Add pipeline code

_pipeline/audiomeasure.py

@@ -0,0 +1,51 @@
+"""Measure audio with ffmpeg: RMS and peak per 50 ms window. The checker and the assembler both use this."""
+import math
+import os
+import re
+import subprocess
+import tempfile
+
+
+def ffmpeg(args):
+    r = subprocess.run(["ffmpeg", "-nostdin", "-hide_banner", "-loglevel", "error", "-y"] + args, capture_output=True, text=True)
+    if r.returncode != 0:
+        raise RuntimeError("ffmpeg failed: " + r.stderr.strip()[-500:])
+
+
+def probe(path):
+    out = subprocess.run(["ffprobe", "-v", "error", "-select_streams", "a:0", "-show_entries",
+                          "stream=codec_name,sample_rate,channels,bit_rate:format=duration", "-of", "default=nw=1", path],
+                         capture_output=True, text=True, check=True).stdout
+    return dict(line.split("=", 1) for line in out.strip().splitlines())
+
+
+def windows(path):
+    """[(rms_db, peak_db)] for consecutive 50 ms windows of the decoded file."""
+    sr = int(probe(path)["sample_rate"])
+    n = sr // 20
+    with tempfile.TemporaryDirectory() as tmp:
+        rms_file, peak_file = os.path.join(tmp, "rms.txt"), os.path.join(tmp, "peak.txt")
+        af = ("asetnsamples=n=%d:p=0,astats=metadata=1:reset=1,"
+              "ametadata=print:key=lavfi.astats.Overall.RMS_level:file=%s,"
+              "ametadata=print:key=lavfi.astats.Overall.Peak_level:file=%s" % (n, rms_file, peak_file))
+        r = subprocess.run(["ffmpeg", "-nostdin", "-hide_banner", "-loglevel", "error", "-i", path, "-af", af, "-f", "null", "-"],
+                           capture_output=True, text=True)
+        if r.returncode != 0:
+            raise RuntimeError("ffmpeg failed: " + r.stderr.strip()[-500:])
+        read = lambda p, key: [float(v) for v in re.findall(key + r"=(-?[\d.]+|-inf|inf)", open(p).read())]
+        rms, peak = read(rms_file, "RMS_level"), read(peak_file, "Peak_level")
+    return list(zip(rms, peak))
+
+
+def db_mean_power(values):
+    """Overall level in dB of equal-length windows given as dB RMS."""
+    if not values:
+        return float("-inf")
+    p = sum(10 ** (v / 10) for v in values) / len(values)
+    return 10 * math.log10(p) if p > 0 else float("-inf")
+
+
+def overall(path):
+    """(rms_db, peak_db) of the whole file."""
+    w = windows(path)
+    return db_mean_power([r for r, _ in w]), max(p for _, p in w)