PaperShow / Paper2Poster /utils /logo_utils.py
ZaynZhu
Clean version without large assets
7c08dc3
"""
Logo management utilities for Paper2Poster project.
Handles searching, downloading, and retrieving logos for conferences and institutions.
Uses file-based matching - just drop PNG files in conferences/ or institutes/ folders.
"""
import os
import requests
import logging
from typing import Optional, Dict, List, Tuple
from pathlib import Path
from PIL import Image
from io import BytesIO
import re
from difflib import SequenceMatcher
from datetime import datetime
# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
class LogoManager:
"""Manages logo storage and retrieval using file-based matching."""
def __init__(self, base_path: str = "logo_store"):
"""
Initialize the LogoManager.
Args:
base_path: Base directory for logo storage
"""
self.base_path = Path(base_path)
self._setup_directories()
def _setup_directories(self):
"""Create necessary directories for logo storage."""
directories = [
self.base_path,
self.base_path / "conferences",
self.base_path / "institutes",
self.base_path / "raw_downloads"
]
for directory in directories:
directory.mkdir(parents=True, exist_ok=True)
def _normalize_name(self, name: str) -> str:
"""Normalize a name for matching."""
# Remove common suffixes like years
name = re.sub(r'\s*\d{4}\s*$', '', name)
# Convert to lowercase and replace spaces/special chars
name = name.lower()
name = re.sub(r'[^a-z0-9]+', '_', name)
name = name.strip('_')
return name
def _fuzzy_match(self, query: str, candidates: List[str]) -> Tuple[Optional[str], float]:
"""
Find the best fuzzy match for a query among candidates.
Args:
query: The search query
candidates: List of candidate strings
Returns:
Best matching candidate and similarity score (0-1)
"""
query_norm = self._normalize_name(query)
best_match = None
best_score = 0.0
for candidate in candidates:
# Check exact match first
if query_norm == candidate:
return candidate, 1.0
# Check if query is contained in candidate or vice versa
if query_norm in candidate or candidate in query_norm:
score = 0.9
if score > best_score:
best_match = candidate
best_score = score
continue
# Use sequence matching for fuzzy comparison
score = SequenceMatcher(None, query_norm, candidate).ratio()
if score > best_score:
best_match = candidate
best_score = score
# Return match only if score is high enough
if best_score >= 0.6: # 60% similarity threshold
return best_match, best_score
return None, 0.0
def _scan_directory(self, directory: Path) -> Dict[str, Path]:
"""
Scan a directory for PNG files.
Returns:
Dictionary mapping normalized names to file paths
"""
logos = {}
if directory.exists():
for file in directory.glob("*.png"):
# Use the filename stem as the key
name = file.stem.lower()
logos[name] = file
return logos
def get_logo_path(self, name: str, category: str = "auto", use_google: bool = False) -> Optional[Path]:
"""
Get the path to a logo file using fuzzy matching.
Args:
name: Name of the conference/institution
category: Type of logo ("conference", "institute", or "auto")
use_google: Whether to use Google Custom Search for web search
Returns:
Path to the logo file or None if not found
"""
print(f"\n 🔍 Looking for logo: '{name}' (category: {category})")
# Scan available logos
conference_logos = self._scan_directory(self.base_path / "conferences")
institute_logos = self._scan_directory(self.base_path / "institutes")
# Determine which directories to search
if category == "conference":
search_dirs = [("conferences", conference_logos)]
print(f" 📂 Searching in: conferences/ ({len(conference_logos)} logos)")
elif category == "institute":
search_dirs = [("institutes", institute_logos)]
print(f" 📂 Searching in: institutes/ ({len(institute_logos)} logos)")
else: # auto
search_dirs = [("conferences", conference_logos), ("institutes", institute_logos)]
print(f" 📂 Searching in: conferences/ ({len(conference_logos)} logos), institutes/ ({len(institute_logos)} logos)")
# Try to find best match
best_match = None
best_score = 0.0
best_path = None
best_dir = None
for dir_name, logos in search_dirs:
if logos:
match, score = self._fuzzy_match(name, list(logos.keys()))
if match and score > best_score:
best_match = match
best_score = score
best_path = logos[match]
best_dir = dir_name
if best_path and best_path.exists():
print(f" ✅ MATCH FOUND: '{best_match}' in {best_dir}/ (similarity: {best_score:.1%})")
print(f" 📄 File: {best_path.name}")
return best_path
# If no match found, try to download
print(f" ❌ No local match found (threshold: 60%)")
print(f" 🌐 Attempting to download from web...")
return self._download_and_save_logo(name, category, use_google=use_google)
def _download_and_save_logo(self, name: str, category: str, use_google: bool = False) -> Optional[Path]:
"""
Try to download a logo from the web and save it.
