export_apple_notes.py

import os
import sqlite3
import html2text
import gzip
import io
import string

# Path to the Notes database
DB_PATH = os.path.expanduser('~/Library/Group Containers/group.com.apple.notes/NoteStore.sqlite')
EXPORT_DIR = 'exported_notes_txt'

def clean_filename(name):
    # Remove characters not allowed in filenames
    return "".join(c for c in name if c.isalnum() or c in (" ", "_", "-")).rstrip()

def try_decompress_gzip(data):
    try:
        with gzip.GzipFile(fileobj=io.BytesIO(data)) as f:
            return f.read().decode('utf-8', errors='replace')
    except Exception:
        return None

def is_mostly_printable(line, threshold=0.7):
    printable = set(string.printable)
    if not line.strip():
        return False
    ratio = sum(1 for c in line if c in printable) / len(line)
    return ratio > threshold and any(c.isalpha() for c in line)

def extract_notes():
    if not os.path.exists(DB_PATH):
        print(f"Database not found at {DB_PATH}")
        return

    os.makedirs(EXPORT_DIR, exist_ok=True)
    conn = sqlite3.connect(DB_PATH)
    cursor = conn.cursor()

    # Get mapping of Z_PK to note titles from ZICCLOUDSYNCINGOBJECT
    try:
        cursor.execute("""
            SELECT Z_PK, ZTITLE1 FROM ZICCLOUDSYNCINGOBJECT WHERE ZTITLE1 IS NOT NULL
        """)
        pk_to_title = {row[0]: row[1] for row in cursor.fetchall()}
    except Exception as e:
        print(f"Error reading titles: {e}")
        conn.close()
        return

    # Now get note data from ZICNOTEDATA
    try:
        cursor.execute("""
            SELECT ZNOTE, ZDATA FROM ZICNOTEDATA WHERE ZDATA IS NOT NULL
        """)
        notes = cursor.fetchall()
    except Exception as e:
        print(f"Error reading note data: {e}")
        conn.close()
        return

    h = html2text.HTML2Text()
    h.ignore_links = True

    for idx, (znote, zdata) in enumerate(notes, 1):
        title = pk_to_title.get(znote, f'note_{idx}')
        filename = clean_filename(title) or f'note_{idx}'
        file_path = os.path.join(EXPORT_DIR, f"{filename}.txt")
        if not zdata:
            continue
        # Try to decompress as gzip
        html_content = try_decompress_gzip(zdata)
        if html_content is None:
            # Fallback: try to decode as utf-8
            try:
                html_content = zdata.decode('utf-8', errors='replace')
            except Exception:
                html_content = str(zdata)
        plain_text = h.handle(html_content)
        # Filter out mostly gibberish lines
        cleaned_lines = [l for l in plain_text.splitlines() if is_mostly_printable(l)]
        cleaned_text = "\n".join(cleaned_lines)
        with open(file_path, 'w', encoding='utf-8') as f:
            f.write(cleaned_text)
        print(f"Exported: {file_path}")

    conn.close()
    print(f"Exported {len(notes)} notes to {EXPORT_DIR}/")

if __name__ == "__main__":
    extract_notes()