Improve text safety and add video fades

movmaker.py

@@ -208,15 +208,19 @@ def media_date(path: Path) -> datetime:
     return dt
 
 
-def slugify(value: str) -> str:
+def ascii_safe(value: str) -> str:
     replacements = {
-        "æ": "ae", "Æ": "ae",
+        "æ": "ae", "Æ": "Ae",
-        "ø": "o", "Ø": "o",
+        "ø": "o", "Ø": "O",
-        "å": "a", "Å": "a",
+        "å": "a", "Å": "A",
-        "đ": "d", "Đ": "d",
+        "đ": "d", "Đ": "D",
     }
     value = "".join(replacements.get(ch, ch) for ch in value)
-    value = unicodedata.normalize("NFKD", value).encode("ascii", "ignore").decode("ascii")
+    return unicodedata.normalize("NFKD", value).encode("ascii", "ignore").decode("ascii")
+
+
+def slugify(value: str) -> str:
+    value = ascii_safe(value)
     value = value.strip().lower()
     value = re.sub(r"[^a-z0-9]+", "_", value)
     value = value.strip("_")
@@ -316,8 +320,13 @@ def parse_data_file(path: Path | None) -> MovieData:
 
 
 def drawtext_escape(text: str) -> str:
-    # Escape for ffmpeg drawtext text='...'
+    # Escape for ffmpeg drawtext text=... without surrounding quotes.
-    return text.replace("\\", "\\\\").replace(":", "\\:").replace("'", "\\'").replace("%", "\\%")
+    return (
+        text.replace("\\", "\\\\")
+        .replace(":", "\\:")
+        .replace(",", "\\,")
+        .replace("%", "\\%")
+    )
 
 
 def normalize_clip(item: MediaItem, out: Path, width: int, height: int, fps: int, preset: str, crf: int) -> None:
@@ -369,17 +378,29 @@ def build_video(
     audio: Path | None,
     output: Path,
     data: MovieData,
+    creation_date: datetime,
     captions: list[str],
     fade: float,
     audio_fade: float,
+    audio_fade_in: float,
+    video_fade: float,
     width: int,
     height: int,
     fps: int,
     preset: str,
     crf: int,
+    text_dir: Path,
 ) -> None:
     output.parent.mkdir(parents=True, exist_ok=True)
 
+    text_file_count = 0
+    def textfile_arg(text: str) -> str:
+        nonlocal text_file_count
+        path = text_dir / f"drawtext_{text_file_count:04d}.txt"
+        text_file_count += 1
+        path.write_text(ascii_safe(text), encoding="utf-8")
+        return f"textfile='{path}':expansion=none"
+
     cmd = [tool_path(FFMPEG, "ffmpeg"), "-y"]
     for clip in clips:
         cmd += ["-i", str(clip)]
@@ -407,6 +428,15 @@ def build_video(
             current = next_label
             total_duration += durations[i] - fade
 
