Python script to clean messy sales CSV and generate a summary report

Hi there, I run a small online store and every month I export our sales transactions from our point-of-sale system as a CSV file. The export is messy — inconsistent date formats, prices with dollar signs and commas, extra whitespace, some duplicate rows, and the occasional blank/invalid row. Right now I clean it by hand in Excel and it takes me an hour or two every month. I'd like to stop doing that. I'm looking for a **command-line Python script** that I can run on my own laptop. I point it at the raw CSV export and it (1) cleans the data and saves a tidy CSV, and (2) produces a short summary report so I can see how the month went at a glance. I'm not a developer but I can run a command in the terminal if you give me clear instructions. Please keep it simple to install and run — I'd strongly prefer **standard Python with no heavy setup** (or minimal, well-documented dependencies). ## What the raw data looks like Each row is one sale. Columns (header names roughly as below, but please be a little tolerant of extra spaces/case): `order_id, date, customer, product, quantity, unit_price` Real-world messiness you'll need to handle: - Dates come in mixed formats: `2026-01-05`, `01/05/2026`, `Jan 5, 2026`. - `unit_price` may look like `$1,299.00`, `1299`, or ` 19.99 `. - `quantity` may have stray spaces or be blank. - Leading/trailing whitespace in text fields; inconsistent capitalization in product names. - Fully duplicated rows (same order_id and everything) should be de-duplicated. - Rows missing an order_id, product, or with non-numeric/zero/negative price or quantity are **invalid** — skip them, but tell me how many were skipped and why. ## Deliverables / Requirements 1. **A single Python CLI script** (e.g. `python sales_report.py input.csv`) that accepts: - the path to the input CSV (required), and - an optional output folder (default: current directory). 2. **Cleaned CSV output** (`cleaned_sales.csv`) with: - normalized dates (ISO `YYYY-MM-DD`), - numeric `unit_price` and `quantity` (no symbols), - a computed `line_total` column (quantity × unit_price), - trimmed text, title-cased product names, - duplicates removed and invalid rows excluded. 3. **A summary report** (both a human-readable `report.txt` / Markdown **and** a machine-readable `summary.json`) containing: - total revenue, total number of valid orders, total units sold, average order value, - revenue and units **by product** (sorted high → low), - revenue **by month**, - **top 5 customers** by revenue, - a data-quality section: how many rows were read, skipped, and de-duplicated, with reasons. 4. **A short README** with setup and exact run instructions, plus a small **sample input CSV** and the **example output** it produces, so I can confirm it works before I run it on my real data. 5. Clean, commented code. Sensible error message if the file is missing or the columns don't match (don't just crash with a stack trace). ## Nice to have (not required) - Works with just the Python standard library so I don't have to install anything. - A `--top` option to change how many top customers/products are shown. ## How to apply Please confirm you've read the requirements, mention any clarifying questions, and let me know your timeline. A quick note on how you'd handle the messy date/price parsing is a plus. Thanks!