feat: core/waveform — peak reduction + ffmpeg region decode

core/waveform.py

@@ -0,0 +1,35 @@
+import subprocess
+
+import numpy as np
+
+from core.paths import _bin
+
+
+def peaks(samples, buckets: int = 128) -> list[float]:
+    """Reduce a 1-D sample array to *buckets* normalized peak magnitudes (0..1)."""
+    if samples is None or len(samples) == 0:
+        return [0.0] * buckets
+    a = np.abs(np.asarray(samples, dtype="float32"))
+    idx = np.linspace(0, len(a), buckets + 1).astype(int)
+    out = [float(a[idx[i]:idx[i + 1]].max()) if idx[i + 1] > idx[i] else 0.0
+           for i in range(buckets)]
+    m = max(out) or 1.0
+    return [v / m for v in out]
+
+
+def load_region_samples(path: str, start: float, duration: float,
+                        sr: int = 8000) -> np.ndarray:
+    """Decode a [start, start+duration] mono slice via ffmpeg for waveform
+    preview (low sample rate). Returns a float32 numpy array, empty on failure."""
+    cmd = [
+        _bin("ffmpeg"), "-ss", str(start), "-i", path, "-t", str(duration),
+        "-vn", "-ac", "1", "-ar", str(sr), "-f", "f32le",
+        "-loglevel", "error", "pipe:1",
+    ]
+    try:
+        proc = subprocess.run(cmd, capture_output=True, timeout=60)
+    except Exception:
+        return np.zeros(0, dtype="float32")
+    if proc.returncode != 0 or not proc.stdout:
+        return np.zeros(0, dtype="float32")
+    return np.frombuffer(proc.stdout, dtype="float32")