fork-db-cleaner

A lightweight Node.js CLI that processes MySQL dump files and removes INSERT INTO statements for specified tables (e.g. ntakmessages). Uses streaming to handle large dumps without loading the entire file into memory.

Requirements

• Node.js 18+

Install

npm install -g @fork-hu/db-cleaner

Or run via npx without installing:

npx @fork-hu/db-cleaner -i dump.sql -o cleaned.sql

Usage

db-cleaner --input <path> --output <path> [--tables table1,table2,...] [--before YYYY-MM-DD] [--created-at-column createdAt]

Alias: fdclean (shorthand)

| Option | Short | Required | Default | Description | |-----------------------|-------|----------|-----------------|--------------------------------------------------------------------------| | --input | -i | Yes | — | Path to the source MySQL dump file | | --output | -o | Yes | — | Path to write the cleaned SQL file | | --tables | -t | No | ntakmessages | Comma-separated table names to process | | --before | -b | No | — | Cutoff date: keep only rows where createdAt ≥ this date (YYYY-MM-DD) | | --created-at-column | — | No | createdAt | Column name used for date filtering (when --before is set) |

Examples

Remove only ntakmessages inserts (default):

db-cleaner -i dump.sql -o cleaned.sql

Remove inserts for multiple tables:

db-cleaner -i dump.sql -o cleaned.sql -t ntakmessages,audit_log,sessions

PowerShell users: Quote arguments containing commas, or use multiple -t flags:

db-cleaner -i dump.sql -o cleaned.sql -t "ntakmessages,orderitems" -b 2025-01-01
db-cleaner -i dump.sql -o cleaned.sql -t ntakmessages -t orderitems -b 2025-01-01

Keep only recent records (filter by createdAt):

db-cleaner -i dump.sql -o cleaned.sql -b 2025-01-01

Output

The tool prints a summary when done:

Done. 1,204,300 lines processed, 98,440 lines removed. Output: cleaned.sql

With --before, it also reports rows kept vs removed:

Done. 196,701 lines processed, 7,715 lines removed. Rows kept: 9,003, rows removed: 7,179. Output: dump-filtered.sql

The output file is a valid MySQL dump and can be imported with:

mysql -u user -p database < cleaned.sql