Python Checkpoint: Read, Validate, and Export Structured Data

Combine reading, schema validation and exporting in one short repeatable exercise.

How It Works

A data pipeline should make every phase explicit: load source data, validate a normalized record, then serialize the output. Separate the phases so you can explain failures without corrupting files. This checkpoint uses in-memory data to allow safe experimentation.

Validate before export and test the output by reading it back.

Step-by-Step Example

import csv
import json
from io import StringIO

source = '[{"sku": "A-1", "count": 4}, {"sku": "B-2", "count": 2}]'
items = json.loads(source)
for item in items:
    if not isinstance(item["count"], int) or item["count"] < 0:
        raise ValueError("Invalid count")
output = StringIO(newline="")
writer = csv.DictWriter(output, fieldnames=["sku", "count"])
writer.writeheader()
writer.writerows(items)
print(output.getvalue().splitlines()[-1])

Understanding the Example

The expression '\'[{"sku": "A-1", "count": 4}, {"sku": "B-2", "count": 2}]\'' supplies the value stored as 'source'. This value is then used or checked by the later statements.
The expression 'json.loads(source)' supplies the value stored as 'items'. This value is then used or checked by the later statements.

A successful run of this example prints the following output:

B-2,2

Where This Fits in a Real Program

For event timestamps, distinguish the displayed local clock from the instant that it represents. Store or exchange an unambiguous aware value, then convert it when displaying it to a user. For repeated calendar appointments, use a geographic timezone so daylight-saving rules are not lost.

Practice: Try It Yourself

Add a third record with a missing sku field and decide how the validator should respond.

For this exercise, predict the result before running the changed code. Then explain the difference between the original and modified cases using the actual values, not just whether the program completed.

Common Beginner Mistakes

Exporting a table before checking that the source records contain the expected columns.
Treating a successful file write as evidence that the exported values are correct.

Key Takeaways

Combine reading, schema validation and exporting in one short repeatable exercise.
A data pipeline should make every phase explicit: load source data, validate a normalized record, then serialize the output.
Validate before export and test the output by reading it back.