Source code for pyked.identifier
"""Validation helpers for chemical species and author identifiers.
Provides InChI/SMILES parsing checks backed by RDKit and ORCID lookups
against the ORCID public API.
"""
import httpx2 as httpx
from rdkit import Chem, rdBase
from rdkit.Chem import inchi
INCHI_PREFIX = "InChI="
headers = {"Accept": "application/json"}
[docs]
def search_orcid(orcid):
"""
Search the ORCID public API
Specfically, return a dictionary with the personal details
(name, etc.) of the person associated with the given ORCID
Args:
orcid (`str`): The ORCID to be searched
Returns:
`dict`: Dictionary with the JSON response from the API
Raises:
`~httpx.HTTPStatusError`: If the given ORCID cannot be found, an
`~httpx.HTTPStatusError` is raised with status code 404
"""
url = f"https://pub.orcid.org/v3.0/{orcid}/person"
r = httpx.get(url, headers=headers)
r.raise_for_status()
return r.json()
"""Validation helpers for chemical species identifiers."""
[docs]
def normalize_inchi(value: str):
"""Return an InChI with the conventional ``InChI=`` prefix.
ChemKED historically permits both complete InChIs and the prefixless form
used by existing PyKED files, such as ``1S/H2/h1H``.
"""
value = value.strip()
if value.startswith(INCHI_PREFIX):
return value
return f"{INCHI_PREFIX}{value}"
[docs]
def valid_inchi(value: str):
"""Return whether *value* can be parsed as an InChI."""
if not isinstance(value, str) or not value.strip():
return False
try:
with rdBase.BlockLogs():
molecule = inchi.MolFromInchi(normalize_inchi(value))
except (inchi.InchiReadWriteError, RuntimeError, ValueError):
return False
return molecule is not None
[docs]
def valid_smiles(value: str):
"""Return whether *value* can be parsed and sanitized as SMILES."""
if not isinstance(value, str) or not value.strip():
return False
try:
with rdBase.BlockLogs():
molecule = Chem.MolFromSmiles(value.strip(), sanitize=True)
except (RuntimeError, ValueError):
return False
return molecule is not None