Args:
name: Name to search for
category: Category for saving
use_google: Whether to use Google Custom Search
Returns:
Path to downloaded logo or None
"""
search_query = f"{name} logo"
print(f" 🔎 Web search query: '{search_query}'")
if use_google:
print(f" 🌐 Using Google Custom Search API")
url = self.search_logo_web(search_query, use_google=use_google)
if not url:
print(f" ❌ No logo found online for: {name}")
return None
print(f" 🌐 Found URL: {url[:80]}...")
# Determine save directory
if category == "conference":
save_dir = self.base_path / "conferences"
else:
save_dir = self.base_path / "institutes"
# Generate filename
filename = self._normalize_name(name) + ".png"
save_path = save_dir / filename
print(f" 💾 Downloading to: {save_path}")
if self.download_logo(url, save_path):
print(f" ✅ Successfully downloaded and saved: {filename}")
return save_path
else:
print(f" ❌ Failed to download/convert logo")
return None
def search_logo_web(self, query: str, use_google: bool = False) -> Optional[str]:
"""
Search for a logo on the web using DuckDuckGo or Google.
Args:
query: Search query
use_google: Whether to use Google Custom Search (requires API key)
Returns:
URL of the found logo image or None
"""
# Try DuckDuckGo first (no API key required)
try:
from duckduckgo_search import DDGS
with DDGS() as ddgs:
# Search for images
results = ddgs.images(
f"{query} official transparent PNG SVG",
max_results=5
)
# Filter for likely logo images
for result in results:
url = result.get('image')
if url and any(ext in url.lower() for ext in ['.png', '.svg', '.jpg', '.jpeg']):
logger.info(f"Found potential logo: {url}")
return url
except Exception as e:
logger.warning(f"DuckDuckGo search failed: {e}")
# Try Google Custom Search if enabled and API key is available
if use_google:
try:
google_api_key = os.getenv('GOOGLE_SEARCH_API_KEY')
google_engine_id = os.getenv('GOOGLE_SEARCH_ENGINE_ID')
if google_api_key and google_engine_id:
url = "https://www.googleapis.com/customsearch/v1"
params = {
'key': google_api_key,
'cx': google_engine_id,
'q': f"{query} official logo transparent PNG",
'searchType': 'image',
'num': 5,
'fileType': 'png|svg'
}
response = requests.get(url, params=params)
if response.status_code == 200:
data = response.json()
items = data.get('items', [])
if items:
return items[0].get('link')
else:
logger.warning("Google API keys not found in environment")
except Exception as e:
logger.warning(f"Google search failed: {e}")
return None
def download_logo(self, url: str, save_path: Path) -> bool:
"""
Download a logo from a URL.
Args:
url: URL of the logo
save_path: Path where to save the logo
Returns:
True if successful, False otherwise
"""
try:
headers = {
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
}
response = requests.get(url, headers=headers, timeout=10)
response.raise_for_status()
# Save the file
save_path.parent.mkdir(parents=True, exist_ok=True)
# If it's an SVG, try to convert to PNG
if url.lower().endswith('.svg'):
try:
import cairosvg
# Convert SVG to PNG
png_bytes = cairosvg.svg2png(bytestring=response.content, output_width=800)
img = Image.open(BytesIO(png_bytes))
if img.mode != 'RGBA':
img = img.convert('RGBA')
img.save(save_path, 'PNG', optimize=True)
logger.info(f"Converted SVG to PNG and saved to {save_path}")
return True
except Exception as e:
logger.warning(f"Could not convert SVG: {e}")
return False
# If it's another image format, convert to PNG
elif any(ext in url.lower() for ext in ['.jpg', '.jpeg', '.gif', '.bmp', '.png']):
try:
img = Image.open(BytesIO(response.content))
# Convert to RGBA for transparency support
if img.mode != 'RGBA':
img = img.convert('RGBA')
img.save(save_path, 'PNG')
logger.info(f"Downloaded and saved logo to {save_path}")
return True
except Exception as e:
logger.warning(f"Could not process image: {e}")
return False
else:
logger.warning(f"Unsupported file format: {url}")
return False
except Exception as e:
logger.error(f"Failed to download logo from {url}: {e}")
return False
def list_available_logos(self) -> Dict[str, List[str]]:
"""List all available logos in the system."""
conference_logos = self._scan_directory(self.base_path / "conferences")
institute_logos = self._scan_directory(self.base_path / "institutes")
return {
"conferences": sorted(conference_logos.keys()),
"institutes": sorted(institute_logos.keys())
}
def extract_first_author_institution(self, paper_content: str) -> Optional[str]:
"""
Extract the first author's institution from paper content.
Args:
paper_content: Text content of the paper (markdown format)
Returns:
First author's institution if found and matched with available logos
"""
print(" 📝 Looking for first author's institution...")
