Files
ComfyUI-Lora-Manager/py/utils/example_images_metadata.py
T
Will Miao 7d963b27b5 fix(example-images): read real dimensions for imported videos, fixes #1115
Example videos added through the "Add examples" flow were stored with a
hardcoded 720x1280 entry. The dimension probe next to it only ran for
images (PIL cannot open .mp4/.webm files), so every video entry stayed
portrait regardless of the source. The showcase viewer then sizes its
container straight from that value (--media-aspect in showcase.css), so
landscape clips were letterboxed inside a 9:16 box. CivitAI-sourced
examples were unaffected because their dimensions come from the API.

PIL cannot read video containers, so add a dependency-free reader that
parses the container headers instead: moov/trak/tkhd for ISO base media
(with the sample description as a fallback), Segment/Tracks/Pixel* for
WebM/Matroska, and RIFF/WebP for animated examples saved with a video
extension. The sniffed signature decides which reader runs, so a .mp4
that is really WebM still reports the right size; the extension is only
a fallback. Both readers seek past mdat rather than reading it, so a
large file costs the same as a small one.

Imported entries now record the file's real size and keep the previous
placeholder only when the file cannot be parsed.

Existing libraries keep their wrong entries, so backfill them once via
the existing naming migration: bump CURRENT_NAMING_VERSION to 3 and
repair each model's empty-url entries from the files on disk, then sync
the scanner cache. Only entries with no remote url are touched -- those
have no other source, which makes the rewrite lossless -- and entries
already carrying the right size are left byte-identical, so the pass is
idempotent and a no-op for libraries that never imported a video.
2026-09-17 21:42:32 +08:00

711 lines
28 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import inspect
import logging
import os
import re
from typing import TYPE_CHECKING, Any, Dict, Mapping, MutableMapping, Optional
from ..recipes.constants import GEN_PARAM_KEYS
from ..services.metadata_service import get_default_metadata_provider, get_metadata_provider
from ..services.metadata_sync_service import MetadataSyncService
from ..services.preview_asset_service import PreviewAssetService
from ..services.settings_manager import get_settings_manager
from ..services.downloader import get_downloader
from ..utils.constants import SUPPORTED_MEDIA_EXTENSIONS
from ..utils.exif_utils import ExifUtils
from ..utils.metadata_manager import MetadataManager
from ..utils.video_metadata import get_video_dimensions
logger = logging.getLogger(__name__)
# Placeholder dimensions written when the real ones cannot be determined.
# Kept for backwards compatibility with pre-existing metadata entries.
_DEFAULT_MEDIA_WIDTH = 720
_DEFAULT_MEDIA_HEIGHT = 1280
# Example metadata entries carry a marker: ``customImages`` use their ``id``
# while ``images`` use the positional index. Either way the marker must be a
# plain filename-safe token, never a path fragment.
_ENTRY_MARKER_PATTERN = re.compile(r"^(?:custom_|image_)?([^./\\]+)$")
_preview_service = PreviewAssetService(
metadata_manager=MetadataManager,
downloader_factory=get_downloader,
exif_utils=ExifUtils,
)
_metadata_sync_service: MetadataSyncService | None = None
_metadata_sync_service_settings: Optional["SettingsManager"] = None
if TYPE_CHECKING: # pragma: no cover - import for type checkers only
from ..services.settings_manager import SettingsManager
async def update_cache_from_metadata(
scanner: Any, file_path: str, metadata: Dict[str, Any]
) -> bool:
"""Update the scanner cache from a metadata dict using the in-place sync path.
``sync_cache_from_metadata`` patches the existing cache entry incrementally
(tag/hash/version indexes, targeted single-row SQL update) and only resorts
when a sort-key field changed. This avoids the ``O(n)`` full-list resort and
full cache rewrite that ``update_single_model_cache`` performs on every call,
which is critical for libraries with 100k+ models.
Falls back to the legacy full update when the scanner does not expose an
async ``sync_cache_from_metadata`` method.
Returns:
``True`` if the cache entry was updated, ``False`` otherwise.
"""
sync_method = getattr(scanner, "sync_cache_from_metadata", None)
if inspect.iscoroutinefunction(sync_method):
return await sync_method(file_path, metadata)
return await scanner.update_single_model_cache(file_path, file_path, metadata)
def _build_metadata_sync_service(settings_manager: "SettingsManager") -> MetadataSyncService:
"""Construct a metadata sync service bound to the provided settings."""
return MetadataSyncService(
metadata_manager=MetadataManager,
preview_service=_preview_service,
settings=settings_manager,
default_metadata_provider_factory=get_default_metadata_provider,
metadata_provider_selector=get_metadata_provider,
)
def _read_media_dimensions(path: str, is_video: bool) -> tuple[int, int]:
"""Return ``(width, height)`` for an example image or video file.
