ReShiki for Python

Draw, convert and analyze chemical structures with the same Rust engine as the ReShiki editor. Python 3.10 or later is required. The API is experimental until 1.0.

uv add reshiki  # or: pip install reshiki

import reshiki

aspirin = reshiki.Molecule.from_smiles("CC(=O)Oc1ccccc1C(=O)O")
figure = aspirin.clean().with_style("JACS / ACS", font_size=12)
figure.save("aspirin.png", width=300)
print(aspirin.properties().formula)

Transforms return new objects. The original molecule keeps its geometry and style. In Jupyter, displaying an object shows its SVG drawing.

Spreadsheets

Install the Excel integration with uv add 'reshiki[excel]', then add drawings beside a column of SMILES:

import reshiki

reshiki.excel.add_structures(
    "compounds.xlsx",
    smiles="A",
    into="B",
    out="compounds-drawn.xlsx",
    width=300,
    header=True,
)

The helper uses openpyxl and embeds PNG images. Editable Office pictures are a follow-up to issue #462. Use header=False for a sheet whose first row contains a structure, or start_row= to choose an explicit first row.

Other entry points

crystal = reshiki.Drawing.open("rutile.vasp")
svg = crystal.svg()
png = reshiki.render("c1ccccc1O", format="png", width=300)
mol = reshiki.convert("CCO", to="mol")
properties = reshiki.properties("CCO")
pngs = reshiki.render_many(["CCO", "C1CC"], format="png", width=300)
print(pngs.errors)

reshiki.parse(text, format=None) detects chemical text, native drawings and crystal structures. Explicit constructors include from_smiles, from_molfile, from_inchi, from_cdxml and offline from_name. An online name lookup raises NotImplementedError: the current PubChem client belongs to the desktop app.

PNG dimensions follow the drawing style's physical bond length and dpi by default. width and height optionally fit the image into a pixel box without changing its aspect ratio. reshiki.styles() lists the app's drawing presets; reshiki.Style(name, **overrides) makes a reusable style.

Install reshiki[pandas] and explicitly import reshiki.pandas to register the pandas Series .reshiki accessor with png(), svg() and properties(). Optional dependencies load when their integration is used. RDKit conversion uses Molecule.from_rdkit(mol) and molecule.to_rdkit() when RDKit is already installed.

Outputs include PNG, SVG, PDF, CDXML, MOL/SDF, SMILES, InChI and native .rsk. EMF raises NotImplementedError pending a headless exporter. Wheels for macOS, Windows and Linux on x64 and ARM64 follow each ReShiki release.

See the Python guide for the API and local build commands.