A gist to rename Markdown files from their Zotero PDF filenames to their citekey (and embed some metadata)

bib_rename.py

import bibtexparser
import pandas as pd
from pathlib import Path
import os
import re
from datetime import datetime
import yaml

def load_bibtex(bib_path):
    """Load bibtex file and convert to DataFrame."""
    with open(bib_path, 'r', encoding='utf-8') as bibtex_file:
        bib_database = bibtexparser.load(bibtex_file)
    
    # Convert to DataFrame
    df_bib = pd.DataFrame(bib_database.entries)
    # Set ID as index
    df_bib.set_index('ID', inplace=True)
    return df_bib

def clean_author_names(author_name_string):
    """Convert author names from 'Last, First' format to 'First Last' format.
    
    Args:
        author_name_string (str): String of author names in format 'Last, First; Last, First'
        
    Returns:
        str: Author names in format 'First Last, First Last'
    """
    if not isinstance(author_name_string, str):
        print("bad author name string", author_name_string)
        return ""
    # Split multiple authors
    authors = author_name_string.split(';')
    
    # Process each author
    cleaned_names = []
    for author in authors:
        # Split into last and first name
        name_parts = author.strip().split(',')
        if len(name_parts) == 2:
            last_name = name_parts[0].strip()
            first_name = name_parts[1].strip()
            cleaned_names.append(f"{first_name} {last_name}")
    
    # Join all authors with commas
    return ', '.join(cleaned_names)
    

def create_merged_dataset(bib_path='library.bib', csv_path='library.csv'):
    """Create a merged dataset from bibtex and csv files."""
    # Load bibtex
    df_bib = load_bibtex(bib_path)
    df_bib['citekey'] = df_bib.index
    
    # Load CSV
    df_csv = pd.read_csv(csv_path)
    
    # Extract just the filename from the full paths
    df_bib['filename'] = df_bib['file'].str.extract(r'/([^/]+)$')
    df_csv['filename'] = df_csv['File Attachments'].str.extract(r'/([^/]+)$')
    
    # Drop rows with no filename
    df_bib = df_bib[df_bib['filename'].notna()]
    df_csv = df_csv[df_csv['filename'].notna()]

    # Merge on the extracted filenames
    merged_data = df_bib.merge(df_csv, on='filename', how='inner')
    
    clean_data = []
    for _, entry in merged_data.iterrows():
        entry_data = {
            'citekey': entry['citekey'],
            'title': entry['title'].replace('{', '').replace('}', ''),
            'year': entry['year'],
            'authors_raw': entry['Author'],
            'authors': clean_author_names(entry['Author']),
            'type': entry['Item Type'],
            'abstract': entry['abstract'],
            'file': entry['file'],
            'downloadURI': f"zotero://select/library/items/{entry['Key']}",
            'tags': ', '.join(entry['Manual Tags'].split(';')) if pd.notna(entry['Manual Tags']) else ''
        }
        clean_data.append(entry_data)
    
    return pd.DataFrame(clean_data)


def create_markdown_content(entry_data):
    """Create markdown content from template."""
    frontmatter = {
        'year': entry_data['year'],
        'authors': entry_data['authors_raw'],
        'type': entry_data['type']
    }
    
    # Create the content
    content = f"""---
{yaml.dump(frontmatter, sort_keys=False)}---
## {entry_data['title']}
{entry_data['authors']}
{entry_data['year']}
[Open in Zotero]({entry_data['downloadURI']})

> [!info] 
> - **Abstract:** {entry_data['abstract']}
"""

    if entry_data['tags']:
        content += f"\n> - **Tags:** {entry_data['tags']}"

    content += """

>[!Personal Notes]
> 

## Annotations 
"""
    
    return content

def process_files(merged_data, limit=None):
    """Main function to process files.
    
    Args:
        merged_data: DataFrame of entries to process
        limit: Optional int, maximum number of entries to process
    """
    
    processed = 0
    # Process each entry
    for _, entry_data in merged_data.iterrows():
        if limit and processed >= limit:
            print("hit the limit")
            break
            
        # Extract original filename from file field
        original_filename = entry_data['file'].split('/')[-1].replace('.pdf', '')
        original_md_path = f"{original_filename}.md"
        new_md_path = f"@{entry_data['citekey']}.md"
        
        # Check if original markdown file exists
        if not os.path.exists(original_md_path):
            print(f"Warning: No markdown file found for {original_filename}")
            continue
            
        # Create new content
        new_content = create_markdown_content(entry_data)
        
        # Read existing content (might want to preserve some of it)
        with open(original_md_path, 'r', encoding='utf-8') as f:
            existing_content = f.read()
            
        # Write new file
        with open(new_md_path, 'w', encoding='utf-8') as f:
            f.write(new_content)
            if existing_content:
                f.write(existing_content)
        
        # Remove old file if it's different from the new one
        if original_md_path != new_md_path and os.path.exists(original_md_path):
            os.remove(original_md_path)
            
        print(f"Processed: {original_md_path} → {new_md_path}")
        processed += 1

# if __name__ == "__main__":
merged_data = create_merged_dataset()
merged_data.to_csv('library-merged.csv', index=False)

# Process only first entry for testing
process_files(merged_data)
# Or process all entries
# process_files(merged_data)

I used this to update my Obsidian collection of literature notes. Files were named Author Year - Title.md. This renames them to @citekey.md.

In order to use this, you must export your Zotero library both in BibTeX format using the better bibtex plugin, and in CSV using File --> "Export Library...". Rename those files to library.csv and library.bib and put them in the directory with this script and with all the markdown files. Create a backup of the folder before you do anything else. Add the limit=1 parameter to the process_files call at the end to make sure you like the result before you do the mass move.