add root readme

cht/summary/output.py

@@ -0,0 +1,75 @@
+"""Format merged segments as a sequential LLM-ready transcript.
+
+Direct port of mts/meetus/transcript_merger.py:_format_detailed (line 249).
+"""
+
+import json
+import logging
+from pathlib import Path
+
+log = logging.getLogger(__name__)
+
+
+def format_detailed(merged_segments: list[dict], *, frames_relative_to: Path | None = None) -> str:
+    """Render the interleaved transcript.
+
+    If `frames_relative_to` is given, frame paths are rewritten relative to it.
+    """
+    lines = []
+    lines.append("=" * 80)
+    lines.append("ENHANCED MEETING TRANSCRIPT")
+    lines.append("Audio transcript + Screen frames")
+    lines.append("=" * 80)
+    lines.append("")
+
+    for seg in merged_segments:
+        ts = _format_timestamp(seg["timestamp"])
+        if seg["type"] == "audio":
+            speaker = seg.get("speaker") or "SPEAKER"
+            lines.append(f"[{ts}] {speaker}:")
+            lines.append(f"  {seg['text']}")
+            lines.append("")
+        else:
+            lines.append(f"[{ts}] SCREEN CONTENT:")
+            fp = seg.get("frame_path")
+            if fp:
+                if frames_relative_to is not None:
+                    try:
+                        fp = str(Path(fp).resolve().relative_to(frames_relative_to.resolve()))
+                    except ValueError:
+                        fp = str(fp)
+                else:
+                    fp = str(fp)
+                lines.append(f"  Frame: {fp}")
+            lines.append("")
+
+    return "\n".join(lines)
+
+
+def _format_timestamp(seconds: float) -> str:
+    seconds = int(seconds)
+    h, rem = divmod(seconds, 3600)
+    m, s = divmod(rem, 60)
+    if h:
+        return f"{h:02d}:{m:02d}:{s:02d}"
+    return f"{m:02d}:{s:02d}"
+
+
+def write_outputs(session_dir: Path, merged: list[dict], *, name: str | None = None) -> Path:
+    """Write `<name>_enhanced.txt` and `merged.json` under `session_dir/summary`.
+
+    Returns the path of the enhanced transcript.
+    """
+    summary_dir = session_dir / "summary"
+    summary_dir.mkdir(parents=True, exist_ok=True)
+    name = name or session_dir.name
+
+    text = format_detailed(merged, frames_relative_to=session_dir)
+    text_path = summary_dir / f"{name}_enhanced.txt"
+    text_path.write_text(text)
+
+    merged_path = summary_dir / "merged.json"
+    merged_path.write_text(json.dumps(merged, indent=2, default=str))
+
+    log.info("Wrote %s (%d entries)", text_path, len(merged))
+    return text_path