CSVQLIZER

Disclaimer: I am on the week 4 problem set of Harvard's CS50 course 2025 and realized I could create this useful thing for myself.

This is a very rough project that aims to accomplish a common work task:

Given a CSV export from a random records management system, how can I rename the columns to be dbeaver sql-autocomplete compatible and create a table to import to where the data type is varchar?

Why varchar?

I find editing/transforming/analyzing CSVs to be easier in postgres than in excel, because I can write whatever logic I want. I also prefer having empty values as null (as opposed to making everything text) It is a lot easier for me to treat everything as a string and do operations like split_part, concat_ws, and ilike '%text%' and transform the data into a new CSV for loading into an existing script I have with appropriate data types.

What do you mean by dbeaver sql-autocomplete compatible?

dbeaver's autocomplete is very good if you have created a table with columns like:

create table deployment.example

(

column_one varchar,

column_two varchar

)

and doesn't guess well if create a table like this:

create table deployment.bad_example

(

"column one" varchar,

"column two" varchar

)

What does this currently do?

• Identify the first line of a CSV with unix linebreaks (windows forthcoming because that's what I work on)
• Prints out a fixed first line of the CSV in the terminal to replace the first line in the input CSV
• Prints out columns in the example format above to help with creating a sql table to import the data into

Planned Features

• Detect windows or linux line endings and behave appropriately
• Produce a file ending with _sqlized.csv with the corrected first line
• Produce a basic sql file giving the entire skeleton of the varchar table to be created