Videos are read from their container headers (PIL cannot open them) so the
showcase viewer sizes the gallery to the real aspect ratio. Falls back to
the legacy ``720x1280`` placeholder when the dimensions cannot be
determined — e.g. an unreadable file or an exotic codec — which only
affects the displayed aspect ratio, never the file itself.
"""
dimensions = None
if is_video:
dimensions = get_video_dimensions(path)
else:
try:
from PIL import Image
if os.path.exists(path):
with Image.open(path) as img:
dimensions = img.size
except Exception:
dimensions = None
if dimensions:
width, height = dimensions
if width > 0 and height > 0:
return int(width), int(height)
return _DEFAULT_MEDIA_WIDTH, _DEFAULT_MEDIA_HEIGHT
def _is_video_entry(file_path: Optional[str], entry: Mapping[str, Any]) -> bool:
"""Return True when an example entry points at a video file.
The local file extension wins over the recorded ``type`` because files in
the wild are frequently mislabelled (animated WebP saved as ``.mp4``);
``_read_media_dimensions`` handles that correctly either way.
"""
if file_path:
ext = os.path.splitext(file_path)[1].lower()
if ext in SUPPORTED_MEDIA_EXTENSIONS["videos"]:
return True
if ext in SUPPORTED_MEDIA_EXTENSIONS["images"]:
return False
return str(entry.get("type", "")).lower() == "video"
def _resolve_local_file(
entry: Mapping[str, Any],
index: int,
local_files: Mapping[str, str],
) -> Optional[str]:
"""Map a metadata entry onto its example file inside the model folder.
Reads the entry's own marker (``id`` for ``customImages``, positional
``index`` for ``images``) with an anchored regex, so the identifier can
never bleed into a neighbouring filename the way a prefix comparison can.
"""
marker = entry.get("id")
if not isinstance(marker, str) or not marker:
marker = str(index)
match = _ENTRY_MARKER_PATTERN.fullmatch(marker)
if not match:
return None
return local_files.get(match.group(1))
def repair_local_video_dimensions(
metadata: MutableMapping[str, Any],
local_files: Mapping[str, str],
*,
dry_run: bool = False,
) -> int:
"""Backfill real video dimensions for an entry that has local files.
Only entries with an empty ``url`` are considered: those have no remote
source, so the local file is the single source of truth for their size and
rewriting them cannot discard API-supplied data. Entries whose dimensions
already match the file are left byte-identical.
Args:
metadata: Raw metadata payload (mutated in place unless ``dry_run``).
local_files: ``{identifier: path}`` for files present in the model's
example folder, where the identifier is the entry's ``id`` (for
``customImages``) or its positional index (for ``images``).
dry_run: Count the fixes without mutating ``metadata``.
Returns:
The number of entries that were (or would be) repaired.
"""
civitai = metadata.get("civitai")
if not isinstance(civitai, dict):
return 0
repaired = 0
for key in ("customImages", "images"):
entries = civitai.get(key)
if not isinstance(entries, list) or not entries:
continue
for index, entry in enumerate(entries):
if not isinstance(entry, dict):
continue
if entry.get("url", "") != "":
# Remote-backed entry: never rebuilt from local state.
continue
file_path = _resolve_local_file(entry, index, local_files)
if not file_path or not os.path.isfile(file_path):
continue
dimensions = _read_media_dimensions(
file_path, _is_video_entry(file_path, entry)
)
width, height = dimensions
if width <= 0 or height <= 0:
continue
if entry.get("width") == width and entry.get("height") == height:
continue
if not dry_run:
entry["width"] = width
entry["height"] = height
repaired += 1
return repaired
def _get_metadata_sync_service() -> MetadataSyncService:
"""Return the shared metadata sync service, initialising it lazily."""