# Look for authors section in the beginning of the paper
lines = paper_content.split('\n')[:100] # Focus on first 100 lines where authors usually appear
# Common institution patterns
institution_patterns = [
r"(?:University of|University) [\w\s]+",
r"[\w\s]+ University",
r"[\w\s]+ Institute of Technology",
r"[\w\s]+ Institute",
r"MIT|CMU|UCLA|UCSD|NYU|ETH|EPFL|Stanford|Berkeley|Harvard|Princeton|Oxford|Cambridge",
r"Google Research|DeepMind|Microsoft Research|Facebook AI Research|OpenAI|NVIDIA Research",
r"Max Planck Institute",
r"[\w\s]+ College",
r"[\w\s]+ Research",
r"[\w\s]+ Lab",
r"[\w\s]+ Laboratory"
]
all_pattern = '|'.join(f'({p})' for p in institution_patterns)
# First pass: Look for the first line with superscript 1 (¹) which typically indicates first author affiliation
first_institution = None
for i, line in enumerate(lines):
# Stop at abstract or introduction
if 'abstract' in line.lower() or 'introduction' in line.lower():
break
# Look for lines with ¹ (first affiliation marker) at the beginning
if '¹' in line:
# Check if this line contains an institution
matches = re.findall(all_pattern, line, re.IGNORECASE)
if matches:
for match_groups in matches:
for inst in match_groups:
if inst:
first_institution = inst.strip()
print(f" 🎯 Found first author institution (from affiliation marker): {first_institution}")
break
if first_institution:
break
if first_institution:
break
# Second pass: If no superscript found, look for institution after the first author name
if not first_institution:
# Look for patterns like "Author Name (Institution)" or similar
for i, line in enumerate(lines):
if 'abstract' in line.lower() or 'introduction' in line.lower():
break
# Skip title lines (usually the first few lines)
if i < 2:
continue
# Look for lines that might contain author + institution
if '(' in line and ')' in line:
# Extract content in parentheses
paren_content = re.findall(r'\((.*?)\)', line)
for content in paren_content:
inst_matches = re.findall(all_pattern, content, re.IGNORECASE)
if inst_matches:
for match_groups in inst_matches:
for inst in match_groups:
if inst:
first_institution = inst.strip()
print(f" 🎯 Found first author institution (from parentheses): {first_institution}")
break
if first_institution:
break
if first_institution:
break
if first_institution:
break
# Third pass: If still nothing, just find the first institution mentioned
if not first_institution:
for line in lines[:30]: # Only check first 30 lines for general search
if 'abstract' in line.lower() or 'introduction' in line.lower():
break
matches = re.findall(all_pattern, line, re.IGNORECASE)
if matches:
for match_groups in matches:
for inst in match_groups:
if inst:
first_institution = inst.strip()
print(f" 🎯 Found institution (general search): {first_institution}")
break
if first_institution:
break
if first_institution:
break
if not first_institution:
print(" ❌ No institution found in author section")
return None
# Return the institution name regardless of whether there's a local match
# The get_logo_path() method will handle downloading if not found locally
print(f" ✅ Extracted institution: '{first_institution}'")
return first_institution
def extract_institution_from_paper(self, paper_content: str) -> List[str]:
"""
Extract institution names from paper content.
DEPRECATED: Use extract_first_author_institution() instead for better accuracy.
Args:
paper_content: Text content of the paper
Returns:
List of detected institution names that match available logos
"""
institutions = []
# Common patterns for affiliations
patterns = [
r"(?:University of|University) [\w\s]+",
r"[\w\s]+ University",
r"[\w\s]+ Institute of Technology",
r"[\w\s]+ Institute",
r"MIT|CMU|UCLA|UCSD|NYU|ETH|EPFL|Stanford|Berkeley|Harvard|Princeton",
r"Google Research|DeepMind|Microsoft Research|Facebook AI Research|OpenAI|NVIDIA Research",
r"Max Planck Institute",
r"[\w\s]+ College"
]
# Extract potential institution names
potential_institutions = []
print(" 📝 Searching for institution patterns...")
for pattern in patterns:
matches = re.findall(pattern, paper_content, re.IGNORECASE)
if matches:
potential_institutions.extend(matches)
# Normalize and deduplicate
potential_institutions = list(set([inst.strip() for inst in potential_institutions]))
if potential_institutions:
print(f" 🔎 Found {len(potential_institutions)} potential institutions in paper")
# Show first 5 for brevity
for i, inst in enumerate(potential_institutions[:5], 1):
print(f" {i}. {inst}")
if len(potential_institutions) > 5:
print(f" ... and {len(potential_institutions) - 5} more")
# Get available logos
available_logos = self.list_available_logos()
all_available = available_logos["institutes"]
print(f" 🔍 Matching against {len(all_available)} available institute logos...")
