What you’ll learn to build
- Read and inspect a CSV without changing it
- Explain normalization, validation and deduplication by testing a real example
- Ship a repeatable command with an audit report and regression tests
Who it’s for
Beginners who have run one Python script and want practical data engineering foundations. Start after running and modifying a small Python script.
Two focused lessons
- 01
Run and understand the working tool
- 02
Test, extend and document the workflow
Explore the curriculum topics
- CSV is a format, not a clean database
- Start with reversible inputs
- Dictionary rows over positional guessing
- Whitespace and the first normalized row
- Case-insensitive deduplication
- Required fields are a policy
- Read the audit counts correctly
- Quoted values and newline discipline
- Unknown headers should fail loudly
- CSV formula risks
- Output overwrite controls
- Build a reliable command log
- First milestone: verify on a new dataset
- Treat a dedupe key as a domain decision
- Empty lines versus incomplete rows
- Malformed CSV structure
- Header quality
- Unicode and encoding
- Idempotence as a useful property
- Use tempfile instead of your desktop
- A focused unittest assertion
- Reconcile before trusting automation
- Privacy-aware sample design
- Bound memory use
- Test the failure path
- Portfolio capstone and next step
Project boundaries
Processes CSV data in memory. Email normalization does not verify deliverability. Spreadsheet-safe escaping changes exported values.