import re

def convert_mysql_to_pg(input_file, output_file):
    print(f"Converting {input_file} to {output_file}...")
    with open(input_file, 'r', encoding='utf-8') as f, open(output_file, 'w', encoding='utf-8') as out:
        out.write("-- Converted PostGIS Migration Script for PostgreSQL (Compatibility Mode)\n")
        out.write("SET statement_timeout = 0;\n")
        out.write("SET client_encoding = 'UTF8';\n")
        # Enable MySQL-style backslash escaping in strings
        out.write("SET standard_conforming_strings = off;\n")
        out.write("SET escape_string_warning = off;\n")
        out.write("CREATE EXTENSION IF NOT EXISTS postgis;\n\n")

        for line in f:
            if line.startswith('--') or not line.strip() or line.startswith('/*') or line.startswith('SET ') or line.startswith('UNSET '):
                continue
            
            if "COMMIT;" in line or "START TRANSACTION;" in line:
                continue

            # 1. Standard Conversion (backticks to double quotes for identifiers)
            # We must be careful not to replace backticks inside strings (though unlikely in these dumps)
            line = line.replace('`', '"')
            
            # 2. Cleanup column definitions
            line = re.sub(r"CHARACTER SET \w+ COLLATE \w+", "", line)
            
            if "CREATE TABLE" in line:
                line = line.replace("int(11)", "integer")
                line = line.replace("int NOT NULL", "integer NOT NULL")
                line = line.replace("bigint(20)", "bigint")
                line = line.replace("double NOT NULL", "double precision NOT NULL")
                line = line.replace("POINT NOT NULL SRID 4326", "geometry(Point, 4326)")
                line = line.replace("point NOT NULL", "geometry(Point, 4326)")
                line = re.sub(r"ENGINE=InnoDB.*?;", ";", line)

            # 3. Handle Hex data (NULL works best, triggers will fix it)
            line = re.sub(r"0x[0-9a-fA-F]+", "NULL", line)
            
            # Since standard_conforming_strings = off, we DON'T need to replace \' with ''
            # Postgres will understand \' as a literal quote.
            # We just need to ensure backslashes themselves are handled if they were double-escaped.
            # Actually, with standard_conforming_strings = off, MySQL's escaping should work AS IS.
            # But wait, my previous script changed \' to ''. Let's revert that and just clean Definitions.

            # 4. Fix AUTO_INCREMENT
            if "AUTO_INCREMENT" in line:
                line = re.sub(r"integer NOT NULL AUTO_INCREMENT", "SERIAL PRIMARY KEY", line)

            out.write(line)

    print("Success!")

if __name__ == "__main__":
    convert_mysql_to_pg('geocodeDB (1).sql', 'converted_data.sql')