asciichem (Python)

Python implementation of AsciiChem — parse ASCII chemistry into a semantic model (atoms, bonds, isotopes, charges, reactions, electron configurations), then render canonical text or the canonical wire JSON. The defining fix over AsciiMath: ^14C binds the isotope to the atom, never to a phantom {} carrier.

Strategy A of the ecosystem's language-binding ADR: a native implementation conforming to the shared contracts — asciichem-model (v1 wire form) and asciichem-tests (the corpus every implementation passes). The Ruby gem is the reference; this package, like the TypeScript one, must agree with it on every corpus case. Zero runtime dependencies.

Usage

from asciichem import parse_text
from asciichem.text import render
from asciichem.wire import to_model_json, from_model_json

from asciichem.smiles import parse_smiles, write_smiles

formula = parse_text("2H_2 + O_2 ->[heat] 2H_2O")
aspirin = parse_smiles("CC(=O)OC1=CC=CC=C1C(=O)O")  # same model, same renderers
write_smiles(aspirin)  # deterministic writer; round-trips exactly
render(formula)          # canonical AsciiChem text (round-trip contract)
to_model_json(formula)   # asciichem-model v1 wire JSON
back = from_model_json(to_model_json(formula))

Conformance claim

Verified in CI against the asciichem-tests corpus and asciichem-model schemas (both cloned as siblings):

Development

pip install -e ".[dev]"
python -m pytest           # unit + corpus conformance (siblings cloned)

License

BSD-2-Clause — same as the rest of the AsciiChem ecosystem.