Skip to content

An AI skill that turns books into animated, narrated, interactive web experiences

fix(scripts): silent beats ({ dur }, no clip) in export, lint, music and check

yzr78b17186c17e
diff --git a/.claude/skills/papermorph/scripts/check.py b/.claude/skills/papermorph/scripts/check.py
index 794cc23..a0b767f 100644
--- a/.claude/skills/papermorph/scripts/check.py
+++ b/.claude/skills/papermorph/scripts/check.py
@@ -11,7 +11,8 @@ stall due to ReadPixels" (preserveDrawingBuffer, which the export needs) and thr
Then, from the first browser, each beat's voice length (pre + its clip) against its real length (P.end). A beat runs to
its last tween plus HOLD (0.6 s), so a tween or a clockRun past D lengthens it silently: the overrun past voice + HOLD
is printed, and marked when over 1 s (a helper without the beat's D: clockRun(f, 0, { tail: true }) runs to the end
-and never lengthens it). Mirrored beats meant to run alike show here too.
+and never lengthens it). Mirrored beats meant to run alike show here too. A silent beat ({ dur: s }, no clip) counts
+its dur as its voice and is labelled so.
"""
import argparse, asyncio, re, sys
from playwright.async_api import async_playwright
@@ -34,7 +35,7 @@ async def chapter(pg, cur, url):
cur["beat"] = bid
await pg.goto(f"{url}?beat={i}&t=0"); await pg.wait_for_timeout(300)
try:
- end, pre, dur = await pg.evaluate("[P.end, BEATS[P.i].pre || 0, TIMINGS[BEATS[P.i].id].dur]")
+ end, pre, dur, quiet = await pg.evaluate("[P.end, BEATS[P.i].pre || 0, TIMINGS[BEATS[P.i].id].dur, !!TIMINGS[BEATS[P.i].id].silent]")
except Exception as e:
cur["errs"].append((bid, str(e).splitlines()[0])); continue
for t in (end / 2, end - .05):
@@ -42,7 +43,7 @@ async def chapter(pg, cur, url):
await pg.evaluate(f"P.t = {t}; evalTo({t})")
except Exception as e:
cur["errs"].append((f"{bid} t={t:.2f}", str(e).splitlines()[0]))
- rows.append((bid, pre + dur, end))
+ rows.append((bid + (" (silent)" if quiet else ""), pre + dur, end))
return rows
diff --git a/.claude/skills/papermorph/scripts/export.py b/.claude/skills/papermorph/scripts/export.py
index d1c8cdd..6d870b8 100644
--- a/.claude/skills/papermorph/scripts/export.py
+++ b/.claude/skills/papermorph/scripts/export.py
@@ -113,10 +113,11 @@ async def main(a):
video, part = out / f'{a.chapter}.mp4', out / f'{a.chapter}.tmp.mp4' # renamed to video when the encoder is done
# Pass 1: the sound track and the captions, beat by beat at the beat's length on screen.
end_s = END_SECONDS if asks or a.link else 0 # a picture book ends on its own last scene
- lengths, starts, counts, music_off, cues, offset, waits = [], [], [], [], [], 0.0, {}
+ lengths, starts, counts, music_off, cues, offset, waits, silent = [], [], [], [], [], 0.0, {}, set()
for i, bid in beats:
await open_beat(pg, base, i)
end, T = await pg.evaluate('[P.end, TIMINGS[BEATS[P.i].id]]')
+ if T.get('silent'): silent.add(bid) # { dur } in the beat's fourth slot: no clip, its sound is silence
if info['beats'][i]['ask']: waits[bid] = end; end += info['beats'][i]['hold'] # from P.end the page waits
# a scene lists its sfx() when its beat is built, so they are collected beat by beat
for e in await pg.evaluate(SFX_JS):
@@ -136,7 +137,7 @@ async def main(a):
track = np.zeros((round(total * RATE) + 1, 2), np.float32)
for k, (i, bid) in enumerate(beats):
o, pre = starts[k], info['beats'][i]['pre']
- put(track, o + pre, decode(f'{a.url}{a.chapter}/audio/{info["lang"]}/{bid}.mp3'), 1, o + lengths[k])
+ if bid not in silent: put(track, o + pre, decode(f'{a.url}{a.chapter}/audio/{info["lang"]}/{bid}.mp3'), 1, o + lengths[k])
for e in info['sfx']: # effects sit at half volume under the voices; a new scene stops them
if e['beat'] == bid and 0 <= e['at'] < lengths[k]:
put(track, o + e['at'], decode(e['src']), .5 * e['vol'], o + lengths[k])
diff --git a/.claude/skills/papermorph/scripts/lint.py b/.claude/skills/papermorph/scripts/lint.py
index 20502d9..d8f11b3 100644
--- a/.claude/skills/papermorph/scripts/lint.py
+++ b/.claude/skills/papermorph/scripts/lint.py
@@ -10,7 +10,8 @@ Errors (exit status 1):
declared by the engine or another of its scripts; a second const/let/class stops the page script, a
second function silently replaces one
math the chapter uses a maths helper from lib/math.js but does not load it after engine.js
- beat a beat id has no entry in audio/<lang>/timings.js or no MP3
+ beat a beat id has no entry in audio/<lang>/timings.js or no MP3 (a silent beat, { dur: s } in its fourth slot,