global _metadata_sync_service, _metadata_sync_service_settings
settings_manager = get_settings_manager()
if isinstance(_metadata_sync_service, MetadataSyncService):
if _metadata_sync_service_settings is not settings_manager:
_metadata_sync_service = _build_metadata_sync_service(settings_manager)
_metadata_sync_service_settings = settings_manager
elif _metadata_sync_service is None:
_metadata_sync_service = _build_metadata_sync_service(settings_manager)
_metadata_sync_service_settings = settings_manager
else:
# Tests may inject stand-ins that do not match the sync service type. Preserve
# those injections while still updating our cached settings reference so the
# next real service instantiation uses the current configuration.
_metadata_sync_service_settings = settings_manager
return _metadata_sync_service
class MetadataUpdater:
"""Handles updating model metadata related to example images"""
@staticmethod
async def refresh_model_metadata(model_hash, model_name, scanner_type, scanner, progress: dict[str, Any] | None = None):
"""Refresh model metadata from CivitAI
Args:
model_hash: SHA256 hash of the model
model_name: Model name (for logging)
scanner_type: Scanner type ('lora' or 'checkpoint')
scanner: Scanner instance for this model type
Returns:
bool: True if metadata was successfully refreshed, False otherwise
"""
try:
# Find the model in the scanner cache
cache = await scanner.get_cached_data()
model_data = None
for item in cache.raw_data:
if item.get('sha256') == model_hash:
model_data = item
break
if not model_data:
logger.warning(f"Model {model_name} with hash {model_hash} not found in cache")
return False
file_path = model_data.get('file_path')
if not file_path:
logger.warning(f"Model {model_name} has no file path")
return False
# Track that we're refreshing this model
if progress is not None:
progress['refreshed_models'].add(model_hash)
async def update_cache_func(old_path, new_path, metadata):
return await update_cache_from_metadata(scanner, new_path, metadata)
await MetadataManager.hydrate_model_data(model_data)
success, error = await _get_metadata_sync_service().fetch_and_update_model(
sha256=model_hash,
file_path=file_path,
model_data=model_data,
update_cache_func=update_cache_func,
)
if success:
logger.info(f"Successfully refreshed metadata for {model_name}")
return True
else:
logger.warning(f"Failed to refresh metadata for {model_name}, {error}")
return False
except Exception as e:
error_msg = f"Error refreshing metadata for {model_name}: {str(e)}"
logger.error(error_msg, exc_info=True)
if progress is not None:
progress['errors'].append(error_msg)
progress['last_error'] = error_msg
return False
@staticmethod
async def get_updated_model(model_hash, scanner):
"""Load the most recent metadata for a model identified by hash."""
cache = await scanner.get_cached_data()
target = None
for item in cache.raw_data:
if item.get('sha256') == model_hash:
target = item
break
if not target:
return None
file_path = target.get('file_path')
if not file_path:
return target
model_cls = getattr(scanner, 'model_class', None)
if model_cls is None:
metadata, should_skip = await MetadataManager.load_metadata(file_path)
else:
metadata, should_skip = await MetadataManager.load_metadata(file_path, model_cls)
if should_skip or metadata is None:
return target
rich_metadata = metadata.to_dict()
rich_metadata.setdefault('folder', target.get('folder', ''))
return rich_metadata
@staticmethod
async def update_metadata_from_local_examples(model_hash, model, scanner_type, scanner, model_dir):
"""Update model metadata with local example image information
Args:
model_hash: SHA256 hash of the model
model: Model data dictionary
scanner_type: Scanner type ('lora' or 'checkpoint')
scanner: Scanner instance for this model type
model_dir: Model images directory
Returns:
bool: True if metadata was successfully updated, False otherwise
"""
try:
# Collect local image paths
local_images_paths = []
if os.path.exists(model_dir):
for file in os.listdir(model_dir):
file_path = os.path.join(model_dir, file)
if os.path.isfile(file_path):
file_ext = os.path.splitext(file)[1].lower()
is_supported = (file_ext in SUPPORTED_MEDIA_EXTENSIONS['images'] or
file_ext in SUPPORTED_MEDIA_EXTENSIONS['videos'])
if is_supported:
local_images_paths.append(file_path)
await MetadataManager.hydrate_model_data(model)
civitai_data = model.setdefault('civitai', {})
# Check if metadata update is needed (no civitai field or empty images)
needs_update = not civitai_data or not civitai_data.get('images')
if needs_update and local_images_paths:
logger.debug(f"Found {len(local_images_paths)} local example images for {model.get('model_name')}, updating metadata")
# Create or get civitai field
# Create images array
images = []
# Generate metadata for each local image/video
for path in local_images_paths:
# Determine if video or image
file_ext = os.path.splitext(path)[1].lower()
is_video = file_ext in SUPPORTED_MEDIA_EXTENSIONS['videos']
width, height = _read_media_dimensions(path, is_video)
# Create image metadata entry
image_entry = {
"url": "", # Empty URL as required
"nsfwLevel": 0,
"width": width,
"height": height,
"type": "video" if is_video else "image",
"meta": None,
"hasMeta": False,
"hasPositivePrompt": False
}
images.append(image_entry)
# Update the model's civitai.images field
civitai_data['images'] = images
# Save metadata to .metadata.json file
file_path = model.get('file_path')
model_copy: Optional[Dict[str, Any]] = None
try:
model_copy = model.copy()
if model_copy is not None:
model_copy.pop('folder', None)
await MetadataManager.save_metadata(file_path, model_copy)
logger.info(f"Saved metadata for {model.get('model_name')}")
except Exception as e:
logger.error(f"Failed to save metadata for {model.get('model_name')}: {str(e)}")
