Module manubot.cite.curie
Compact Uniform Resource Identifiers
Manubot keeps a local versions of the Bioregistry. Repository developers can run the following commands to update the Manubot version.
# regenerate manubot/cite/curie/bioregistry.json
python manubot/cite/curie/__init__.py
# if bioregistry.json has changed, the following test will likely fail:
pytest manubot/cite/tests/test_handlers.py::test_prefix_to_handler
# copy captured stdout from failed test_prefix_to_handler to
# manubot.cite.handlers.prefix_to_handler. Pre-commit hook will reformat file.
References:
- https://bioregistry.io/
- https://github.com/biopragmatics/bioregistry
- https://en.wikipedia.org/wiki/CURIE
- https://cthoyt.com/2021/10/07/biopragmatics-glossary.html
- https://github.com/manubot/manubot/issues/305
- https://identifiers.org/
- https://github.com/manubot/manubot/issues/218
- https://docs.identifiers.org/articles/api.html
- https://n2t.net/e/compact_ids.html
- https://n2t.net/e/cdl_ebi_prefixes.yaml
- https://en.wikipedia.org/wiki/MIRIAM_Registry
- Identifiers.org and MIRIAM Registry: community resources to provide persistent identification
- On the road to robust data citation
- Uniform Resolution of Compact Identifiers for Biomedical Data
View Source
"""
Compact Uniform Resource Identifiers
Manubot keeps a local versions of the Bioregistry.
Repository developers can run the following commands to update the Manubot version.
```shell
# regenerate manubot/cite/curie/bioregistry.json
python manubot/cite/curie/__init__.py
# if bioregistry.json has changed, the following test will likely fail:
pytest manubot/cite/tests/test_handlers.py::test_prefix_to_handler
# copy captured stdout from failed test_prefix_to_handler to
# manubot.cite.handlers.prefix_to_handler. Pre-commit hook will reformat file.
```
References:
- https://bioregistry.io/
- https://github.com/biopragmatics/bioregistry
- https://en.wikipedia.org/wiki/CURIE
- https://cthoyt.com/2021/10/07/biopragmatics-glossary.html
- https://github.com/manubot/manubot/issues/305
- https://identifiers.org/
- https://github.com/manubot/manubot/issues/218
- https://docs.identifiers.org/articles/api.html
- https://n2t.net/e/compact_ids.html
- https://n2t.net/e/cdl_ebi_prefixes.yaml
- https://en.wikipedia.org/wiki/MIRIAM_Registry
- [Identifiers.org and MIRIAM Registry: community resources to provide persistent identification](https://doi.org/10.1093/nar/gkr1097)
- [On the road to robust data citation](https://doi.org/10.1038/sdata.2018.95)
- [Uniform Resolution of Compact Identifiers for Biomedical Data](https://doi.org/10.1038/sdata.2018.29)
"""
import dataclasses
import functools
import json
import pathlib
import re
import typing
from manubot.cite.citekey import CiteKey
from manubot.cite.handlers import Handler
_keep_bioregistry_fields = {
"deprecated",
"example",
"uri_format",
"name",
"pattern",
"preferred_prefix",
"synonyms",
}
bioregistry_path = pathlib.Path(__file__).parent.joinpath("bioregistry.json")
valid_prefix_pattern = re.compile(r"^[a-z0-9][a-z0-9._-]+?$")
"""
Ignore Bioregistry prefixes/synonyms that do not adhere to this pattern.
More permissive than the pattern at <https://github.com/biopragmatics/bioregistry/issues/158>,
because there are existing Bioregistry prefixes, even preferred prefixes, that are bad.
For example, starting with a number. We primarily care whether the prefix will work as part
of a pandoc citation key without requiring escaping.
