Takes a log file with dbt's output and indicates what models/tests are taking longer

parse_dbt_log.py

from decimal import Decimal
import re

REGEX = r'^\d\d:\d\d:\d\d  \d+ of \d+ (OK|PASS) (.*) \.*\s+\[(PASS|SUCCESS \d+) in (\d+\.\d+)s\]$'


def parse_file(file_path: str) -> list[tuple[str, str, Decimal]]:
    with open(file_path) as f:
        lines = f.readlines()

    regex = re.compile(REGEX)
    results = []
    for line in lines:
        match = regex.match(line)
        if match:
            _, text, _, timing = match.groups()
            if 'created sql incremental model' in text:
                type = 'model'
            elif 'created sql view model' in text:
                type = 'view'
            else:
                type = 'test'
            results.append(
                (
                    text.replace('created sql incremental model ', '').replace('created sql view model ', ''),
                    type,
                    Decimal(timing)
                )
            )

    return sorted(results, key=lambda x: x[2], reverse=True)


def main() -> None:
    for model, type, timing in parse_file('log.log'):
        print(f'{timing}s - \t{type}:{model}')


if __name__ == '__main__':
    main()