# Save updated metadata to scanner cache. sync_cache_from_metadata
# returns False both for "already in sync" and for actual failures,
# so the cache sync result is deliberately not treated as an error;
# the return value reflects whether the metadata was persisted.
if file_path and model_copy is not None:
await update_cache_from_metadata(scanner, file_path, model_copy)
logger.info(f"Successfully updated metadata for {model.get('model_name')} with {len(images)} local examples")
return True
logger.warning(f"Failed to update metadata for {model.get('model_name')}")
return False
return False
except Exception as e:
logger.error(f"Error updating metadata from local examples: {str(e)}", exc_info=True)
return False
@staticmethod
async def update_metadata_after_import(model_hash, model_data, scanner, newly_imported_paths):
"""Update model metadata after importing example images
Args:
model_hash: SHA256 hash of the model
model_data: Model data dictionary
scanner: Scanner instance (lora or checkpoint)
newly_imported_paths: List of paths to newly imported files
Returns:
tuple: (regular_images, custom_images) - Both image arrays
"""
try:
await MetadataManager.hydrate_model_data(model_data)
civitai_data = model_data.get('civitai')
if not isinstance(civitai_data, dict):
civitai_data = {}
model_data['civitai'] = civitai_data
custom_images = civitai_data.get('customImages')
if not isinstance(custom_images, list):
custom_images = []
civitai_data['customImages'] = custom_images
# Add new image entry for each imported file
for path_tuple in newly_imported_paths:
path, short_id = path_tuple
# Determine if video or image
file_ext = os.path.splitext(path)[1].lower()
is_video = file_ext in SUPPORTED_MEDIA_EXTENSIONS['videos']
width, height = _read_media_dimensions(path, is_video)
# Create image metadata entry
image_entry = {
"url": "", # Empty URL as requested
"id": short_id,
"nsfwLevel": 0,
"width": width,
"height": height,
"type": "video" if is_video else "image",
"meta": None,
"hasMeta": False,
"hasPositivePrompt": False
}
# Extract and parse metadata if this is an image
if not is_video:
try:
# Extract metadata from image
extracted_metadata = ExifUtils.extract_image_metadata(path)
if extracted_metadata:
# Parse the extracted metadata to get generation parameters
parsed_meta = MetadataUpdater._parse_image_metadata(extracted_metadata)
if parsed_meta:
image_entry["meta"] = parsed_meta
image_entry["hasMeta"] = True
image_entry["hasPositivePrompt"] = bool(parsed_meta.get("prompt", ""))
logger.debug(f"Extracted metadata from {os.path.basename(path)}")
except Exception as e:
logger.warning(f"Failed to extract metadata from {os.path.basename(path)}: {e}")
# Append to existing customImages array
custom_images.append(image_entry)
# Save metadata to .metadata.json file
file_path = model_data.get('file_path')
model_copy: Optional[Dict[str, Any]] = None
if file_path:
try:
model_copy = model_data.copy()
if model_copy is not None:
model_copy.pop('folder', None)
await MetadataManager.save_metadata(file_path, model_copy)
logger.info(f"Saved metadata for {model_data.get('model_name')}")
except Exception as e:
logger.error(f"Failed to save metadata: {str(e)}")
# Save updated metadata to scanner cache
if file_path and model_copy is not None:
await update_cache_from_metadata(scanner, file_path, model_copy)
# Get regular images array (might be None)
regular_images = civitai_data.get('images', [])
# Return both image arrays
return regular_images, custom_images
except Exception as e:
logger.error(f"Failed to update metadata after import: {e}", exc_info=True)
return [], []
@staticmethod
def _parse_image_metadata(user_comment):
"""Parse metadata from image to extract generation parameters
Args:
user_comment: Metadata string extracted from image
Returns:
dict: Parsed metadata with generation parameters
"""
if not user_comment:
return None
try:
# Initialize metadata dictionary
metadata = {}
# Split on Negative prompt if it exists
if "Negative prompt:" in user_comment:
parts = user_comment.split('Negative prompt:', 1)
prompt = parts[0].strip()
negative_and_params = parts[1] if len(parts) > 1 else ""
else:
# No negative prompt section
param_start = re.search(r'Steps: \d+', user_comment)
