ETL across formats¶
Materializes a tiny CSV, reads it back through smart dispatch (csv → list[dict]), filters with a for loop, and saves to JSON — same save() builtin picks the JSON handler from the extension. Demonstrates the smart-dispatch read/save roundtrip.
Source: examples/03_etl.c4
# Tiny ETL: synthesize a CSV, transform it, save as JSON.
# Demonstrates: smart read/save roundtrip across formats, for-loop
# over rows, in-place dict construction.
# Materialize a small dataset on disk.
seed = [
{"name": "ada", "age": 36},
{"name": "lin", "age": 41},
{"name": "tim", "age": 22},
{"name": "bea", "age": 17},
]
save(seed, "./_etl_input.csv")
# Read back through smart-dispatch (csv → list[dict]).
rows = read("./_etl_input.csv")
# Transform: keep adults, normalize names to upper case.
adults = []
for r in rows {
age = int(r["age"])
if age >= 18 {
adults.append({"name": r["name"].upper(), "age": age})
}
}
# Save in a different format — handler picks JSON automatically.
save(adults, "./_etl_output.json")
log("etl done", input=len(rows), output=len(adults))