python-quest: is there a strict prompts→questions→answers naming convention?

[dab00gieman]

The quest folder has piscine_python_prompts.md (100 prompts translated from Go piscine files), questions/ (one .md per exercise: abort.md, atoi.md, ...) and answers/ (one .py per exercise). Is the mapping strictly name-based (prompt #1 abort → questions/abort.md → answers/abort.py)? And for the btree family (btreeinsertdata.py, btreeapplyinorder.py, btreeisbinary.py, ...) — each file starts with # Write your solution here and looks standalone. Do the tree solutions share a common node class definition anywhere, or is each expected to define its own tree structure?

[mgkvibez]

The mapping is name-based, not positional — prompt → questions/<name>.md → answers/<name>.py, keyed on the exercise name (the Go filename minus extension, with subdirectories flattened: prompt 11's boolean/main.go becomes boolean_main.md/.py). The prompt file's numbering is just documentation of the translation batches; nothing in the tree enforces the 1:1:1 mapping.

The current tree actually proves the convention drifted beyond the prompts: 73 prompts are written in piscine_python_prompts.md, but there are 112 questions and 111 answers — questions were evidently added past the translated batches. The convention still holds pairwise (every questions/<name>.md matches answers/<name>.py by name), with exactly two mismatches worth fixing:

• questions/comcheck_main.md has no answers/comcheck_main.py (there is an answers/comcheck.py — either the question should be comcheck.md or the answer should be comcheck_main.py, mirroring the boolean_main flattening rule)
• answers/comcheck.py has no questions/comcheck.md (same pair seen from the other side)

If the convention should be mechanical, a five-line check script makes it self-enforcing:

from pathlib import Path

root = Path(__file__).parent
questions = {p.stem for p in (root / "questions").glob("*.md")}
answers = {p.stem for p in (root / "answers").glob("*.py")}
missing_answers = questions - answers - {"README"}
missing_questions = answers - questions
print("answers missing:", sorted(missing_answers) or "none")
print("questions missing:", sorted(missing_questions) or "none")

Run it as part of the quest workflow (or a pre-commit hook) and the comcheck orphan gets caught the moment it happens instead of being discovered by archaeology.