pyobo 0.11.2__py3-none-any.whl → 0.12.0__py3-none-any.whl

pyobo/getters.py

@@ -1,38 +1,48 @@
 """Utilities for OBO files."""
 
+from __future__ import annotations
+
 import datetime
 import gzip
 import json
 import logging
 import pathlib
 import subprocess
+import time
 import typing
 import urllib.error
+import zipfile
 from collections import Counter
-from collections.abc import Iterable, Mapping, Sequence
+from collections.abc import Callable, Iterable, Mapping, Sequence
 from pathlib import Path
-from typing import (
-    Callable,
-    Optional,
-    TypeVar,
-    Union,
-)
+from textwrap import indent
+from typing import TypeVar
 
 import bioregistry
+import click
+import pystow.utils
 from bioontologies import robot
+from tabulate import tabulate
 from tqdm.auto import tqdm
+from typing_extensions import Unpack
 
-from .constants import DATABASE_DIRECTORY
-from .identifier_utils import MissingPrefixError, wrap_norm_prefix
+from .constants import (
+    DATABASE_DIRECTORY,
+    GetOntologyKwargs,
+    IterHelperHelperDict,
+    SlimGetOntologyKwargs,
+)
+from .identifier_utils import ParseError, wrap_norm_prefix
 from .plugins import has_nomenclature_plugin, run_nomenclature_plugin
+from .reader import from_obo_path, from_obonet
 from .struct import Obo
 from .utils.io import get_writer
 from .utils.path import ensure_path, prefix_directory_join
 from .version import get_git_hash, get_version
 
 __all__ = [
-    "get_ontology",
     "NoBuildError",
+    "get_ontology",
 ]
 
 logger = logging.getLogger(__name__)
@@ -48,7 +58,14 @@ class UnhandledFormatError(NoBuildError):
 
 #: The following prefixes can not be loaded through ROBOT without
 #: turning off integrity checks
-REQUIRES_NO_ROBOT_CHECK = {"clo", "vo", "orphanet.ordo", "orphanet"}
+REQUIRES_NO_ROBOT_CHECK = {
+    "clo",
+    "vo",
+    "orphanet.ordo",
+    "orphanet",
+    "foodon",
+    "caloha",
+}
 
 
 @wrap_norm_prefix
@@ -56,58 +73,85 @@ def get_ontology(
     prefix: str,
     *,
     force: bool = False,
-    rewrite: bool = False,
-    strict: bool = True,
-    version: Optional[str] = None,
+    force_process: bool = False,
+    strict: bool = False,
+    version: str | None = None,
     robot_check: bool = True,
+    upgrade: bool = True,
+    cache: bool = True,
+    use_tqdm: bool = True,
 ) -> