#!/usr/bin/env python3
"""
Format .bbl file to use Author Year. format instead of {Author} (Year).

This script processes the bibliography file generated by BibTeX and reformats
entries to display as "Author Year." instead of "{Author} (Year)." in the
references section.

Usage:
    python format_bbl.py <bbl_file>
"""

import re
import sys


def format_bibliography(bbl_file):
    """
    Format bibliography entries to Author Year. format.
    Also fixes Vietnamese name order from "Given Family" back to "Family Given".

    Args:
        bbl_file: Path to the .bbl file to format
    """
    with open(bbl_file, 'r', encoding='utf-8') as f:
        content = f.read()

    # First, join multiline bibitem entries into single lines
    # Pattern: \bibitem[...](year)]{key} that may span multiple lines
    content = re.sub(r'\\bibitem\[([^\]]+)\]\{([^}]+)\}\s*\n\s*', r'\\bibitem[\1]{\2}\n', content)

    lines = content.split('\n')
    output = []
    i = 0

    # Vietnamese family names to detect
    viet_families = ['Cao', 'Đoàn', 'Hoàng', 'Mai', 'Nguyễn', 'Vũ', 'Vương']

    while i < len(lines):
        line = lines[i]

        # Match bibitem line with format: \bibitem[{Author}(Year)]{key}
        # Handle various formats including fontencoding
        match = re.match(r'\\bibitem\[(.+?)\((\d+[a-z]?|{\\.+?})\)\]\{([^}]+)\}', line)
        # Also handle bibitem without year (editors)
        if not match:
            match = re.match(r'\\bibitem\[(.+?)\]\{([^}]+)\}', line)
            has_year = False
        else:
            has_year = True

        if match:
            if has_year:
                author_part = match.group(1).strip()
                year = match.group(2)
                key = match.group(3)
            else:
                author_part = match.group(1).strip()
                year = None
                key = match.group(2)

            # Write the bibitem line unchanged
            output.append(line)
            i += 1

            # Next line should be the author - replace with Author Year.
            if i < len(lines):
                next_line = lines[i]
                # Skip if line is empty or starts with \newblock
                if next_line.strip() and not next_line.strip().startswith('\\newblock'):
                    # Handle fontencoding patterns
                    clean_line = re.sub(r'\{\\fontencoding\{[^}]+\}\\selectfont ([^}]+)\}', r'\1', next_line)
                    clean_line = re.sub(r'^\{(.+)\}\.?\s*$', r'\1', clean_line.strip())

                    # Fix Vietnamese name order: "Given~Family" -> "Family Given"
                    # and wrap Vietnamese names in T5 encoding
                    for family in viet_families:
                        # Match "Given~Family" or "Given Family"
                        pattern = r'(\S+)~(' + re.escape(family) + r')\b'
                        if re.search(pattern, clean_line):
                            # Swap to "Family Given" and wrap in T5
                            clean_line = re.sub(pattern, r'{\\fontencoding{T5}\\selectfont \2 \1}', clean_line)

                    # Also wrap any remaining Vietnamese family names in T5 if they appear standalone
                    for family in viet_families:
                        # Match "Family Given" pattern not already wrapped
                        pattern = r'\b(' + re.escape(family) + r')\s+(\S+)\b'
                        # Only wrap if not already in fontencoding and contains Vietnamese characters
                        if re.search(pattern, clean_line) and 'fontencoding' not in clean_line:
                            # Check if the second part contains Vietnamese characters
                            match = re.search(pattern, clean_line)
                            if match:
                                given = match.group(2)
                                # Check for Vietnamese characters
                                if any(c in given for c in 'ạảãàáâậầấẩẫăặắằẳẵêệềếểễôộồốổỗơởỡớờợưựứừửữđíìịỉĩóòỏõọúùủũụýỳỷỹỵ'):
                                    clean_line = re.sub(pattern, r'{\\fontencoding{T5}\\selectfont \1 \2}', clean_line, count=1)

                    # Add year with period
                    if has_year and year and not clean_line.endswith(year + '.') and 'natexlab' not in clean_line:
                        clean_line = clean_line.rstrip('.') + ' ' + year + '.'

                    output.append(clean_line)
                    i += 1
                else:
                    output.append(next_line)
                    i += 1
        else:
            output.append(line)
            i += 1

    # Write back the formatted content
    with open(bbl_file, 'w', encoding='utf-8') as f:
        f.write('\n'.join(output))

    print(f"Formatted bibliography in {bbl_file}")


if __name__ == '__main__':
    if len(sys.argv) != 2:
        print("Usage: python format_bbl.py <bbl_file>")
        sys.exit(1)

    bbl_file = sys.argv[1]
    format_bibliography(bbl_file)