linearscript 3.1.0

SCRIPT V3.1: A deterministic, RDKit-independent molecular notation with 100% round-trip stereo parity, materials science extensions, biopolymer support, and formal LALR grammar.

SCRIPT: Structural Chemical Representation In Plain Text

SCRIPT (linearscript on PyPI) is a deterministic molecular notation system with an RDKit-independent core engine. Every molecule has exactly one canonical SCRIPT string. No ambiguity. No post-hoc sanitization.

Aspirin in SMILES:  CC(=O)Oc1ccccc1C(=O)O  (one of many valid forms)
Aspirin in SCRIPT:  CC(=O)OC:C:C:C:C:C&6:C(=O)O  (always and only this)

Install

# Core (no RDKit required)
pip install linearscript

# With RDKit bridge for SMILES interop
pip install linearscript[rdkit]

---

Key Features

Feature	SMILES	SCRIPT
Canonical	No	Yes (DFS + Morgan)
Human-readable	Yes	Yes
Validation on parse	No	Yes (Sandhi state machine)
Organometallics	Partial	Full (dative, haptic, coordinate)
Alloys / Fractional occupancy	No	Yes (<~0.9>)
Crystallographic context	No	Yes ([[Rutile]])
Surface chemistry	No	Yes (`
Electronic / excited states	No	Yes (<s:3>, <*>)
Biopolymers (peptide / nucleic)	No	Yes ({A.G.S})
Query atoms (SMARTS-style)	No	Yes ([#6], [R], [v3])
Polymers / stochastic chains	No	Yes ({[CC]}n)
Reactions	Partial	Yes (3-part R>A>P)
RDKit-free core	No	Yes

---

Quick Start

Parse and canonicalize

from script.parser import SCRIPTParser
from script.canonical import SCRIPTCanonicalizer

parser = SCRIPTParser()
result = parser.parse("CC(=O)Oc1ccccc1C(=O)O")  # aspirin from SMILES-style input

mol = result["molecule"]
print(len(mol.atoms))   # 13
print(len(mol.bonds))   # 13

canon = SCRIPTCanonicalizer().canonicalize_core(mol)
print(canon)   # CC(=O)OC:C:C:C:C:C&6:C(=O)O

Stereochemistry

result = parser.parse("C[C@H](O)C(=O)O")   # L-Lactic acid
mol = result["molecule"]
# Chirality stored in mol.chiral_centers — DFS-invariant, CIP-verified

Reactions

result = parser.parse("[C:1]OCO>>[C:1]O")  # reaction with atom mapping
rxn = result["molecule"]                    # Reaction object
print(rxn.reactants, rxn.products)

Materials Science

# Alloy with fractional site occupancy
result = parser.parse("Ti<~0.9>N<~0.1>")
mol = result["molecule"]
print(mol.atoms[0].occupancy)   # 0.9

# Crystallographic phase
result = parser.parse("[[Rutile]] Ti(O)2")
print(result["molecule"].macroscopic_context)   # "Rutile"

# Surface adsorption
result = parser.parse("[[Pt_111]] | >C=O")
print(result["success"])   # True

# Triplet oxygen
result = parser.parse("O=O<s:3>")
print(result["molecule"].atoms[-1].spin)   # 3

Biopolymers

# Peptide chain (expands to atomic graph)
result = parser.parse("{A.G.S}")

# DNA oligonucleotide
result = parser.parse("{dA.dG.dC.dT}")

# Nucleotide modifications
result = parser.parse("{m5C.m6A.psU}")

RDKit interop

from rdkit import Chem
from script.rdkit_bridge import SCRIPTFromMol, MolFromSCRIPT

# SMILES -> SCRIPT
mol = Chem.MolFromSmiles("CN1CCC[C@H]1c2cccnc2")   # Nicotine
script_str = SCRIPTFromMol(mol)

# SCRIPT -> RDKit mol (100% InChI parity verified)
mol_back = MolFromSCRIPT(script_str)

---

Benchmark

Tested on a diverse 97-compound set (alkanes, rings, aromatics, stereocenters, drugs, natural products):

• 100% InChI round-trip parity (SCRIPT -> RDKit -> InChI matches original)
• 100% native round-trip (SCRIPT -> CoreMolecule -> SCRIPT)
• 22/22 Materials Science tests passing (Alloys, Surfaces, Excited States)

---

License

MIT with Commons Clause. Free for academic and non-commercial use. Commercial licensing available separately.

---

Developed by SCRIPT Development Team. GitHub: script-notation/script