+    fade_label = current
+    if video_fade > 0:
+        fade_out_start = max(0.0, total_duration - video_fade)
+        fade_label = "video_faded"
+        filters.append(
+            f"[{current}]fade=t=in:st=0:d={video_fade:.3f},fade=t=out:st={fade_out_start:.3f}:d={video_fade:.3f}[{fade_label}]"
+        )
+        current = fade_label
+
     title_lines = [x for x in [data.title, data.date, data.place] if x]
     final_label = "vout"
     title_label = "titled"
@@ -422,28 +452,28 @@ def build_video(
         if subtitle:
             title_y = f"(h-text_h)/2-{title_font_size // 2}"
             filters.append(
-                f"[{current}]drawtext=fontfile='{title_font}':text='{drawtext_escape(main_title)}':x=(w-text_w)/2:y={title_y}:"
+                f"[{current}]drawtext=fontfile='{title_font}':{textfile_arg(main_title)}:x=(w-text_w)/2:y={title_y}:"
                 f"fontsize={title_font_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=3:shadowy=3:"
                 f"alpha='{title_alpha}':enable='between(t,0,6)'[title0]"
             )
             filters.append(
-                f"[title0]drawtext=fontfile='{title_font}':text='{drawtext_escape(main_title)}':x=(w-text_w)/2+1:y={title_y}:"
+                f"[title0]drawtext=fontfile='{title_font}':{textfile_arg(main_title)}:x=(w-text_w)/2+1:y={title_y}:"
                 f"fontsize={title_font_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=3:shadowy=3:"
                 f"alpha='{title_alpha}':enable='between(t,0,6)'[title0b]"
             )
             filters.append(
-                f"[title0b]drawtext=fontfile='{title_font}':text='{drawtext_escape(subtitle)}':x=(w-text_w)/2:y=(h-text_h)/2+{subtitle_font_size}:"
+                f"[title0b]drawtext=fontfile='{title_font}':{textfile_arg(subtitle)}:x=(w-text_w)/2:y=(h-text_h)/2+{subtitle_font_size}:"
                 f"fontsize={subtitle_font_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=2:shadowy=2:"
                 f"alpha='{title_alpha}':enable='between(t,0,6)'[{title_label}]"
             )
         else:
             filters.append(
-                f"[{current}]drawtext=fontfile='{title_font}':text='{drawtext_escape(main_title)}':x=(w-text_w)/2:y=(h-text_h)/2:"
+                f"[{current}]drawtext=fontfile='{title_font}':{textfile_arg(main_title)}:x=(w-text_w)/2:y=(h-text_h)/2:"
                 f"fontsize={title_font_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=3:shadowy=3:"
                 f"alpha='{title_alpha}':enable='between(t,0,6)'[title0]"
             )
             filters.append(
-                f"[title0]drawtext=fontfile='{title_font}':text='{drawtext_escape(main_title)}':x=(w-text_w)/2+1:y=(h-text_h)/2:"
+                f"[title0]drawtext=fontfile='{title_font}':{textfile_arg(main_title)}:x=(w-text_w)/2+1:y=(h-text_h)/2:"
                 f"fontsize={title_font_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=3:shadowy=3:"
                 f"alpha='{title_alpha}':enable='between(t,0,6)'[{title_label}]"
             )
@@ -481,13 +511,13 @@ def build_video(
             y_expr = f"h-text_h-62-{block_height - line_idx * line_gap}"
             next_label = f"caption{idx}_{line_idx}"
             filters.append(
-                f"[{caption_label}]drawtext=fontfile='{stamp_bold_font}':text='{drawtext_escape(caption_line)}':x=(w-text_w)/2:y={y_expr}:"
+                f"[{caption_label}]drawtext=fontfile='{stamp_bold_font}':{textfile_arg(caption_line)}:x=(w-text_w)/2:y={y_expr}:"
                 f"fontsize={caption_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=2:shadowy=2:"
                 f"alpha='{caption_alpha}':enable='between(t,{start:.3f},{end:.3f})'[{next_label}]"
             )
             bold_label = f"caption{idx}_{line_idx}b"
             filters.append(
-                f"[{next_label}]drawtext=fontfile='{stamp_bold_font}':text='{drawtext_escape(caption_line)}':x=(w-text_w)/2+1:y={y_expr}:"
+                f"[{next_label}]drawtext=fontfile='{stamp_bold_font}':{textfile_arg(caption_line)}:x=(w-text_w)/2+1:y={y_expr}:"
                 f"fontsize={caption_size}:fontcolor=white:borderw=0:shadowcolor=black@0.75:shadowx=2:shadowy=2:"
                 f"alpha='{caption_alpha}':enable='between(t,{start:.3f},{end:.3f})'[{bold_label}]"
             )
@@ -499,22 +529,33 @@ def build_video(
     stamp_left_label = "stamp_left"
     if stamp_text:
         filters.append(
-            f"[{caption_label}]drawtext=fontfile='{stamp_font}':text='{stamp_text}':x=24:y=h-text_h-18:"
+            f"[{caption_label}]drawtext=fontfile='{stamp_font}':{textfile_arg(' | '.join(stamp_lines))}:x=24:y=h-text_h-18:"
             f"fontsize={stamp_size}:fontcolor=white:borderw=1:bordercolor=black[{stamp_left_label}]"
         )
     else:
         filters.append(f"[{caption_label}]copy[{stamp_left_label}]")
     filters.append(
-        f"[{stamp_left_label}]drawtext=fontfile='{stamp_font}':text='@bubulescu':x=w-text_w-24:y=h-text_h-18:"
+        f"[{stamp_left_label}]drawtext=fontfile='{stamp_font}':{textfile_arg('@bubulescu')}:x=w-text_w-24:y=h-text_h-18:"
         f"fontsize={stamp_size}:fontcolor=white:borderw=1:bordercolor=black[{final_label}]"
     )
 