if param_start:
prompt = user_comment[:param_start.start()].strip()
negative_and_params = user_comment[param_start.start():]
else:
prompt = user_comment.strip()
negative_and_params = ""
# Add prompt if it's in GEN_PARAM_KEYS
if 'prompt' in GEN_PARAM_KEYS:
metadata['prompt'] = prompt
# Extract negative prompt and parameters
if negative_and_params:
# If we split on "Negative prompt:", check for params section
if "Negative prompt:" in user_comment:
param_start = re.search(r'Steps: ', negative_and_params)
if param_start:
neg_prompt = negative_and_params[:param_start.start()].strip()
if 'negative_prompt' in GEN_PARAM_KEYS:
metadata['negative_prompt'] = neg_prompt
params_section = negative_and_params[param_start.start():]
else:
if 'negative_prompt' in GEN_PARAM_KEYS:
metadata['negative_prompt'] = negative_and_params.strip()
params_section = ""
else:
# No negative prompt, entire section is params
params_section = negative_and_params
# Extract generation parameters
if params_section:
# Extract basic parameters
param_pattern = r'([A-Za-z\s]+): ([^,]+)'
params = re.findall(param_pattern, params_section)
for key, value in params:
clean_key = key.strip().lower().replace(' ', '_')
# Skip if not in recognized gen param keys
if clean_key not in GEN_PARAM_KEYS:
continue
# Convert numeric values
if clean_key in ['steps', 'seed']:
try:
metadata[clean_key] = int(value.strip())
except ValueError:
metadata[clean_key] = value.strip()
elif clean_key in ['cfg_scale']:
try:
metadata[clean_key] = float(value.strip())
except ValueError:
metadata[clean_key] = value.strip()
else:
metadata[clean_key] = value.strip()
# Extract size if available and add if a recognized key
size_match = re.search(r'Size: (\d+)x(\d+)', params_section)
if size_match and 'size' in GEN_PARAM_KEYS:
width, height = size_match.groups()
metadata['size'] = f"{width}x{height}"
# Return metadata if we have any entries
return metadata if metadata else None
except Exception as e:
logger.error(f"Error parsing image metadata: {e}", exc_info=True)
return None
@staticmethod
async def prune_stale_example_images(metadata) -> bool:
"""Remove example-image metadata entries whose files no longer exist on disk.
Checks ``civitai.customImages`` (by ``id``) and ``civitai.images`` entries
that have an empty ``url`` (no remote fallback) against actual files in
the model's example-image folder. Stale entries are removed in-place so
the caller can persist the cleaned metadata afterwards.
Args:
metadata: A ``BaseModelMetadata`` instance (modified in place).
Returns:
True if at least one entry was removed.
"""
from ..utils.example_images_paths import get_model_folder
model_hash = getattr(metadata, "sha256", None)
if not model_hash:
return False
model_folder = get_model_folder(model_hash)
if not model_folder or not os.path.isdir(model_folder):
return False
civitai = getattr(metadata, "civitai", None)
if not isinstance(civitai, dict):
return False
# Read the directory listing once so every image entry reuses it.
try:
dir_entries = os.listdir(model_folder)
except OSError:
dir_entries = []
has_changes = False
custom_images = civitai.get("customImages")
if isinstance(custom_images, list) and custom_images:
stale: list[int] = []
for idx, img in enumerate(custom_images):
img_id = img.get("id", "")
if not img_id:
continue
prefix = f"custom_{img_id}"
found = any(
f.startswith(prefix) and os.path.isfile(
os.path.join(model_folder, f)
)
for f in dir_entries
)
if not found:
stale.append(idx)
if stale:
for idx in reversed(stale):
custom_images.pop(idx)
has_changes = True
logger.info(
"Pruned %d stale custom image(s) for %s",
len(stale),
getattr(metadata, "model_name", model_hash),
)
images = civitai.get("images")
if isinstance(images, list) and images:
stale_images: list[int] = []
for idx, img in enumerate(images):
if img.get("url", ""):
# Has a remote fallback keep it even if the local copy
# is gone.
continue
prefix = f"image_{idx}."
if not any(f.startswith(prefix) for f in dir_entries):
stale_images.append(idx)
if stale_images:
for idx in reversed(stale_images):
images.pop(idx)
has_changes = True
logger.info(
"Pruned %d stale image entry(ies) for %s",
len(stale_images),
getattr(metadata, "model_name", model_hash),
)
return has_changes