feat(parse): add ExtractImages stream scanner for OSC 9 markers

Implements ExtractImages() which scans filter stdout line-by-line,
extracts valid OSC 9 image markers into ImageRef structs, and replaces
marker lines with null-delimited placeholders (\x00IMG:N\x00). Invalid
markers (missing path) pass through unchanged. Also clarifies the Index
field comment in ImageRef to indicate it is set by ExtractImages, not
ParseImageOSC.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

lib/parse/extract_images.go

@@ -0,0 +1,41 @@
+package parse
+
+import (
+	"bufio"
+	"bytes"
+	"fmt"
+	"io"
+
+	"git.sr.ht/~rjarry/aerc/lib/log"
+)
+
+// ExtractImages scans a reader for OSC 9 image markers, replaces them with
+// null-byte-delimited placeholders (\x00IMG:N\x00), and returns the cleaned
+// reader plus a slice of ImageRef structs.
+//
+// Lines containing valid image markers are replaced entirely with the
+// placeholder. Lines with invalid markers (e.g. missing path) pass through
+// unchanged.
+func ExtractImages(r io.Reader) (io.Reader, []ImageRef) {
+	var images []ImageRef
+	buf := bytes.NewBuffer(nil)
+	scanner := bufio.NewScanner(r)
+	scanner.Buffer(nil, 1024*1024*1024)
+	for scanner.Scan() {
+		line := scanner.Bytes()
+		if bytes.Contains(line, oscImagePrefix) {
+			if ref, ok := ParseImageOSC(line); ok {
+				ref.Index = len(images)
+				images = append(images, *ref)
+				fmt.Fprintf(buf, "\x00IMG:%d\x00\n", ref.Index)
+				continue
+			}
+		}
+		buf.Write(line)
+		buf.WriteByte('\n')
+	}
+	if err := scanner.Err(); err != nil {
+		log.Warnf("ExtractImages: scan error: %v", err)
+	}
+	return buf, images
+}

lib/parse/extract_images_test.go

@@ -0,0 +1,101 @@
+package parse
+
+import (
+	"io"
+	"strings"
+	"testing"
+)
+
+func TestExtractImages_NoMarkers(t *testing.T) {
+	input := "Hello\nWorld\nNo images here\n"
+	r, images := ExtractImages(strings.NewReader(input))
+	if len(images) != 0 {
+		t.Errorf("got %d images, want 0", len(images))
+	}
+	out, _ := io.ReadAll(r)
+	if string(out) != input {
+		t.Errorf("output = %q, want %q", string(out), input)
+	}
+}
+
+func TestExtractImages_SingleMarker(t *testing.T) {
+	input := "Line 1\n\033]9;image:path=/tmp/img.png;alt=photo\007\nLine 2\n"
+	r, images := ExtractImages(strings.NewReader(input))
+	if len(images) != 1 {
+		t.Fatalf("got %d images, want 1", len(images))
+	}
+	if images[0].Path != "/tmp/img.png" {
+		t.Errorf("path = %q", images[0].Path)
+	}
+	if images[0].Alt != "photo" {
+		t.Errorf("alt = %q", images[0].Alt)
+	}
+	if images[0].Index != 0 {
+		t.Errorf("index = %d, want 0", images[0].Index)
+	}
+	out, _ := io.ReadAll(r)
+	expected := "Line 1\n\x00IMG:0\x00\nLine 2\n"
+	if string(out) != expected {
+		t.Errorf("output = %q, want %q", string(out), expected)
+	}
+}
+
+func TestExtractImages_MultipleMarkers(t *testing.T) {
+	input := "Text\n\033]9;image:path=/tmp/a.png\007\nMiddle\n\033]9;image:path=/tmp/b.jpg;alt=second\007\nEnd\n"
+	r, images := ExtractImages(strings.NewReader(input))
+	if len(images) != 2 {
+		t.Fatalf("got %d images, want 2", len(images))
+	}
+	if images[0].Index != 0 || images[1].Index != 1 {
+		t.Errorf("indices: %d, %d", images[0].Index, images[1].Index)
+	}
+	if images[1].Path != "/tmp/b.jpg" {
+		t.Errorf("second path = %q", images[1].Path)
+	}
+	out, _ := io.ReadAll(r)
+	if !strings.Contains(string(out), "\x00IMG:0\x00") {
+		t.Error("missing placeholder 0")
+	}
+	if !strings.Contains(string(out), "\x00IMG:1\x00") {
+		t.Error("missing placeholder 1")
+	}
+}
+
+func TestExtractImages_EmptyInput(t *testing.T) {
+	r, images := ExtractImages(strings.NewReader(""))
+	if len(images) != 0 {
+		t.Errorf("got %d images, want 0", len(images))
+	}
+	out, _ := io.ReadAll(r)
+	if len(out) != 0 {
+		t.Errorf("expected empty output, got %q", string(out))
+	}
+}
+
+func TestExtractImages_MarkerAtStartAndEnd(t *testing.T) {
+	input := "\033]9;image:path=/tmp/first.png\007\nContent\n\033]9;image:path=/tmp/last.png\007\n"
+	r, images := ExtractImages(strings.NewReader(input))
+	if len(images) != 2 {
+		t.Fatalf("got %d images, want 2", len(images))
+	}
+	out, _ := io.ReadAll(r)
+	lines := strings.Split(strings.TrimSuffix(string(out), "\n"), "\n")
+	if lines[0] != "\x00IMG:0\x00" {
+		t.Errorf("first line = %q, want placeholder", lines[0])
+	}
+	if lines[2] != "\x00IMG:1\x00" {
+		t.Errorf("last line = %q, want placeholder", lines[2])
+	}
+}
+
+func TestExtractImages_InvalidMarkerPassesThrough(t *testing.T) {
+	input := "Text\n\033]9;image:alt=no path\007\nMore text\n"
+	r, images := ExtractImages(strings.NewReader(input))
+	if len(images) != 0 {
+		t.Errorf("got %d images, want 0 (invalid marker)", len(images))
+	}
+	out, _ := io.ReadAll(r)
+	if !strings.Contains(string(out), "Text") {
+		t.Error("text content lost")
+	}
+}

lib/parse/imageref.go

@@ -7,7 +7,7 @@ import (
 
 // ImageRef represents an image referenced by a filter via OSC 9 marker.
 type ImageRef struct {
-	Index int    // Sequential position in the message
+	Index int    // Sequential position in the message; set by ExtractImages, not ParseImageOSC
 	Path  string // Absolute path to image file on disk
 	Alt   string // Alt text fallback
 }