reparatio-cli 0.1.0

Command-line tool for the Reparatio data conversion API (alpha)

reparatio — CLI

Alpha software. Commands, options, and output formats may change without notice between versions.

Command-line tool for the Reparatio data conversion API.

Convert, inspect, merge, append, and query CSV, Excel, Parquet, JSON, GeoJSON, SQLite, and 30+ other formats directly from your terminal.

See also: reparatio-mcp (MCP server for AI assistants) · reparatio-sdk (Python SDK)

---

Installation

Via pipx (recommended — isolated, no dependency conflicts):

pipx install reparatio-cli

Via uv:

uv tool install reparatio-cli

Via pip:

pip install reparatio-cli

All three methods install a reparatio command on your PATH.

---

Quick start

# 1. Set your API key (get one at reparatio.app)
reparatio key set rp_YOUR_KEY

# 2. Inspect a file
reparatio inspect sales.csv

# 3. Convert it
reparatio convert sales.csv sales.parquet

---

Pricing

Plan	Price	Max file	Formats	API / CLI / MCP
Free (anonymous)	$0	10 MB	Basic (csv, tsv, xlsx, json, jsonl, parquet)	No
Free (registered)	$0	10 MB	Basic	No
Standard	$29/month	500 MB	All formats	No
Professional	$79/month	2 GB	All formats + Fixed-Width + EBCDIC	Yes
Credits	$10 = 25 conversions	500 MB	All formats (no API/FWF/EBCDIC)	No

The Professional plan is required to use the CLI, REST API, Python/JS SDKs, MCP server, Fixed-Width parser, and EBCDIC encoding support.

Get an API key at reparatio.app.

---

Authentication

The API key can be configured in three ways, in order of precedence:

1. Environment variable: export REPARATIO_API_KEY=rp_...
2. Stored key: reparatio key set rp_... (saved to ~/.config/reparatio/config.json)
3. No key: only inspect and formats work without a key

---

Command reference

reparatio key

Manage your API key.

reparatio key set rp_YOUR_KEY   # Save to disk
reparatio key show               # Print the current key
reparatio key clear              # Remove the stored key

---

reparatio me

Show subscription details for the current API key.

$ reparatio me

  Email      [e-mail]
  Plan       pro
  Active     yes
  API access yes
  Expires    2026-04-01T00:00:00Z

---

reparatio formats

List all supported input and output formats.

reparatio formats          # Pretty table
reparatio formats --json   # Raw JSON

No API key required.

---

reparatio inspect FILE

Inspect a file: encoding, row count, column types, null counts, unique counts, and a data preview.

reparatio inspect sales.csv
reparatio inspect report.xlsx --sheet Q3
reparatio inspect data.csv --preview-rows 20
reparatio inspect data.tsv --delimiter $'\t'
reparatio inspect file.csv --json    # Machine-readable output

Options:

Option	Default	Description
--preview-rows N	8	Number of preview rows (1–100)
--no-header	off	Treat first row as data
--no-fix-encoding	off	Disable encoding repair
--delimiter TEXT	auto	Custom delimiter
--sheet TEXT	first	Sheet or table name (Excel, ODS, SQLite)
--json	off	Output raw JSON

No API key required.

---

reparatio convert INPUT [OUTPUT]

Convert a file to any supported format.

The output format is inferred from the OUTPUT filename extension. If OUTPUT is omitted, the result is written next to INPUT with the appropriate extension.

# Basic
reparatio convert sales.csv sales.parquet
reparatio convert report.xlsx report.csv

# Output format without specifying a full output path
reparatio convert sales.csv --format json.gz

# Select columns
reparatio convert big.csv slim.csv --select date,region,revenue

# Rename columns
reparatio convert data.csv renamed.csv --rename Date,Region,Revenue

# Deduplicate
reparatio convert events.csv clean.csv --deduplicate

# Sample
reparatio convert huge.csv sample.csv --sample-n 10000
reparatio convert huge.csv sample.csv --sample-frac 0.05

# Encoding issues (legacy Windows CSV)
reparatio convert legacy.csv fixed.xlsx

# EBCDIC mainframe file (cp037 = EBCDIC US, cp500 = EBCDIC International)
reparatio convert mainframe.dat output.csv --encoding cp037
reparatio convert ibm.dat output.csv --encoding cp500

# Read a specific Excel sheet
reparatio convert workbook.xlsx q3.csv --sheet Q3

# Compressed output
reparatio convert data.csv data.csv.gz
reparatio convert data.csv data.parquet    # already compressed by Parquet

Options:

Option	Description
--format TEXT	Output format (inferred from OUTPUT if omitted)
--no-header	Treat first row as data
--no-fix-encoding	Disable encoding repair
--delimiter TEXT	Custom delimiter for CSV-like input
--sheet TEXT	Sheet or table name
--select col1,col2	Columns to include in output
--rename col1,col2	New column names in order
--deduplicate	Remove duplicate rows
--sample-n N	Random sample of N rows
--sample-frac F	Random sample fraction (e.g. 0.1 for 10%)
--geometry-column TEXT	WKT geometry column for GeoJSON output
--null-values TEXT	Comma-separated strings to treat as null, e.g. "N/A,NULL,-"
--cast COL=TYPE[:FORMAT]	Override a column type (repeatable)
--encoding TEXT	Force a specific encoding (bypasses auto-detection), e.g. cp037 for EBCDIC US, cp500 for EBCDIC International

--null-values example:

# Treat "N/A", "NULL", and "-" as null values when loading
reparatio convert legacy.csv clean.parquet --null-values "N/A,NULL,-"

--cast examples:

# Cast price to Float64
reparatio convert data.csv out.parquet --cast price=Float64

# Parse dates in day/month/year format
reparatio convert data.csv out.parquet --cast date=Date:"%d/%m/%Y"

# Multiple overrides
reparatio convert data.csv out.parquet --cast price=Float64 --cast qty=Int32

Supported types: String, Int8–Int64, UInt8–UInt64, Float32, Float64, Boolean, Date (optionally Date:FORMAT), Datetime (optionally Datetime:FORMAT), Time.

