289 lines
14 KiB
Python
289 lines
14 KiB
Python
# /ComfyUI-Prompt-Formatter/categorized_prompt_analyzer.py
|
|
|
|
import yaml
|
|
import re
|
|
import os
|
|
import sys
|
|
from collections import defaultdict
|
|
from pathlib import Path
|
|
|
|
|
|
# --- Dependency Check ---
|
|
try:
|
|
import yaml
|
|
except ImportError:
|
|
print("PyYAML is not installed. Please run install.py or install it manually.", file=sys.stderr)
|
|
|
|
# --- Constants ---
|
|
NODE_NAME_ANALYZER = "Categorized Prompt Analyzer" # Specific name
|
|
INCLUDE_DIRECTIVE = "$include"
|
|
TAGS_KEY = "tags"
|
|
|
|
|
|
def get_node_directory_analyzer():
|
|
return Path(__file__).parent
|
|
|
|
def find_yaml_file_analyzer(filename):
|
|
if not filename: return None
|
|
if Path(filename).is_absolute():
|
|
if Path(filename).is_file(): return Path(filename)
|
|
else: print(f"Warning [Analyzer]: Absolute path specified but not found: {filename}"); return None
|
|
node_dir = get_node_directory_analyzer()
|
|
relative_to_node = node_dir / filename
|
|
if relative_to_node.is_file(): return relative_to_node
|
|
try:
|
|
current_dir = Path.cwd()
|
|
input_dir = current_dir / 'input'
|
|
if input_dir.exists():
|
|
input_dir_path = input_dir / filename
|
|
if input_dir_path.is_file(): return input_dir_path
|
|
search_dir = node_dir
|
|
for _ in range(5):
|
|
input_dir = search_dir / 'input'
|
|
if input_dir.exists():
|
|
input_dir_path = input_dir / filename
|
|
if input_dir_path.is_file(): return input_dir_path
|
|
if (search_dir / 'ComfyUI').exists() or (search_dir / 'main.py').exists():
|
|
input_dir = search_dir / 'input'
|
|
if input_dir.exists():
|
|
input_dir_path = input_dir / filename
|
|
if input_dir_path.is_file(): return input_dir_path
|
|
break
|
|
parent = search_dir.parent
|
|
if parent == search_dir: break
|
|
search_dir = parent
|
|
except Exception as e: print(f"Warning [Analyzer]: Error searching for ComfyUI input directory: {e}")
|
|
print(f"Warning [Analyzer]: YAML file '{filename}' not found as absolute, relative to node, or in input dir.")
|
|
return None
|
|
|
|
def resolve_category_tags_analyzer(category_name, yaml_data, resolved_cache, recursion_guard=None):
|
|
category_name = str(category_name).strip()
|
|
if recursion_guard is None: recursion_guard = set()
|
|
if len(recursion_guard) > 20: return set() # Depth limit
|
|
if category_name in recursion_guard: print(f"Warning [Analyzer]: Circular dependency: '{category_name}'."); return set()
|
|
if category_name in resolved_cache: return resolved_cache[category_name]
|
|
if category_name not in yaml_data: return set()
|
|
recursion_guard.add(category_name)
|
|
category_data = yaml_data[category_name]
|
|
final_tags = set()
|
|
if isinstance(category_data, list):
|
|
for item in category_data:
|
|
item_str = str(item).strip();
|
|
if not item_str: continue
|
|
is_full_include = item_str.startswith('$') and len(item_str) > 1 and ' ' not in item_str and item_str[1:].replace('_', '').isalnum()
|
|
if is_full_include:
|
|
included_category = item_str[1:]
|
|
final_tags.update(resolve_category_tags_analyzer(included_category, yaml_data, resolved_cache, recursion_guard.copy()))
|
|
else:
|
|
match = re.search(r'\$(\w+)', item_str)
|
|
if match:
|
|
ref_category_name = match.group(1); placeholder = match.group(0)
|
|
resolved_ref_tags = resolve_category_tags_analyzer(ref_category_name, yaml_data, resolved_cache, recursion_guard.copy())
|
|
if not resolved_ref_tags: pass # Optionally add literal item_str here
|
|
else:
|
|
for resolved_tag in resolved_ref_tags:
|
|
new_tag = item_str.replace(placeholder, str(resolved_tag).strip(), 1)
|
|
final_tags.add(new_tag)
|
|
else: final_tags.add(item_str)
|
|
elif isinstance(category_data, dict):
|
|
if INCLUDE_DIRECTIVE in category_data:
|
|
includes = category_data[INCLUDE_DIRECTIVE]; include_list = []
|
|
if isinstance(includes, list): include_list = [str(inc).strip() for inc in includes if str(inc).strip()]
|
|
elif isinstance(includes, str): include_list = [includes.strip()] if includes.strip() else []
|
|
for included_category in include_list:
|
|
final_tags.update(resolve_category_tags_analyzer(included_category, yaml_data, resolved_cache, recursion_guard.copy()))
|
|
if TAGS_KEY in category_data and isinstance(category_data[TAGS_KEY], list):
|
|
final_tags.update(str(tag).strip() for tag in category_data[TAGS_KEY] if str(tag).strip())
|
|
resolved_cache[category_name] = final_tags
|
|
return final_tags
|
|
|
|
def parse_tag_analyzer(tag, handle_weights=True):
|
|
original_tag = str(tag).strip(); base_tag = original_tag
|
|
if not handle_weights: return original_tag, base_tag
|
|
patterns = [ r"^\s*\(\s*(.*?)\s*(?::\s*[\d.]+\s*)?\)\s*$", r"^\s*\[\s*(.*?)\s*\]\s*$", r"^\s*\{\s*(.*?)\s*\}\s*$" ]
|
|
for pattern in patterns:
|
|
match = re.match(pattern, original_tag)
|
|
if match:
|
|
base_tag = match.group(1).strip()
|
|
nested_match = re.match(r"^\s*[\(\[\{](.*?)[\)\]\}]\s*$", base_tag)
|
|
if nested_match: base_tag = nested_match.group(1).strip()
|
|
return original_tag, base_tag
|
|
return original_tag, base_tag.strip()
|
|
|
|
def clean_output_string_analyzer(text, delimiter=", "):
|
|
if not text: return ""
|
|
text = text.strip().strip(delimiter.strip()).strip()
|
|
delimiter_pattern = r'\s*' + re.escape(delimiter.strip()) + r'\s*'
|
|
text = re.sub(f'({delimiter_pattern})+', delimiter, text)
|
|
if text.startswith(delimiter): text = text[len(delimiter):].lstrip()
|
|
if text.endswith(delimiter.rstrip()): text = text[:-len(delimiter.rstrip())].rstrip()
|
|
return text
|
|
|
|
# --- The Analyzer Node Class ---
|
|
|
|
class CategorizedPromptAnalyzer:
|
|
"""
|
|
A ComfyUI node to analyze tag occurrences in a prompt based on specific
|
|
tags or categories defined in a YAML file.
|
|
"""
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"prompt": ("STRING", {"multiline": True, "default": ""}),
|
|
"analyze_targets": ("STRING", {"multiline": False, "default": "quality, eyes, 1girl"}),
|
|
"category_definition_file": ("STRING", {"default": "prompt_categories.yaml"}),
|
|
},
|
|
"optional": {
|
|
"input_delimiter": ("STRING", {"default": ","}),
|
|
"target_delimiter": ("STRING", {"default": ","}),
|
|
"output_delimiter": ("STRING", {"default": ", "}), # For list outputs
|
|
"strip_whitespace": ("BOOLEAN", {"default": True}),
|
|
"case_sensitive_matching": ("BOOLEAN", {"default": False}),
|
|
"handle_weights": ("BOOLEAN", {"default": True}),
|
|
"match_underscores_spaces": ("BOOLEAN", {"default": True}),
|
|
"generate_detailed_output": ("BOOLEAN", {"default": False}),
|
|
"generate_unmatched_output": ("BOOLEAN", {"default": False}),
|
|
}
|
|
}
|
|
|
|
RETURN_TYPES = ("STRING", "INT", "STRING", "STRING")
|
|
RETURN_NAMES = ("analysis_summary", "total_matched_count", "details", "unmatched_tags")
|
|
FUNCTION = "analyze_prompt"
|
|
CATEGORY = "text/analysis"
|
|
|
|
def analyze_prompt(self, prompt, analyze_targets, category_definition_file,
|
|
input_delimiter=",", target_delimiter=",", output_delimiter=", ",
|
|
strip_whitespace=True, case_sensitive_matching=False,
|
|
handle_weights=True, match_underscores_spaces=True,
|
|
generate_detailed_output=False, generate_unmatched_output=False):
|
|
|
|
# --- 1. Load & Resolve YAML ---
|
|
resolved_category_tags = {}
|
|
raw_yaml_data = None
|
|
yaml_path = find_yaml_file_analyzer(category_definition_file)
|
|
|
|
if yaml_path and yaml_path.exists():
|
|
try:
|
|
with open(yaml_path, 'r', encoding='utf-8') as f: raw_yaml_data = yaml.safe_load(f)
|
|
if not isinstance(raw_yaml_data, dict): raw_yaml_data = None
|
|
except Exception as e: print(f"Error [Analyzer]: Loading YAML {yaml_path}: {e}")
|
|
|
|
if raw_yaml_data:
|
|
resolved_tags_cache = {}
|
|
all_category_names = list(raw_yaml_data.keys())
|
|
for category_name in all_category_names:
|
|
if str(category_name).strip() != INCLUDE_DIRECTIVE and str(category_name).strip() != TAGS_KEY:
|
|
# Resolve tags using the helper
|
|
resolved_tags_set = resolve_category_tags_analyzer(category_name, raw_yaml_data, resolved_tags_cache)
|
|
# Store resolved tags, applying case sensitivity NOW for matching later
|
|
normalized_category_name = str(category_name).strip() if case_sensitive_matching else str(category_name).strip().lower()
|
|
normalized_tags = set()
|
|
for tag in resolved_tags_set:
|
|
# Also normalize tags based on case sensitivity for matching
|
|
normalized_tags.add(tag if case_sensitive_matching else tag.lower())
|
|
resolved_category_tags[normalized_category_name] = normalized_tags
|
|
