Parse two JSON objects
Input
parse_json_lines('{"a": 1}\n{"b": 2}')
Output
[{'a': 1}, {'b': 2}]
Two JSON objects are parsed into a list of dictionaries.
Full lesson preview
Stream-parse JSON lines into records
Parse newline-delimited JSON (JSON Lines) into Python records, skipping bad lines and optionally extracting keys.
Python practice22 minModules & Standard LibraryAdvancedLast updated March 23, 2026
Problem statement
You are given a string containing multiple JSON objects separated by newlines (JSON Lines). Implement parse_json_lines(text, keys=None) that reads the input text and returns a list of parsed records. Behavior:
- Empty lines should be ignored.
- Malformed JSON lines should be skipped (do not raise).
- If keys is None, return a list of dictionaries (the parsed JSON objects).
- If keys is a list of strings, for each valid JSON object return a tuple of values corresponding to those keys (use None for missing keys).
- Preserve nested JSON objects as Python dicts in the output.
This function is useful for streaming logs or data pipelines where some lines may be corrupted and you still want to salvage valid records.
Task
Implement a robust parser for JSON Lines that skips blank or malformed lines and can extract specified keys, returning consistent Python types.
Examples
Extract specific keys
Input
parse_json_lines('{"a":1,"b":2}\n{"a":3}', ['a','b'])
Output
[(1, 2), (3, None)]
When keys are provided, return tuples of values; missing keys yield None.
Input format
A single string argument text containing newline-separated JSON objects. Optional second argument keys as a list of strings or None.
Output format
A Python list of dicts (if keys is None) or a Python list of tuples (if keys provided).
Constraints
- Do not raise on malformed JSON lines; skip them.
- Use only Python standard library (json is allowed).
- Keep memory reasonable: assume input fits in memory for this exercise.
Samples
Sample 1
Input
parse_json_lines('{"x": {"y": 5}, "z": 1}\n{"x": {"y": 6}}', ['x','z'])
Output
[({'y': 5}, 1), ({'y': 6}, None)]
Nested dicts are preserved and missing keys become None.