbf.New(ctx, reader, 4)
 	scanner.SkipWays = true
 	scanner.SkipRelations = true
 	scanner.FilterNode = isMountainPass
@@ -50,6 +64,7 @@ func ExtractMountainPasses(ctx context.Context, pbfPath string, db *sql.DB) (int
 
 	count := 0
 	for scanner.Scan() {
+		report(reader.n, total)
 		node, ok := scanner.Object().(*osm.Node)
 		if !ok {
 			continue
@@ -74,6 +89,19 @@ func ExtractMountainPasses(ctx context.Context, pbfPath string, db *sql.DB) (int
 	return count, nil
 }
 
+// countingReader tracks the number of bytes read from an underlying reader,
+// used to report scan progress against the PBF file size.
+type countingReader struct {
+	r io.Reader
+	n int64
+}
+
+func (c *countingReader) Read(p []byte) (int, error) {
+	n, err := c.r.Read(p)
+	c.n += int64(n)
+	return n, err
+}
+
 func isMountainPass(node *osm.Node) bool {
 	if node.Tags.Find("mountain_pass") == "yes" {
 		return true
diff --git a/mountain_pass/extract_test.go b/mountain_pass/extract_test.go
index 636372c..7eae33b 100644
--- a/mountain_pass/extract_test.go
+++ b/mountain_pass/extract_test.go
@@ -1,10 +1,13 @@
 package mountain_pass
 
 import (
+	"io"
+	"strings"
 	"testing"
 
 	"github.com/paulmach/osm"
 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )
 
 func TestIsMountainPass(t *testing.T) {
@@ -47,3 +50,21 @@ func TestParseElevation(t *testing.T) {
 		assert.Equal(t, c.ok, ok)
 	}
 }
+
+func TestCountingReaderTracksBytes(t *testing.T) {
+	const content = "0123456789abcdef"
+	reader := &countingReader{r: strings.NewReader(content)}
+
+	buf := make([]byte, 4)
+	var total int64
+	for {
+		n, err := reader.Read(buf)
+		total += int64(n)
+		require.Equal(t, total, reader.n)
+		if err == io.EOF {
+			break
+		}
+		require.NoError(t, err)
+	}
+	assert.Equal(t, int64(len(content)), reader.n)
+}