Utility functions that generate a dictionary with column name and corresponding varchar value.

dtypes_handler.py

import pandas as pd
import re

#example string
dtypes_correction = """
[sales] --5
,[id] --50
,[project_id]--50
,[another_column]--31
"""

def dtypes_updater(factor: int, dtypes_correction: str):
    dtypes = {}
    for d in dtypes_correction.split(","):
        new_string = re.sub(r"[\([{})\]]", "", d)
        d = new_string.split("--")
        val = d[1].replace("\n", "")
        dtypes[d[0].strip()] = f"varchar({int(val)* 2 })"

    print(f"dtypes = {dtypes}")
    

def get_df_varchars(df: pd.DataFrame, excess:int=2):
    """Using maximum lengh of data in columns, function genreates varchar dtypes dictionary."""
    max_length_object_cols = {col: df.loc[:, col].astype(str).apply(len).max() for col in df.columns}
    for x in max_length_object_cols:
        maximum = int(max_length_object_cols[x])
        val = (2 * round(maximum/2)) * excess
        if val < 10:
            val = 10
        max_length_object_cols[x] = f'varchar({val})'
    return max_length_object_cols

dtypes utility