---

reparatio batch-convert ZIP_FILE

Convert every file inside a ZIP archive to a common format. Returns a ZIP of converted files. Files that cannot be parsed are skipped with a warning.

reparatio batch-convert monthly_reports.zip --format parquet
reparatio batch-convert raw_data.zip --format csv.gz -o processed.zip
reparatio batch-convert data.zip --format parquet --cast price=Float64

Options:

Option	Description
--format TEXT	Output format (required)
--output FILE	Output ZIP path (default: converted.zip)
--no-header	Treat first row as data
--no-fix-encoding	Disable encoding repair
--delimiter TEXT	Custom delimiter
--select col1,col2	Columns to include from every file
--deduplicate	Remove duplicate rows from each file
--sample-n N	Random sample of N rows per file
--sample-frac F	Random sample fraction per file
--cast COL=TYPE[:FORMAT]	Column type override (repeatable)

---

reparatio merge FILE1 FILE2

Merge or join two files.

# Left join on customer_id, output as Parquet
reparatio merge orders.csv customers.xlsx --op left --on customer_id --format parquet

# Stack two files (append/union)
reparatio merge jan.csv feb.csv --op append --format csv -o q1.csv

# Inner join on multiple columns
reparatio merge a.csv b.csv --op inner --on "user_id,date" --format xlsx

--op values:

Value	Behaviour
append	Stack all rows from both files; missing columns filled with null
left	All rows from FILE1; matching columns from FILE2
right	All rows from FILE2; matching columns from FILE1
outer	All rows from both files; nulls where no match
inner	Only rows present in both files

Options:

Option	Description
--op	Join operation (required)
--format TEXT	Output format (required)
--on col1,col2	Columns to join on (not needed for append)
--output / -o PATH	Output path (defaults to FILE1 directory)
--no-header	Treat first row as data
--no-fix-encoding	Disable encoding repair
--geometry-column TEXT	WKT geometry column for GeoJSON output

---

reparatio append FILES...

Stack rows from two or more files vertically.

Column mismatches are handled gracefully — missing values are filled with null.

# Stack three monthly CSVs into a single Parquet file
reparatio append jan.csv feb.csv mar.csv --format parquet -o q1.parquet

# Using shell glob
reparatio append monthly/*.csv --format csv -o all_months.csv

Options:

Option	Description
--format TEXT	Output format (required)
--output / -o PATH	Output path (defaults to appended.<format> next to the first file)
--no-header	Treat first row as data
--no-fix-encoding	Disable encoding repair

---

reparatio query FILE SQL

Run a SQL query against a file. The file is loaded as a table named data.

# Aggregate
reparatio query events.parquet "SELECT region, SUM(revenue) AS total FROM data GROUP BY region ORDER BY total DESC"

# Filter and limit
reparatio query sales.csv "SELECT * FROM data WHERE amount > 1000 ORDER BY date DESC LIMIT 50" --format xlsx

# Read a specific sheet
reparatio query workbook.xlsx "SELECT COUNT(*) FROM data" --sheet Summary

# Output as JSON
reparatio query data.parquet "SELECT * FROM data LIMIT 10" --format json -o preview.json

Supports SELECT, WHERE, GROUP BY, ORDER BY, LIMIT, aggregations, and most scalar functions.

Options:

Option	Default	Description
--format TEXT	csv	Output format
--output / -o PATH	auto	Output path
--no-header	off	Treat first row as data
--no-fix-encoding	off	Disable encoding repair
--delimiter TEXT	auto	Custom delimiter for CSV-like input
--sheet TEXT	first	Sheet or table name

---

Supported formats

Input

CSV, TSV, CSV.GZ, CSV.BZ2, CSV.ZST, CSV.ZIP, TSV.GZ, TSV.BZ2, TSV.ZST, TSV.ZIP, GZ (any supported format), ZIP (any supported format), BZ2 (any supported format), ZST (any supported format), Excel (.xlsx / .xls), ODS, JSON, JSON.GZ, JSON.BZ2, JSON.ZST, JSON.ZIP, JSON Lines, GeoJSON, Parquet, Feather, Arrow, ORC, Avro, SQLite, YAML, BSON, SRT, VTT, HTML, Markdown, XML, SQL dump, PDF (text layer)

Output

CSV, TSV, CSV.GZ, CSV.BZ2, CSV.ZST, CSV.ZIP, TSV.GZ, TSV.BZ2, TSV.ZST, TSV.ZIP, Excel (.xlsx), ODS, JSON, JSON.GZ, JSON.BZ2, JSON.ZST, JSON.ZIP, JSON Lines, JSON Lines.GZ, JSON Lines.BZ2, JSON Lines.ZST, JSON Lines.ZIP, GeoJSON, GeoJSON.GZ, GeoJSON.BZ2, GeoJSON.ZST, GeoJSON.ZIP, Parquet, Feather, Arrow, ORC, Avro, SQLite, YAML, BSON, SRT, VTT

Run reparatio formats for the authoritative live list.

---

Exit codes

Code	Meaning
0	Success
1	Error (authentication, parse failure, API error, etc.)

---

Privacy

Files are sent to the Reparatio API at reparatio.app for processing. Files are handled in memory and never stored — see the Privacy Policy.