Initial commit: GnommoEditor video pipeline

A code-first, declarative video editing system that compiles text documents into rendered video via FFmpeg. Uses a compiler-style ETL pipeline: Extract (parse inputs) → Validate → Transform (build timeline) → Render (FFmpeg). Features: - Text-based project definition (manuscript, transcript, JSON configs) - Slide markers [S1], [S2] in transcript map to timed overlays - Strict validation with fail-fast error reporting - FFmpeg filter_complex generation with time-based enables - CLI with validate/render/dry-run modes Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-01-12 11:19:38 +01:00
commit d5a8d38c9c
15 changed files with 967 additions and 0 deletions
@@ -0,0 +1,88 @@
+"""Transform stage: resolve timings and build render plan."""
+
+from pathlib import Path
+
+from .models import (
+    ProjectConfig,
+    RenderPlan,
+    SlideDefinition,
+    SlideEvent,
+    TimedWord,
+    VideoSource,
+)
+from .parser import get_video_duration
+
+
+def build_render_plan(
+    project_path: Path,
+    config: ProjectConfig,
+    slides: dict[str, SlideDefinition],
+    videos: dict[str, VideoSource],
+    transcript: list[TimedWord],
+) -> RenderPlan:
+    """
+    Build a complete render plan from parsed and validated data.
+
+    This transforms transcript markers into timed slide events and
+    assembles all information needed for the render stage.
+    """
+    # For POC: use the first video as the talking head
+    talking_head_id = next(iter(videos.keys()))
+    talking_head = videos[talking_head_id]
+
+    # Get video duration for end time calculations
+    video_path = project_path / talking_head.file
+    total_duration = get_video_duration(video_path)
+
+    # Build slide events from transcript markers
+    slide_events = _extract_slide_events(transcript, slides, total_duration)
+
+    return RenderPlan(
+        project_path=project_path,
+        config=config,
+        talking_head=talking_head,
+        slide_events=slide_events,
+        total_duration=total_duration,
+        slides=slides,
+    )
+
+
+def _extract_slide_events(
+    transcript: list[TimedWord],
+    slides: dict[str, SlideDefinition],
+    total_duration: float,
+) -> list[SlideEvent]:
+    """
+    Extract slide events from transcript markers.
+
+    Each marker like [S1] in the transcript becomes a SlideEvent with:
+    - start_time: timestamp of the marker
+    - end_time: timestamp of next marker, or end of video
+    """
+    # Find all markers in transcript
+    marker_times: list[tuple[float, str]] = []
+
+    for timed_word in transcript:
+        if timed_word.is_marker:
+            marker_id = timed_word.marker_id
+            if marker_id and marker_id in slides:
+                marker_times.append((timed_word.time, marker_id))
+
+    # Convert markers to slide events
+    events: list[SlideEvent] = []
+
+    for i, (start_time, marker_id) in enumerate(marker_times):
+        # End time is start of next marker, or end of video
+        if i + 1 < len(marker_times):
+            end_time = marker_times[i + 1][0]
+        else:
+            end_time = total_duration
+
+        events.append(SlideEvent(
+            slide_id=marker_id,
+            start_time=start_time,
+            end_time=end_time,
+            slide_def=slides[marker_id],
+        ))
+
+    return events