Python Standard Library

File paths, directory operations, environment variables, process info. The bridge between Python and the OS.

Command-line arguments (argv), stdin/stdout/stderr, sys.path, sys.modules, interpreter version. Direct access to the Python interpreter.

Object-oriented file paths — join with /, read/write directly, glob/rglob. The modern replacement for os.path string manipulation.

Encode Python to JSON and decode JSON to Python. Custom encoders for datetimes, Decimal, and any object.

Regular expressions — search, match, findall, sub, groups, lookaheads. Find patterns in any text.

Date and time objects, timedelta, strftime/strptime, timezone-aware datetimes, and zoneinfo for the IANA database.

Counter, defaultdict, namedtuple, deque, OrderedDict, ChainMap — specialised containers that go beyond list and dict.

chain, product, combinations, groupby, accumulate — the iterator algebra toolkit for memory-efficient sequence operations.

Async I/O, event loops, coroutines, tasks, and streams. The framework for concurrent Python code.

Test cases, assertions, mocking, and test runners. Python's built-in testing framework.

Structured logging with levels, handlers, formatters, and propagation. The right way to record what your program does.

lru_cache and cache for memoisation, partial, reduce, singledispatch, cached_property, total_ordering.

Read and write CSV correctly — reader, writer, DictReader, DictWriter, dialects, and the newline="" rule.

Run external programs — subprocess.run(), capture output, pipelines, timeouts, Popen, injection safety.

Command-line argument parsing — parsers, positional/optional args, types, choices, subcommands.

Mathematical functions — constants, sqrt, rounding, trig, logs, isclose for float safety.

Pseudo-random numbers — randint, choice, shuffle, sample, distributions, seeding (not for security).

with-statement utilities — @contextmanager, suppress, redirect_stdout, ExitStack.

Cryptographically secure random — tokens, passwords, constant-time compare. Use for all security randomness.

Secure hashes — SHA-256, file integrity, password hashing with pbkdf2/scrypt, HMAC, BLAKE2.

Statistics without NumPy — mean, median, stdev, NormalDist, quantiles, correlation, regression.

Object serialization — save any Python object to bytes. Powerful but unsafe with untrusted data.

Built-in SQL database — zero setup, one file. Connect, query, parameters, transactions, Row factory.

System clock, sleep, and performance timing — timestamps, perf_counter, monotonic, nanosecond clocks.

High-level file operations — copy/move directory trees, rmtree, create and extract archives, disk usage.

Shallow and deep copy — the difference that causes Python’s most confusing bugs. copy() vs deepcopy().

High-level parallelism — run functions across threads or processes with one uniform API. Futures, map, submit.

Thread-safe queues — pass work between threads safely. FIFO, LIFO, priority, producer-consumer, join.

Low-level networking — the foundation beneath every HTTP library. TCP/UDP clients and servers.

Built-in HTTP server — instant file sharing and custom handlers. For development and learning, not production.