This project contains a minimal method-chained API to deal with .jsonl and .csv files.
The API is inspired by dplyr and is designed to be lazy.
Everything is a generator until you call .collect().
- Read
.jsonlfiles from local paths or URLs - Read
.csvfiles from local paths or URLs - Lazy evaluation - data is only loaded as needed
- Method chaining for data transformation
- Support for custom delimiters and field names
from lazylines import read_jsonl, read_csv
# Read JSONL from URL
pokemon = read_jsonl("https://calmcode.io/static/data/pokemon.jsonl")
for p in pokemon.head(5):
print(p['name'])
# Read CSV from local file
data = read_csv("data.csv")
# Read CSV from URL with head() for row limit
iris = read_csv("https://raw.githubusercontent.com/mwaskom/seaborn-data/master/iris.csv").head(100)
# Chain operations
results = (
read_csv("data.csv")
.mutate(new_col=lambda d: d["age"] * 2)
.keep(lambda d: d["age"] > 25)
.select("name", "new_col")
.collect()
)The best way to explore is to check the docs on GitHub pages.