DemonGatanjieuandClaude Opus 5.5
c59927e701
AMB read models only from ordinary drop-downs, so rgthree's Power Lora Loader (rows are
{on, lora, strength} objects) was invisible: Model Check never reported its missing LoRAs.
Now its switched-on rows are Model Check boxes (lora_nodes.js powerLoraRowBoxes: ComfyUI's LoRA
list as choices, a fix changes only the row's LoRA; downloads put files in the same way), saved
fingerprints and a picture's model search read them, the workflow summary lists them, and saved
model references include them and Easy-Use LoRA stack rows (nodeModelValues /
_model_reference_values, mirrored). Only these widely used nodes; ComfyUI's loaders stay the
reference.
recipe_parser.js was over 600 lines: which LoRAs a node loads moves to recipe_loras.js and the
box readers to recipe_widgets.js.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
122 lines
5.8 KiB
Python
122 lines
5.8 KiB
Python
import json
|
|
from pathlib import Path
|
|
import sys
|
|
import tempfile
|
|
import unittest
|
|
from unittest import mock
|
|
|
|
from PIL import Image, PngImagePlugin
|
|
|
|
PLUGIN = Path(__file__).resolve().parents[1]
|
|
sys.path[:0] = [str(PLUGIN), str(PLUGIN.parents[1])]
|
|
from api import image_search, model_resolution
|
|
|
|
|
|
def _png(path, **texts):
|
|
info = PngImagePlugin.PngInfo()
|
|
for key, value in texts.items():
|
|
info.add_text(key, value)
|
|
Image.new('RGB', (4, 4)).save(path, pnginfo=info)
|
|
|
|
|
|
PROMPT = json.dumps({
|
|
'3': {'class_type': 'KSampler', 'inputs': {'seed': 123456789, 'steps': 25, 'model': ['4', 0]}},
|
|
'4': {'class_type': 'CheckpointLoaderSimple', 'inputs': {'ckpt_name': '2D_Anime\\MiaoMiao RealSkin.safetensors'}},
|
|
'6': {'class_type': 'CLIPTextEncode', 'inputs': {'text': '1girl, 赛博朋克 city, neon'}},
|
|
})
|
|
WORKFLOW = json.dumps({'extra': {'anomalous_hashes': {
|
|
'4_2D_Anime\\MiaoMiao RealSkin.safetensors': {'hash': '667B5E3BEDE5A2FD53BB6F0EA5B93A89453D6855FA2E5D95ACDE93CF543C4F7C', 'size': 1},
|
|
}}})
|
|
|
|
|
|
class ImageSearchTests(unittest.TestCase):
|
|
def setUp(self):
|
|
self._tmp = tempfile.TemporaryDirectory()
|
|
self.root = Path(self._tmp.name)
|
|
(self.root / 'sub').mkdir()
|
|
_png(self.root / 'a.png', prompt=PROMPT, workflow=WORKFLOW)
|
|
_png(self.root / 'sub' / 'b.png', prompt=json.dumps({'1': {'class_type': 'CheckpointLoaderSimple', 'inputs': {'ckpt_name': 'renamed_copy.safetensors'}}}))
|
|
_png(self.root / 'c.png', parameters='a cat in rain\nSteps: 20, Seed: 42, Model hash: abcdef1234, Model: dreamshaper')
|
|
Image.new('RGB', (4, 4)).save(self.root / 'plain.jpg')
|
|
self.images = [
|
|
{'filename': 'a.png', 'subfolder': ''},
|
|
{'filename': 'b.png', 'subfolder': 'sub'},
|
|
{'filename': 'c.png', 'subfolder': ''},
|
|
{'filename': 'plain.jpg', 'subfolder': ''},
|
|
]
|
|
image_search._record_cache.clear()
|
|
image_search._model_index.update(expires=0, entries={})
|
|
patcher = mock.patch.object(model_resolution, 'collect_model_hash_index', return_value={
|
|
'renamed_copy.safetensors': {'hash': '667b5e3bede5a2fd53bb6f0ea5b93a89453d6855fa2e5d95acde93cf543c4f7c'},
|
|
})
|
|
patcher.start()
|
|
self.addCleanup(patcher.stop)
|
|
|
|
def tearDown(self):
|
|
self._tmp.cleanup()
|
|
|
|
def names(self, query):
|
|
return [image['filename'] for image in image_search.filter_images(str(self.root), self.images, query)]
