JSON to Excel / CSV CLI

A small, fast command‑line tool to convert JSON to Excel (XLSX) or CSV. It flattens nested objects into columns and can explode arrays into multiple rows. Input JSON is streamed from disk (via ijson) so you can process large files.

Features

• Stream JSON input from disk (does not load entire file into memory)
• Flatten nested objects into dotted columns (configurable separator)
• Explode arrays into multiple rows for analysis
• Join or JSON-encode non‑exploded lists (--list-policy join|json)
• Discover and order headers with sampling; pin first columns
• Exclude entire column trees by prefix (e.g., --exclude customer.address)
• Write CSV (streaming) or Excel XLSX output (uses openpyxl)

Quickstart

# Install (choose one)
pipx install json-to-excel-converter
# or
pip install json-to-excel-converter
# or, if you use uv
uv tool install json-to-excel-converter

# Convert JSON to CSV
json-to-excel-converter INPUT.json output.csv --root items --first-column id

# Convert JSON to Excel (XLSX)
json-to-excel-converter INPUT.json output.xlsx --root items --sheet-name Items

Installation

For End Users

pipx install json-to-excel-converter
# or
pip install json-to-excel-converter
# or
uv tool install json-to-excel-converter

For Development

git clone https://github.com/vlorenzo/json2excel-cli
cd json2excel-cli
uv sync
. .venv/bin/activate

Usage

json-to-excel-converter INPUT.json OUTPUT.(csv|xlsx) \
  --root items \
  --explode attributes \
  --list-policy join \
  --list-sep "," \
  --sep . \
  --sheet-name Items \
  --sample-headers 10 \
  --header-order stable \
  --first-column id \
  --exclude details \
  --include summary \
  --include details

Options

• --root: path to array/object to process (optional, defaults to top-level array)
• --explode: create separate rows for array elements (repeatable)
• --list-policy: handle arrays as join (comma-separated) or json (JSON string)
• --list-sep: separator for joined arrays (default: ";")
• --sample-headers: rows to scan for column discovery (default: 1000)
• --header-order: column ordering stable (first-seen) or alpha (alphabetical)
• --first-column: pin specific columns to the beginning (repeatable)
• --exclude: remove columns by path prefix (repeatable)
• --include: keep only columns whose path equals or starts with this prefix (repeatable). Ordering of groups follows the flag order; pinned columns still appear first. Within each group, --header-order applies.

FAQ

• How do I select the part of JSON to convert? Use --root with a dotted path like orders.items or a JSON Pointer like /orders/items. If your JSON starts with an array, you can omit --root.
• My root is an object, not an array. What happens? By default, the tool expects an array. If your root points to an object, pass --allow-object-values to iterate that object's values.
• How do I explode arrays into rows? Pass --explode path (repeatable) for each array you want to expand. Multiple --explode flags create a cartesian product across those arrays.
• How are lists handled if I don't explode them? Choose --list-policy join (default) to join scalar lists with --list-sep or --list-policy json to JSON‑encode the list.

Examples

Understanding the Sample Data

The sample contains e-commerce order data with nested structures:

{
  "orders": [
    {
      "order_id": "ORD001",
      "customer": {"name": "John Smith", "address": {"city": "New York"}},
      "items": [{"product": "Laptop", "price": 1299.99}, {"product": "Mouse", "price": 29.99}],
      "payment": {"method": "credit_card", "status": "completed", "total": 1359.97},
      "tags": ["priority", "business"]
    }
    // ... 3 orders total with nested customer info, multiple items, payments, tags
  ]
}

Get Sample Data

curl -O https://raw.githubusercontent.com/vlorenzo/json2excel-cli/main/sample.json

1. Basic Flattening (Nested Objects → Columns)

json-to-excel-converter sample.json orders.csv --root orders --first-column order_id

What you get: 3 rows (1 per order), nested objects become dotted columns:

• order_id, customer.name, customer.email, customer.address.street, customer.address.city, payment.method, payment.status, etc.
• Arrays as strings: items becomes JSON, tags becomes "priority;business"

2. Array Explosion - One Row Per Item Purchased

json-to-excel-converter sample.json items.csv --root orders --explode items --first-column order_id

What you get: 6 rows (one per item across all orders):

• order_id, customer.name, items.product, items.brand, items.price, items.quantity
• ORD001 creates 2 rows (Laptop + Mouse), ORD003 creates 3 rows (Keyboard + Monitor + Cable)

3. Multiple Array Explosions - Cartesian Product

json-to-excel-converter sample.json detailed.csv --root orders --explode items --explode tags

What you get: 12 rows (items × tags combinations)

• Every item gets a row for each tag of that order

4. Remove Unwanted Columns

json-to-excel-converter sample.json clean.csv --root orders --exclude customer.address --first-column order_id

What you get: Same structure but removes customer.address.street, customer.address.city, etc.

5. Excel Output with Custom Sheet Name

json-to-excel-converter sample.json orders.xlsx --root orders --sheet-name "Customer Orders"

6. Working with Your Own Data

Replace sample.json with your JSON file:

# For JSON starting with an array (no --root needed)
json-to-excel-converter your-data.json output.csv --first-column id

# For JSON with nested "products" array
json-to-excel-converter your-data.json output.csv --root products --first-column id

# Remove unwanted columns
json-to-excel-converter your-data.json output.csv --exclude personal_info --exclude internal

### 7. Include Only Certain Columns (and Order by Flag Order)
```bash
json-to-excel-converter sample.json selected.csv \
  --root orders \
  --first-column order_id \
  --include payment \
  --include customer

What you get: only order_id, then all payment.* columns, then all customer.* columns. Within each group, ordering follows --header-order.

### Row Count Summary
- **Basic**: 3 rows (1 per order)
- **Explode items**: 6 rows (multiple items per order)
- **Explode tags**: 5 rows (different tag counts)
- **Explode both**: 12 rows (items × tags cartesian product)


## When to prefer CSV over XLSX

XLSX writing is convenient but the workbook is kept in memory by `openpyxl`,
so for very large outputs CSV is usually faster and more memory‑efficient.

Consider using CSV when any of these apply:

- Your input JSON is larger than ~200 MB
- You approach Excel’s sheet limit of 1,048,576 rows

This tool streams JSON input, so it typically handles files up to around 1 GB
without issues when writing CSV.


## Contributing

### Development Setup
```bash
git clone https://github.com/vlorenzo/json2excel-cli
cd json2excel-cli
uv sync
. .venv/bin/activate

Running Tests

uv run pytest -q
# Run only CLI tests
uv run pytest -q tests/cli

Code Quality

uv run black .
uv run ruff check .