     cmd += ["-filter_complex", ";".join(filters), "-map", f"[{final_label}]"]
+    metadata = {
+        "title": ascii_safe(data.title),
+        "date": ascii_safe(data.date),
+        "location": ascii_safe(data.place),
+        "artist": "@bubulescu",
+        "comment": "Created with movmaker",
+        "creation_time": creation_date.isoformat(),
+    }
+    for key, value in metadata.items():
+        if value:
+            cmd += ["-metadata", f"{key}={value}"]
 
     if audio:
         audio_index = len(clips)
         audio_fade_out_start = max(0.0, total_duration - audio_fade)
-        audio_filter = f"afade=t=in:st=0:d={audio_fade:.3f},afade=t=out:st={audio_fade_out_start:.3f}:d={audio_fade:.3f}"
+        audio_filter = f"afade=t=in:st=0:d={audio_fade_in:.3f},afade=t=out:st={audio_fade_out_start:.3f}:d={audio_fade:.3f}"
         cmd += ["-map", f"{audio_index}:a:0", "-af", audio_filter, "-c:a", "aac", "-b:a", "192k", "-shortest"]
     else:
         cmd += ["-an"]
@@ -530,7 +571,9 @@ def main(argv: list[str] | None = None) -> int:
     parser.add_argument("--out-dir", type=Path, default=Path("out"), help="directory for auto-named output videos")
     parser.add_argument("--image-duration", type=float, default=6.0, help="seconds each image is visible")
     parser.add_argument("--fade", type=float, default=3.0, help="crossfade duration in seconds")
-    parser.add_argument("--audio-fade", type=float, default=10.0, help="music fade-in/fade-out duration in seconds")
+    parser.add_argument("--audio-fade", type=float, default=10.0, help="music fade-out duration in seconds")
+    parser.add_argument("--audio-fade-in", type=float, default=2.0, help="music fade-in duration in seconds")
+    parser.add_argument("--video-fade", type=float, default=1.5, help="video fade-in/fade-out duration in seconds")
     parser.add_argument("--first-transition-at", type=float, default=12.0, help="seconds before first transition starts")
     parser.add_argument("--resolution", type=parse_resolution, default=(1920, 1080), help="output resolution, e.g. 1920x1080")
     parser.add_argument("--fps", type=int, default=30, help="output frames per second")
@@ -554,10 +597,13 @@ def main(argv: list[str] | None = None) -> int:
         crf = 28
 
     media, audio_files, data = scan_input(args.input, args.image_duration)
+    first_media_date = media_date(media[0].path)
     if not data.date:
-        data.date = display_date(media_date(media[0].path))
+        data.date = display_date(first_media_date)
     fade = min(args.fade, max(0.0, args.image_duration - 0.1))
     audio_fade = max(0.0, args.audio_fade)
+    audio_fade_in = max(0.0, args.audio_fade_in)
+    video_fade = max(0.0, args.video_fade)
 
     output = args.output or default_output_path(media[0].path, data, args.out_dir)
 
@@ -602,7 +648,7 @@ def main(argv: list[str] | None = None) -> int:
 
         captions = [data.captions.get(item.path.name, data.captions.get(str(item.path), "")) for item in media]
         audio = concat_or_mix_audio(audio_files, tmp / "music.m4a")
-        build_video(clips, durations, audio, output, data, captions, fade, audio_fade, width, height, fps, preset, crf)
+        build_video(clips, durations, audio, output, data, first_media_date, captions, fade, audio_fade, audio_fade_in, video_fade, width, height, fps, preset, crf, tmp)
 
     print(f"Done: {output}")
     return 0