"""
@dataclasses.dataclass
class Handler_CURIE(Handler):
def __post_init__(self):
try:
self.resource = get_prefix_to_resource()[self.prefix_lower]
except KeyError as err:
raise ValueError(f"Unrecognized CURIE prefix {self.prefix_lower}") from err
self.standard_prefix = (
self.resource.get("preferred_prefix") or self.resource["prefix"]
)
self.prefixes = self.resource["all_prefixes"]
if "pattern" in self.resource:
self.accession_pattern = self.resource["pattern"]
def get_csl_item(self, citekey: CiteKey):
from ..url import get_url_csl_item
url = self.get_url(accession=citekey.standard_accession)
return get_url_csl_item(url)
def inspect(self, citekey: CiteKey) -> typing.Optional[str]:
pattern = self._get_pattern("accession_pattern")
if pattern and not pattern.fullmatch(citekey.accession):
return f"{citekey.accession} does not match regex {pattern.pattern}"
def get_url(self, accession: str) -> str:
if "uri_format" in self.resource:
return self.resource["uri_format"].replace("$1", accession)
return f"https://bioregistry.io/{self.standard_prefix}:{accession}"
def get_curie_handlers():
"""Get all possible CURIE handlers"""
registries = get_bioregistry(compile_patterns=True)
handlers = [Handler_CURIE(reg["prefix"]) for reg in registries]
return handlers
def _download_bioregistry() -> None:
"""
Download the Bioregistry consensus registry adding the following fields for each registry:
- prefix: the standard lowercase registry prefix
- all_prefixes: all distinct valid lowercase prefixes including synonyms
"""
import requests
url = "https://github.com/biopragmatics/bioregistry/raw/main/exports/registry/registry.json"
response = requests.get(url)
response.raise_for_status()
results = response.json()
assert isinstance(results, dict)
registry = []
for prefix, resource in results.items():
assert isinstance(resource, dict)
if not resource.get("uri_format"):
# discard unresolvable prefixes
continue
for field in set(resource) - _keep_bioregistry_fields:
del resource[field]
resource["prefix"] = prefix
all_prefixes = {
prefix,
*(x.lower() for x in resource.pop("synonyms", [])),
}
# remove invalid prefixes as per https://github.com/manubot/manubot/pull/306#discussion_r744125504
resource["all_prefixes"] = sorted(
filter(valid_prefix_pattern.fullmatch, all_prefixes)
)
registry.append(resource)
json_text = json.dumps(registry, indent=2, ensure_ascii=False)
bioregistry_path.write_text(json_text + "\n", encoding="utf-8")
def get_bioregistry(compile_patterns=False) -> dict:
with bioregistry_path.open(encoding="utf-8-sig") as read_file:
registry = json.load(read_file)
assert isinstance(registry, list)
if compile_patterns:
for resource in registry:
if "pattern" in resource:
resource["compiled_pattern"] = re.compile(resource["pattern"])
return registry
@functools.lru_cache
def get_prefix_to_resource() -> typing.Dict[str, typing.Dict]:
prefix_to_resource = {}
for resource in get_bioregistry():
for prefix in resource["all_prefixes"]:
prefix_to_resource[prefix] = resource
return prefix_to_resource
def standardize_curie(curie: str) -> str:
"""
Return CURIE with Bioregistry preferred prefix capitalization.
`curie` should be in `prefix:accession` format.
If `curie` is malformed or uses an unrecognized prefix, raise ValueError.
"""
if not isinstance(curie, str):
raise TypeError(
f"curie parameter should be string. Received {curie.__class__.__name__} instead for {curie}"
)
try:
prefix, accession = curie.split(":", 1)
except ValueError as err:
raise ValueError(
f"curie must be splittable by `:` and formatted like `prefix:accession`. Received {curie}"
) from err
handler = Handler_CURIE(prefix.lower())
return f"{handler.standard_prefix}:{accession}"
def curie_to_url(curie: str) -> str:
"""
`curie` should be in `prefix:accession` format
"""
curie = standardize_curie(curie)
prefix, accession = curie.split(":", 1)
handler = Handler_CURIE(prefix.lower())
return handler.get_url(accession)
if __name__ == "__main__":
_download_bioregistry()
bioregistry = get_bioregistry()
Variables
bioregistry_path
valid_prefix_pattern
Ignore Bioregistry prefixes/synonyms that do not adhere to this pattern.
More permissive than the pattern at https://github.com/biopragmatics/bioregistry/issues/158, because there are existing Bioregistry prefixes, even preferred prefixes, that are bad. For example, starting with a number. We primarily care whether the prefix will work as part of a pandoc citation key without requiring escaping.
Functions
curie_to_url
def curie_to_url(
curie: str
) -> str
curie should be in prefix:accession format
View Source
def curie_to_url(curie: str) -> str:
"""
`curie` should be in `prefix:accession` format
"""
curie = standardize_curie(curie)
prefix, accession = curie.split(":", 1)
handler = Handler_CURIE(prefix.lower())
return handler.get_url(accession)
get_bioregistry
def get_bioregistry(
compile_patterns=False
) -> dict
View Source
def get_bioregistry(compile_patterns=False) -> dict:
with bioregistry_path.open(encoding="utf-8-sig") as read_file:
registry = json.load(read_file)
assert isinstance(registry, list)
if compile_patterns:
for resource in registry:
if "pattern" in resource:
resource["compiled_pattern"] = re.compile(resource["pattern"])
return registry
get_curie_handlers
def get_curie_handlers(
)
Get all possible CURIE handlers
View Source
def get_curie_handlers():
"""Get all possible CURIE handlers"""
registries = get_bioregistry(compile_patterns=True)
handlers = [Handler_CURIE(reg["prefix"]) for reg in registries]
return handlers
get_prefix_to_resource
def get_prefix_to_resource(
) -> Dict[str, Dict]
View Source
@functools.lru_cache
def get_prefix_to_resource() -> typing.Dict[str, typing.Dict]:
prefix_to_resource = {}
for resource in get_bioregistry():
for prefix in resource["all_prefixes"]:
prefix_to_resource[prefix] = resource
return prefix_to_resource
standardize_curie
def standardize_curie(
curie: str
) -> str
Return CURIE with Bioregistry preferred prefix capitalization.
curie should be in prefix:accession format.
