Downloads
CSV

CSV
Module csv
Author Luma Contributors
Version 1.0.0
License MIT

A pure Luma CSV library for parsing, querying, and converting comma-separated data. No external dependencies — works with Luma’s built-in file I/O and JSON support.

Features

  • Read CSV files or parse CSV strings
  • Custom delimiters (tabs, semicolons, etc.)
  • Quoted field handling (commas inside quotes)
  • Column access by name
  • Filtering rows by column value
  • Convert to maps or JSON
  • Write back to CSV with proper quoting

Quick Start 

csv = import "csv"

// Read a CSV file
table: [[str]] = csv.read("data.csv")

// Access data
print(csv.headers(table))        // [name age city]
print(csv.col(table, "name"))    // [Alice Bob Charlie]
print(csv.get(table, 0, "name")) // Alice

// Filter rows
engineers: [[str]] = csv.where(table, "role", "engineer")

// Convert to JSON
print(csv.to_json(table))

Installation

  1. Download the module file: csv.luma
  2. Place it in your project directory (or a lib/ folder)
  3. Import it in your code:
csv = import "csv"

Design

A CSV table is [[str]] — a list of rows, where each row is a list of field strings. The first row (table[0]) is always the header. Fields are split during parsing, so all accessor functions work with already-parsed data.

API Reference

Function Description
read(path: str) -> [[str]] Read a CSV file (comma-delimited)
read_delim(path: str, delim: str) -> [[str]] Read a file with a custom delimiter
parse(text: str) -> [[str]] Parse a CSV string (comma-delimited)
parse_delim(text: str, delim: str) -> [[str]] Parse a string with a custom delimiter
headers(table: [[str]]) -> [str] Get header names
rows(table: [[str]]) -> [[str]] Get all data rows (without header)
row(table: [[str]], index: int) -> [str] Get a single data row by index (0-based)
col(table: [[str]], name: str) -> [str] Get all values in a column by name
get(table: [[str]], row: int, col: str) -> str Get a single cell by row index and column name
count(table: [[str]]) -> int Count data rows (excluding header)
to_maps(table: [[str]]) -> [{str: str}] Convert to a list of header-keyed maps
to_json(table: [[str]]) -> str Convert to a JSON array of objects
where(table: [[str]], col: str, val: str) -> [[str]] Filter rows where a column matches a value
to_string(table: [[str]]) -> str Convert table back to a CSV string
write(path: str, table: [[str]]) Write table to a CSV file

Usage

Reading CSV 

// From a file
table: [[str]] = csv.read("data.csv")

// From a string
text: str = "name,age,city\nAlice,30,Tallinn\nBob,25,Riga"
table: [[str]] = csv.parse(text)

// Tab-separated
tsv: [[str]] = csv.read_delim("data.tsv", "\t")

Accessing data 

h: [str] = csv.headers(table)       // [name age city]
all: [[str]] = csv.rows(table)      // all data rows
first: [str] = csv.row(table, 0)    // first data row
names: [str] = csv.col(table, "name") // all values in "name" column
val: str = csv.get(table, 0, "name")  // single cell
n: int = csv.count(table)             // number of data rows

Filtering

where returns a new table (with header) containing only matching rows:

engineers: [[str]] = csv.where(table, "role", "engineer")
csv.rows(engineers).walk(r) -> print(sprint(r))

Converting 

// To maps (header name -> value)
maps: [{str: str}] = csv.to_maps(table)

// To JSON
j: str = csv.to_json(table)
print(j)

Writing 

// To string
output: str = csv.to_string(table)

// To file
csv.write("output.csv", table)

Quoted fields

The parser handles quoted fields correctly — commas inside quotes are preserved:

text: str = "name,note\nAlice,\"likes commas, really\""
table: [[str]] = csv.parse(text)
print(csv.get(table, 0, "note"))   // likes commas, really

When writing, fields containing commas, quotes, or newlines are automatically quoted.

Complete Example 

csv = import "csv"

// Read the data
table: [[str]] = csv.read("employees.csv")
print("Loaded ${csv.count(table)} employees")

// Show all names
names: [str] = csv.col(table, "name")
print("Names: " + sprint(names))

// Filter engineers
engineers: [[str]] = csv.where(table, "role", "engineer")
print("Engineers: ${csv.count(engineers)}")

// Convert to JSON and save
file("engineers.json").write(csv.to_json(engineers))

// Write filtered CSV
csv.write("engineers.csv", engineers)
JWT