|
|
|
|
def test_power_lora_loader_rows_name_their_loras(self):
|
|
prompt = json.dumps({'9': {'class_type': 'Power Lora Loader (rgthree)', 'inputs': {
|
|
'PowerLoraLoaderHeaderWidget': {'type': 'PowerLoraLoaderHeaderWidget'},
|
|
'lora_1': {'on': True, 'lora': 'Style\\ink wash.safetensors', 'strength': 0.8},
|
|
'model': ['4', 0]}}})
|
|
_png(self.root / 'd.png', prompt=prompt)
|
|
record = image_search.build_record(str(self.root / 'd.png'))
|
|
self.assertIn('style/ink wash.safetensors', record['models'])
|
|
self.assertIn('ink wash.safetensors', record['models'])
|
|
|
|
def test_prompt_text_models_and_seed(self):
|
|
self.assertEqual(self.names('赛博朋克'), ['a.png'])
|
|
self.assertEqual(self.names('miaomiao realskin'), ['a.png'], 'all terms must match')
|
|
self.assertEqual(self.names('123456789'), ['a.png'])
|
|
self.assertEqual(self.names('seed:123456789'), ['a.png'])
|
|
self.assertEqual(self.names('neon cat'), [])
|
|
|
|
def test_search_blocks_keep_phrases_whole(self):
|
|
# A list is one phrase per block; a plain string still splits on spaces.
|
|
self.assertEqual(self.names(['city, neon']), ['a.png'])
|
|
self.assertEqual(self.names(['neon city']), [])
|
|
self.assertEqual(self.names('neon city'), ['a.png'])
|
|
self.assertEqual(self.names(['cat in rain', 'Seed: 42']), ['c.png'])
|
|
self.assertEqual(image_search.parse_query([' A b ', 'a b', '']), ['a b'])
|
|
|
|
def test_a1111_parameters_and_file_names(self):
|
|
self.assertEqual(self.names('cat'), ['c.png'])
|
|
self.assertEqual(self.names('abcdef1234'), ['c.png'], 'A1111 model hash text')
|
|
self.assertEqual(self.names('plain'), ['plain.jpg'])
|
|
|
|
def test_hash_matches_recorded_hashes_and_local_renamed_files(self):
|
|
# a.png recorded the hash; b.png only names a local file whose scanned hash matches.
|
|
self.assertEqual(self.names('667B5E3B'), ['a.png', 'b.png'])
|
|
self.assertEqual(self.names('667b5e3bede5a2fd'), ['a.png', 'b.png'])
|
|
self.assertEqual(self.names('deadbeef'), [])
|
|
|
|
def test_records_are_cached_by_mtime(self):
|
|
self.names('cat')
|
|
with mock.patch.object(image_search, 'build_record', side_effect=AssertionError('re-read')):
|
|
images = [dict(image, mtime=image_search._record_cache[str(self.root / image['subfolder'] / image['filename'])][0])
|
|
for image in self.images]
|
|
self.assertEqual([i['filename'] for i in image_search.filter_images(str(self.root), images, 'cat')], ['c.png'])
|
|
|
|
def test_only_text_chunks_before_pixels_are_read(self):
|
|
chunks = image_search.read_png_text(self.root / 'a.png')
|
|
self.assertEqual(set(chunks), {'prompt', 'workflow'})
|
|
self.assertIn('赛博朋克', json.loads(chunks['prompt'])['6']['inputs']['text'])
|
|
|
|
|
|
class ModelResolutionImportTests(unittest.TestCase):
|
|
def test_model_scan_uses_model_extensions(self):
|
|
with tempfile.TemporaryDirectory() as directory:
|
|
Path(directory, 'model.safetensors').write_bytes(b'x')
|
|
Path(directory, 'notes.txt').write_text('x')
|
|
with mock.patch.object(model_resolution.folder_paths, 'get_folder_paths', return_value=[directory]):
|
|
candidates = model_resolution._collect_resolution_candidates(['checkpoints'])
|
|
self.assertEqual([c['filename'] for c in candidates], ['model.safetensors'])
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|