If curie is malformed or uses an unrecognized prefix, raise ValueError.
View Source
def standardize_curie(curie: str) -> str:
"""
Return CURIE with Bioregistry preferred prefix capitalization.
`curie` should be in `prefix:accession` format.
If `curie` is malformed or uses an unrecognized prefix, raise ValueError.
"""
if not isinstance(curie, str):
raise TypeError(
f"curie parameter should be string. Received {curie.__class__.__name__} instead for {curie}"
)
try:
prefix, accession = curie.split(":", 1)
except ValueError as err:
raise ValueError(
f"curie must be splittable by `:` and formatted like `prefix:accession`. Received {curie}"
) from err
handler = Handler_CURIE(prefix.lower())
return f"{handler.standard_prefix}:{accession}"
Classes
Handler_CURIE
class Handler_CURIE(
prefix_lower: str
)
Handler_CURIE(prefix_lower: str)
View Source
@dataclasses.dataclass
class Handler_CURIE(Handler):
def __post_init__(self):
try:
self.resource = get_prefix_to_resource()[self.prefix_lower]
except KeyError as err:
raise ValueError(f"Unrecognized CURIE prefix {self.prefix_lower}") from err
self.standard_prefix = (
self.resource.get("preferred_prefix") or self.resource["prefix"]
)
self.prefixes = self.resource["all_prefixes"]
if "pattern" in self.resource:
self.accession_pattern = self.resource["pattern"]
def get_csl_item(self, citekey: CiteKey):
from ..url import get_url_csl_item
url = self.get_url(accession=citekey.standard_accession)
return get_url_csl_item(url)
def inspect(self, citekey: CiteKey) -> typing.Optional[str]:
pattern = self._get_pattern("accession_pattern")
if pattern and not pattern.fullmatch(citekey.accession):
return f"{citekey.accession} does not match regex {pattern.pattern}"
def get_url(self, accession: str) -> str:
if "uri_format" in self.resource:
return self.resource["uri_format"].replace("$1", accession)
return f"https://bioregistry.io/{self.standard_prefix}:{accession}"
Ancestors (in MRO)
- manubot.cite.handlers.Handler
Class variables
prefixes
Methods
get_csl_item
def get_csl_item(
self,
citekey: manubot.cite.citekey.CiteKey
)
Return a CSL_Item with bibliographic details for citekey.
View Source
def get_csl_item(self, citekey: CiteKey):
from ..url import get_url_csl_item
url = self.get_url(accession=citekey.standard_accession)
return get_url_csl_item(url)
get_url
def get_url(
self,
accession: str
) -> str
View Source
def get_url(self, accession: str) -> str:
if "uri_format" in self.resource:
return self.resource["uri_format"].replace("$1", accession)
return f"https://bioregistry.io/{self.standard_prefix}:{accession}"
inspect
def inspect(
self,
citekey: manubot.cite.citekey.CiteKey
) -> Optional[str]
Check citekeys adhere to expected formats. If an issue is detected a
string describing the issue is returned. Otherwise returns None.
View Source
def inspect(self, citekey: CiteKey) -> typing.Optional[str]:
pattern = self._get_pattern("accession_pattern")
if pattern and not pattern.fullmatch(citekey.accession):
return f"{citekey.accession} does not match regex {pattern.pattern}"
standardize_prefix_accession
def standardize_prefix_accession(
self,
accession: str
) -> Tuple[str, str]
Return (prefix, accession) in standardized form.
This method defaults to returning self.standard_prefix
(or self.prefix_lower if standard_prefix is not defined).
Subclasses can override this method with more specific standardization logic.
View Source
def standardize_prefix_accession(self, accession: str) -> Tuple[str, str]:
"""
Return (prefix, accession) in standardized form.
This method defaults to returning `self.standard_prefix`
(or `self.prefix_lower` if standard_prefix is not defined).
Subclasses can override this method with more specific standardization logic.
"""
standard_prefix = getattr(self, "standard_prefix", self.prefix_lower)
standard_accession = accession
return standard_prefix, standard_accession