File size: 7,379 Bytes
0f52e37 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 | r"""Regenerate the field tables in REFERENCE.md from the node's own schema.
REFERENCE.md went stale because it was hand-maintained alongside a 69-field node:
it documented a `total_seconds` input that had been gone for months and was missing
17 fields, including every guard. Hand-copying a schema is a losing game, so the
field-by-field part is generated from INPUT_TYPES and RETURN_NAMES instead. The
prose around it stays hand-written -- reasoning is the part worth writing by hand.
Everything between the AUTOGEN markers is replaced; everything else is untouched.
python gen_reference.py # rewrite REFERENCE.md in place
python gen_reference.py --check # exit 1 if it is out of date (for CI)
Run it from the node directory with ComfyUI importable (the portable python works:
..\..\..\python_embeded\python.exe gen_reference.py
).
"""
import importlib.util
import os
import sys
HERE = os.path.dirname(os.path.abspath(__file__))
BEGIN = "<!-- AUTOGEN:FIELDS BEGIN -- edit gen_reference.py, not this block -->"
END = "<!-- AUTOGEN:FIELDS END -->"
# Which group each input belongs to, so the tables read as sections rather than as
# one 69-row dump. First match wins; anything unmatched lands in "Other".
GROUPS = [
("Wiring", ["model", "clip", "vae", "audio_vae", "first_frame",
"ref_image_1", "ref_image_2", "ref_image_3", "ref_image_4"]),
("The scene", ["prompt", "character_memory", "anchor_override", "beat_split",
"per_beat_length"]),
("Size and length", ["resolution", "megapixels", "shot_seconds",
"allow_oversize_shots", "allow_res_backoff", "fps"]),
("Sampling", ["steps", "cfg", "sampler_name", "scheduler", "seed",
"vary_seed_per_shot", "apply_model_sampling", "shift_video",
"shift_audio"]),
("Chaining shots", ["trim_seam", "handoff_offset", "cleanup_between_shots",
"vram_headroom_gb"]),
("Audio", ["global_soundscape", "non_diegetic_music", "auto_soundscape",
"auto_silence_nonspeech", "mute_nonspeech_audio", "mute_fade_ms"]),
("Consistency guards", ["subject_count_guard", "anatomy_guard", "solidity_guard",
"motion_guard", "contact_guard", "lock_restraints",
"auto_wardrobe", "auto_props", "prevent_nudity",
"exposed_terms"]),
("Reference images", ["ref_mode", "ref_image_size", "ref_noise_aug"]),
("Decode and upscale", ["decode_tile_frames", "decode_tile_size", "upscale",
"upscale_model", "upscale_target_short_edge",
"upscale_batch"]),
("Overlays", ["watermark_text", "watermark_position", "watermark_size",
"watermark_opacity", "watermark_margin", "intro_text",
"intro_position", "intro_seconds", "intro_fade", "intro_size",
"overlay_font", "overlay_stroke"]),
("Preview", ["plan_only"]),
]
def load_node():
sys.path.insert(0, os.path.dirname(os.path.dirname(HERE))) # custom_nodes/
spec = importlib.util.spec_from_file_location(
"_h3ref", os.path.join(HERE, "__init__.py"),
submodule_search_locations=[HERE])
mod = importlib.util.module_from_spec(spec)
sys.modules["_h3ref"] = mod
spec.loader.exec_module(mod)
return mod.NODE_CLASS_MAPPINGS
def describe(spec):
"""(type, constraints) for one input's schema tuple."""
t, opts = spec[0], (spec[1] if len(spec) > 1 and isinstance(spec[1], dict) else {})
if isinstance(t, list):
return "choice", ", ".join(f"`{o}`" for o in t)
bits = []
if opts.get("forceInput"):
bits.append("**input socket**")
if "default" in opts and not opts.get("forceInput"):
d = opts["default"]
bits.append(f"default `{d!r}`" if isinstance(d, str) else f"default `{d}`")
if "min" in opts or "max" in opts:
bits.append(f"range `{opts.get('min', '-')}`–`{opts.get('max', '-')}`")
if opts.get("multiline"):
bits.append("multiline")
return t if isinstance(t, str) else "?", ", ".join(bits) or "—"
def tooltip_of(spec):
opts = spec[1] if len(spec) > 1 and isinstance(spec[1], dict) else {}
tip = " ".join((opts.get("tooltip") or "").split())
return tip or "_(no tooltip)_"
def build(cls):
it = cls.INPUT_TYPES()
req, opt = it.get("required", {}), it.get("optional", {})
everything = {**req, **opt}
seen, out = set(), []
out.append(f"There are **{len(everything)}** inputs and "
f"**{len(cls.RETURN_NAMES)}** outputs. Required inputs are marked "
f"**R**; everything else is optional and has a working default.\n")
for title, names in GROUPS:
rows = [n for n in names if n in everything]
if not rows:
continue
seen.update(rows)
out.append(f"### {title}\n")
out.append("| field | | type | constraints |")
out.append("|---|---|---|---|")
for n in rows:
t, c = describe(everything[n])
out.append(f"| `{n}` | {'**R**' if n in req else ''} | {t} | {c} |")
out.append("")
for n in rows:
tip = tooltip_of(everything[n])
if tip != "_(no tooltip)_":
out.append(f"**`{n}`** — {tip}\n")
leftover = [n for n in everything if n not in seen]
if leftover:
out.append("### Other\n")
out.append("| field | | type | constraints |")
out.append("|---|---|---|---|")
for n in leftover:
t, c = describe(everything[n])
out.append(f"| `{n}` | {'**R**' if n in req else ''} | {t} | {c} |")
out.append("")
for n in leftover:
tip = tooltip_of(everything[n])
if tip != "_(no tooltip)_":
out.append(f"**`{n}`** — {tip}\n")
out.append("### Outputs\n")
out.append("| slot | name | type |")
out.append("|---|---|---|")
for i, (n, t) in enumerate(zip(cls.RETURN_NAMES, cls.RETURN_TYPES)):
out.append(f"| {i} | `{n}` | {t} |")
out.append("")
out.append("Outputs are only ever **appended**. A workflow stores an output link "
"by slot index, so inserting one would silently re-target every wire "
"after it.\n")
return "\n".join(out)
def main(check_only=False):
mapping = load_node()
body = build(mapping["H3LongVideos"])
path = os.path.join(HERE, "REFERENCE.md")
with open(path, encoding="utf-8") as f:
doc = f.read()
if BEGIN not in doc or END not in doc:
sys.exit(f"Markers not found in REFERENCE.md. Add:\n{BEGIN}\n{END}")
head, rest = doc.split(BEGIN, 1)
_, tail = rest.split(END, 1)
new = f"{head}{BEGIN}\n\n{body}\n{END}{tail}"
if new == doc:
print("REFERENCE.md is up to date.")
return
if check_only:
sys.exit("REFERENCE.md is OUT OF DATE. Run: python gen_reference.py")
with open(path, "w", encoding="utf-8") as f:
f.write(new)
print(f"REFERENCE.md field tables regenerated "
f"({len(mapping['H3LongVideos'].INPUT_TYPES().get('required', {}))} required + "
f"{len(mapping['H3LongVideos'].INPUT_TYPES().get('optional', {}))} optional).")
if __name__ == "__main__":
main("--check" in sys.argv)
|