+ needs neither)
mark m('name') is not a mark of any beat
stale the narration's text differs from timings.js, or a narration [[mark]] is missing there: run tts.py again
style style.json is not valid JSON or lacks schema, tier, book, words, look or voice
@@ -115,6 +116,22 @@ def decls(text):
return [(name, kind, code.count('\n', 0, at) + 1) for name, kind, at in sorted(out, key=lambda d: d[2])]
+def slots(code, start):
+ """The top-level slots of the array literal opening at code[start] ('['), in code with strings and comments blanked."""
+ depth, out, last = 0, [], start + 1
+ for k in range(start, len(code)):
+ c = code[k]
+ if c in '([{':
+ depth += 1
+ elif c in ')]}':
+ depth -= 1
+ if depth == 0:
+ return out + [code[last:k]]
+ elif c == ',' and depth == 1:
+ out.append(code[last:k]); last = k + 1
+ return out
+
+
def timings(path):
t = path.read_text(encoding='utf-8')
T = json.loads(t[t.index('{'):t.rindex('}') + 1])
@@ -169,14 +186,18 @@ def lint_chapter(ch, content, out):
# beats, audio, marks
lang = (re.search(r"""language:\s*['"](\w+)['"]""", inline) or [None, 'en'])[1]
tpath = ch / 'audio' / lang / 'timings.js'
- if not tpath.exists():
- return err('beat', f'no audio/{lang}/timings.js')
- T = timings(tpath)
starts = [(m.start(), m.group(1)) for m in BEAT.finditer(inline)]
ids = [b for _, b in starts]
+ code = depth0(inline)[0]
+ silent = {b for pos, b in starts if len(s := slots(code, pos)) > 3 and re.search(r'\bdur\s*:', s[3])} # { dur: s }: no clip
+ if not tpath.exists() and not (ids and set(ids) <= silent):
+ return err('beat', f'no audio/{lang}/timings.js')
+ T = timings(tpath) if tpath.exists() else {}
if not ids:
warn('beat', "no beats recognised (expected ['id', 'title', (m, D) => …]); marks were not checked")
for b in ids:
+ if b in silent and b not in T:
+ continue
if b not in T:
err('beat', f'beat {b} has no timings')
elif not (ch / 'audio' / lang / f'{b}.mp3').exists():
diff --git a/.claude/skills/papermorph/scripts/music.py b/.claude/skills/papermorph/scripts/music.py
index 223067c..81ce238 100644
--- a/.claude/skills/papermorph/scripts/music.py
+++ b/.claude/skills/papermorph/scripts/music.py
@@ -51,7 +51,7 @@ def chapter_url(url, ch):
async def timeline(url):
- """[(beat id, P.end, pre)] for every beat, the chapter's language and its first music() file, gain and beat count."""
+ """[(beat id, P.end, pre, silent)] for every beat (silent: { dur }, no clip), the chapter's language and its first music() file, gain and beat count."""
from playwright.async_api import async_playwright
answers(url)
async with async_playwright() as p:
@@ -65,7 +65,7 @@ async def timeline(url):
beats = []
for i in range(n):
await pg.goto(f"{url}?beat={i}&t=0"); await pg.wait_for_timeout(300)
- beats.append(tuple(await ask("[BEATS[P.i].id, P.end, BEATS[P.i].pre || 0]")))
+ beats.append(tuple(await ask("[BEATS[P.i].id, P.end, BEATS[P.i].pre || 0, !!TIMINGS[BEATS[P.i].id].silent]")))
await b.close()
if errs: print(f"page errors: {'; '.join(sorted(set(errs))[:3])}", file=sys.stderr)
return beats, lang, bgm
@@ -112,7 +112,7 @@ def spans(end, at, rests):
def cut(a):
if a.length is None:
beats, _, _ = asyncio.run(timeline(a.url))
- a.length = sum(e for _, e, _ in beats)
+ a.length = sum(b[1] for b in beats)
end, ends = (a.end, None) if a.end is not None else (last_note(a.src), sound_end(a.src))
at = a.length - a.shift # where the last note falls in the cut
found = silences(a.src, a.rests, .25, max(0, end - at - 120), at + 120) if a.rests is not None else [] # ponytail: rests up to 2 min before the in-point, enough for a few short ones
@@ -155,9 +155,10 @@ def momentary(src, gain=1.0):
def gaps(a):
import numpy as np
beats, lang, bgm = asyncio.run(timeline(a.url))
- L, SR, hop, w = sum(e for _, e, _ in beats), 16000, 800, 6400 # 400 ms windows every 50 ms
+ L, SR, hop, w = sum(b[1] for b in beats), 16000, 800, 6400 # 400 ms windows every 50 ms
voice, t0 = np.zeros(int((L + 1) * SR), np.float32), 0.0
- for bid, end, pre in beats:
+ for bid, end, pre, silent in beats:
+ if silent: t0 += end; continue # no clip: all of it is a gap
x = np.frombuffer(subprocess.run(["ffmpeg", "-v", "error", "-i", f"{a.url}audio/{lang}/{bid}.mp3", "-ac", "1", "-ar", str(SR),
"-f", "f32le", "-"], capture_output=True, check=True).stdout, np.float32)
i = int((t0 + pre) * SR); x = x[:max(0, min(len(x), int((t0 + end) * SR) - i))]; voice[i:i + len(x)] += x; t0 += end