How can I convert a small JSON array of objects to CSV?

Decide the CSV schema first: the exact columns, their order, and what to do when an object does not match. Python’s JSON decoder represents an array as a list and objects as dictionaries, so you can check the structure before writing. [S2]

For example, suppose the input is:

[
  {"name": "Mina", "team": "Design", "city": "Riga"},
  {"name": "Omar", "team": "Support"},
  {"name": "Lee", "team": "Sales", "city": "Oslo", "level": 2}
]

You might choose name, team, city as the fields, in that order. Decide whether a missing city becomes a blank cell and whether an unlisted key such as level makes the record incompatible. Reporting the extra key and skipping that record is one clear policy; adding level to the schema is another. Choose deliberately so data is not lost without notice.

This example writes to a different file, accepts missing listed fields as blank, and reports records with extra fields instead of writing them. Replace the path placeholders. DictWriter uses the explicit fieldnames order for columns; its default behavior raises on extra keys, so the code checks them first. [S1]

import csv
import json
from pathlib import Path

json_path = Path("[input.json]")
csv_path = Path("[output.csv]")
fields = ["name", "team", "city"]

with json_path.open(encoding="utf-8") as source:
    records = json.load(source)

if not isinstance(records, list):
    raise ValueError("Expected the top-level JSON value to be an array")

issues = []
compatible = []

for index, record in enumerate(records):
    if not isinstance(record, dict):
        issues.append(f"Record {index}: expected an object; skipped")
        continue

    missing = [field for field in fields if field not in record]
    extra = [key for key in record if key not in fields]

    if extra:
        issues.append(f"Record {index}: extra keys {extra}; skipped")
        continue

    if missing:
        issues.append(
            f"Record {index}: missing keys {missing}; blank cells written"
        )
    compatible.append(record)

with csv_path.open("w", encoding="utf-8", newline="") as output:
    writer = csv.DictWriter(output, fieldnames=fields, restval="")
    writer.writeheader()
    writer.writerows(compatible)

print(f"Wrote {len(compatible)} records to {csv_path}")
print("Issues:")
for issue in issues:
    print(f"- {issue}")

With the sample, the second record is written with a blank city; the third is skipped and reported because of level. Indexes start at zero, so Record 1 means the second item. The code checks input shape and keys, not whether values suit your project.

The CSV would contain a header and the first two records:

name,team,city
Mina,Design,Riga
Omar,Support,

The report would include Record 1: missing keys ['city']; blank cells written and Record 2: extra keys ['level']; skipped. The CSV column order follows the explicit fields list. [S1]

If you need to retain level, add it intentionally to fields and decide how to handle records where it is missing. Review the report before relying on the CSV.

Check that the CSV has the intended header and order, that the number of written records is expected, and that each reported issue has an intentional resolution. If you revise the field list, rerun the conversion and review the new report.

This example checks that the top-level value is a list, each item is an object, and keys are missing or extra. It does not validate value types or whether the columns meet a particular system’s requirements; define those rules for your intended use.