Reading and writing SMILES¶

Parsing¶

m = omgkit.parse_smiles("O[C@@H](C)c1ccccc1")

Parsing does not sanitize. What you get back is exactly what the string said: the atoms in the order they were written, the bonds as written, aromatic lower-case atoms recorded as aromatic because the author claimed so — not because anything has verified it yet.

Malformed input raises ValueError with a caret view:

>>> omgkit.parse_smiles("CC(C")
Traceback (most recent call last):
  ...
ValueError: CC(C
    ^ 括号不匹配

Two ways to write it back¶

	Depends on	Use it for
`to_smiles()`	the current atom storage order	round-tripping, debugging, seeing what you built
`to_canonical_smiles()`	the structure only	keys, deduplication, comparison

>>> m = omgkit.parse_smiles("OC(=O)c1ccccc1N")
>>> m.to_smiles()
'OC(=O)c1ccccc1N'
>>> m.sanitize(); m.to_canonical_smiles()
'c1cccc(c1C(O)=O)N'

to_canonical_smiles() is the one that keeps the promise: any two writings of the same molecule give the same string.

>>> a = omgkit.parse_smiles("OC(=O)c1ccccc1N"); a.sanitize()
>>> b = omgkit.parse_smiles("Nc1ccccc1C(O)=O"); b.sanitize()
>>> a.to_canonical_smiles() == b.to_canonical_smiles()
True

What survives a round trip¶

Tetrahedral chirality, double-bond geometry, dative bonds, isotopes, formal charges, radical electrons, atom map numbers and explicit hydrogens all parse and write back.

Chirality deserves a note. A tetrahedral tag (@ / @@) is relative to the order the neighbours are stored in. Any operation that changes that order — removing a hydrogen, cutting a bond, reordering for canonical output — has to rebase the tag, or the molecule silently becomes its mirror image. Nothing raises; you just get the wrong enantiomer. This is why the writing layer has its own differential judge for stereochemistry rather than trusting round-trip equality alone.

Explicit hydrogens¶

>>> h = omgkit.parse_smiles("[H]C([H])([H])O"); h.sanitize()
>>> h.num_atoms
5
>>> h.remove_hs()
3
>>> h.num_atoms, h.to_smiles()
(2, 'CO')

remove_hs() folds hydrogens into their neighbour's hydrogen count and returns how many it removed. It modifies in place and every atom index changes — deleting atoms necessarily renumbers.

It deliberately keeps hydrogens that carry information: isotopes, charges, atom map numbers, radicals, bridging hydrogens, and hydrogens that carry a double bond's direction. Keeping one extra hydrogen only costs a node in the graph; removing the wrong one loses information.