# Match against available logos using fuzzy matching
matched_count = 0
for inst in potential_institutions:
match, score = self._fuzzy_match(inst, all_available)
if match and score >= 0.7: # Higher threshold for institution matching
institutions.append(inst)
matched_count += 1
print(f" ✅ MATCH: '{inst}' → '{match}' (similarity: {score:.1%})")
if matched_count == 0 and potential_institutions:
print(f" ❌ No matches found (threshold: 70%)")
return institutions
def main():
"""Example usage of LogoManager."""
# Initialize manager
manager = LogoManager()
# List available logos
available = manager.list_available_logos()
print("Available logos:")
for category, items in available.items():
if items:
print(f"\n{category}: {', '.join(items)}")
# Example: Get a conference logo with fuzzy matching
test_names = ["NeurIPS", "neurips 2024", "NIPS", "neural information"]
for name in test_names:
logo_path = manager.get_logo_path(name, "conference")
if logo_path:
print(f"\nLogo for '{name}' -> {logo_path}")
# Example: Test institute matching
test_institutes = ["MIT", "Massachusetts Institute of Technology", "Stanford University", "stanford"]
for inst in test_institutes:
logo_path = manager.get_logo_path(inst, "institute")
if logo_path:
print(f"\nLogo for '{inst}' -> {logo_path}")
def get_logo_dimensions(logo_path: str, target_height: float) -> Tuple[float, float]:
"""
Calculate logo width to preserve aspect ratio.
Args:
logo_path: Path to logo image
target_height: Desired height in inches
Returns:
Tuple of (width, height) in inches
"""
try:
with Image.open(logo_path) as img:
aspect_ratio = img.width / img.height
target_width = target_height * aspect_ratio
return target_width, target_height
except Exception:
# Fallback to square if can't read image
return target_height, target_height
def add_logos_to_poster_code(
poster_code: str,
width_inch: float,
height_inch: float,
institution_logo_path: Optional[str] = None,
conference_logo_path: Optional[str] = None,
logo_height: float = 2.0,
logo_margin: float = 0.5
) -> str:
"""
Add institution and conference logos to poster code.
Args:
poster_code: Existing poster generation code
width_inch: Width of poster in inches
height_inch: Height of poster in inches
institution_logo_path: Path to institution logo (top-left)
conference_logo_path: Path to conference logo (top-right)
logo_height: Height of logos in inches (default: 2.0)
logo_margin: Margin from edges in inches (default: 0.5)
Returns:
Modified poster code with logos added
"""
import re
logo_code = ""
# Add institution logo to top-left
if institution_logo_path and os.path.exists(institution_logo_path):
inst_width, inst_height = get_logo_dimensions(institution_logo_path, logo_height)
logo_code += f'''
# Add institution logo to top-left (aspect ratio preserved)
institution_logo = add_image(
poster_slide,
'institution_logo',
{logo_margin}, # left
{logo_margin}, # top
{inst_width}, # width (calculated from aspect ratio)
{inst_height}, # height (fixed)
image_path="{institution_logo_path}"
)'''
# Add conference logo to top-right
if conference_logo_path and os.path.exists(conference_logo_path):
conf_width, conf_height = get_logo_dimensions(conference_logo_path, logo_height)
if logo_code: # Add newline if there's already institution logo code
logo_code += '\n'
logo_code += f'''
# Add conference logo to top-right (aspect ratio preserved)
conference_logo = add_image(
poster_slide,
'conference_logo',
{width_inch - conf_width - logo_margin}, # left (right-aligned with calculated width)
{logo_margin}, # top
{conf_width}, # width (calculated from aspect ratio)
{conf_height}, # height (fixed)
image_path="{conference_logo_path}"
)'''
# Insert logo code before saving the presentation
if logo_code:
# Find the position to insert (before save_presentation)
save_pos = poster_code.find('\n# Save the presentation')
if save_pos != -1:
# Insert before the newline that precedes "# Save the presentation"
poster_code = poster_code[:save_pos] + '\n' + logo_code + poster_code[save_pos:]
else:
# Fallback: Find just the comment without newline
save_pos = poster_code.find('# Save the presentation')
if save_pos != -1:
poster_code = poster_code[:save_pos] + logo_code + '\n\n' + poster_code[save_pos:]
else:
# Try alternative search pattern for save_presentation call
pattern = r'(save_presentation\s*\([^)]+\))'
match = re.search(pattern, poster_code)
if match:
# Insert before the save_presentation call
insert_pos = match.start()
poster_code = poster_code[:insert_pos] + logo_code + '\n\n' + poster_code[insert_pos:]
return poster_code
if __name__ == "__main__":
main()