Updated: December 22, 2025
Python helps make a lot of things really easy. The conciseness of the language make it awesome to accomplish simple tasks in as few lines as possible. One of these simple tasks is checking to see if a file exists. read on »
Updated: December 22, 2025
When debugging and profiling Python software you'll find yourself relying on Python's built-in logging module. Something that is very simple to overlook is the fact that the root logger defaults to WARN and not DEBUG. read on »
Updated: December 22, 2025
When using Python for Data Science or general Programming you'll find yourself needing to read and parse very very very large files. The easiest way to accomplish this is by iterating over the actual file object. read on »
Updated: September 17, 2026
Two ways to crawl a directory tree in Python: `os.walk()` and `pathlib`. Both work, but `os.walk()` has a classic trap-`dirnames` and `filenames` are names, not full paths-so one missing `os.path.join()` can send your script wandering into the wrong folder. This post shows safe patterns for each approach, how to prune noisy dirs like `node_modules`, and why `Path(...).rglob()` is usually the "hard to break" choice-especially once you start hunting down those mysterious duplicate CSVs. read on »
Updated: March 18, 2026
Learn how to organize a Python Secret Santa gift exchange using pyge library. Optimize gift matches, set exclusions, and automate your holiday gift exchange with this complete guide. read on »
Updated: September 17, 2026
That "Decimal is not JSON serializable" crash isn't random. JSON only supports a tiny set of types, and Python refuses to guess conversions that can quietly corrupt data (hello, billing). The fix isn't `default=str`. It's picking explicit encoding rules once-Decimal to string, datetime to ISO-8601, UUID to string, Enum to `.value`, dataclasses to `asdict()`-and using them everywhere with `json.dumps(..., default=...)`. read on »
Updated: September 17, 2026
I kept re-learning Python slices because my brain refused to accept "stop is exclusive." Then I started doing LLM work and suddenly slices were everywhere: token windows, chunking, sliding overlap, all of it. This post nails the only three rules you need (`start:stop:step`), the five slice patterns I actually use, and why "exclusive stop" is the reason chunking doesn’t turn into off-by-one misery. read on »