# else: Warning already printed by find_yaml_file
|
|
|
|
# --- 2. Parse & Process Targets ---
|
|
targets_info = [] # List of tuples: (original_target, normalized_target, is_category, {normalized_tags_to_match})
|
|
original_targets_list = [t.strip() for t in analyze_targets.split(target_delimiter) if t.strip()]
|
|
|
|
for original_target in original_targets_list:
|
|
normalized_target = original_target if case_sensitive_matching else original_target.lower()
|
|
is_category = False
|
|
normalized_tags_to_match = set()
|
|
|
|
if normalized_target in resolved_category_tags:
|
|
# It's a category target
|
|
is_category = True
|
|
normalized_tags_to_match = resolved_category_tags[normalized_target]
|
|
else:
|
|
# It's a literal target
|
|
is_category = False
|
|
# Generate variants for the literal target itself for matching
|
|
literal_variants = {normalized_target}
|
|
if match_underscores_spaces:
|
|
literal_variants.add(normalized_target.replace('_', ' '))
|
|
literal_variants.add(normalized_target.replace(' ', '_'))
|
|
normalized_tags_to_match = literal_variants
|
|
|
|
targets_info.append((original_target, normalized_target, is_category, normalized_tags_to_match))
|
|
|
|
# --- 3. Parse Input Prompt ---
|
|
input_tags_processed = [] # List of tuples: (original_tag, base_tag_normalized, {variants})
|
|
all_input_tags_original = set()
|
|
|
|
if prompt:
|
|
raw_tags = prompt.split(input_delimiter)
|
|
for raw_tag in raw_tags:
|
|
tag_original = raw_tag.strip() if strip_whitespace else raw_tag
|
|
if not tag_original: continue
|
|
all_input_tags_original.add(tag_original)
|
|
|
|
# Parse tag, get base form
|
|
_, base_parsed = parse_tag_analyzer(tag_original, handle_weights)
|
|
base_tag_normalized = base_parsed if case_sensitive_matching else base_parsed.lower()
|
|
|
|
# Generate variants for matching
|
|
variants = {base_tag_normalized}
|
|
if match_underscores_spaces:
|
|
variants.add(base_tag_normalized.replace('_', ' '))
|
|
variants.add(base_tag_normalized.replace(' ', '_'))
|
|
|
|
input_tags_processed.append((tag_original, base_tag_normalized, variants))
|
|
|
|
|
|
# --- 4. Perform Counting ---
|
|
counts = defaultdict(int) # original_target -> count
|
|
details_data = defaultdict(list) # original_target -> list_of_matched_input_tags
|
|
matched_input_originals_set = set()
|
|
|
|
for tag_original, _, input_variants in input_tags_processed:
|
|
tag_matched_at_least_once = False
|
|
for original_target, normalized_target, is_category, target_match_set in targets_info:
|
|
# Check if any input variant matches any tag in the target's match set
|
|
found_match = False
|
|
for variant in input_variants:
|
|
if variant in target_match_set:
|
|
found_match = True
|
|
break # Found a match for this input tag against this target
|
|
|
|
if found_match:
|
|
counts[original_target] += 1
|
|
if generate_detailed_output:
|
|
details_data[original_target].append(tag_original)
|
|
matched_input_originals_set.add(tag_original)
|
|
tag_matched_at_least_once = True
|
|
# Note: If an input tag matches multiple targets (e.g., literal and category),
|
|
# it will correctly increment the count for *each* target it matches.
|
|
|
|
|
|
# --- 5. Calculate Unmatched ---
|
|
unmatched_tags_list = []
|
|
if generate_unmatched_output:
|
|
unmatched_tags_list = list(all_input_tags_original - matched_input_originals_set)
|
|
|
|
|
|
# --- 6. Format Outputs ---
|
|
summary_parts = []
|
|
for original_target in original_targets_list: # Keep original order
|
|
summary_parts.append(f"{original_target}: {counts[original_target]}")
|
|
analysis_summary = output_delimiter.join(summary_parts)
|
|
|
|
total_matched_count = sum(counts.values())
|
|
|
|
details_str = ""
|
|
if generate_detailed_output:
|
|
detail_parts = []
|
|
for original_target in original_targets_list:
|
|
matched_list_str = output_delimiter.join(details_data[original_target])
|
|
detail_parts.append(f"{original_target}: {counts[original_target]} [{matched_list_str}]")
|
|
details_str = "\n".join(detail_parts) # Use newline for readability
|
|
|
|
unmatched_tags_str = ""
|
|
if generate_unmatched_output:
|
|
unmatched_tags_str = output_delimiter.join(unmatched_tags_list)
|
|
|
|
|
|
return (analysis_summary, total_matched_count, details_str, unmatched_tags_str) |