examples.md

Interpreter — where it lives outside this repo

Cited, real implementations to study (or point an agent at) when designing or reviewing little-language code.

Python standard library

• ast.literal_eval / ast.NodeVisitor — the safe-evaluation floor: Python parses, you walk only the nodes you allow. This module's safe_eval is the canonical restricted walk. docs.python.org/3/library/ast.html
• re — a complete Interpreter implementation used daily: a pattern language parsed and compiled into a program, evaluated against strings. docs.python.org/3/library/re.html

Major ecosystems

• Django Q objects — query predicates built as composable expression trees (Q(age__gte=18) & Q(country="CA")), interpreted into SQL by the ORM. docs.djangoproject.com/en/stable/topics/db/queries/#complex-lookups-with-q-objects
• SQLAlchemy Core expression language — column expressions form a tree the compiler walks to emit dialect-specific SQL: grammar-as-objects at production scale. docs.sqlalchemy.org/en/latest/core/expression_api.html
• pytest -k expressions — a real shipped mini-language (and/or/ not over test names) with its own tiny parser and evaluator. docs.pytest.org/en/stable/how-to/usage.html#specifying-which-tests-to-run (unverified)
• json-logic — rules-as-JSON evaluated by a small interpreter; the same shape as this unit's flag engine, standardized across languages. jsonlogic.com (unverified)

What to notice across all of them

None of them expose a general-purpose evaluator to user input. Each fixes a closed set of operations (Django's lookups, pytest's three combinators) and validates sentences structurally before evaluating — the two guards (ValueError on unknown operations, bounded depth) that this module treats as part of the pattern, not optional hardening.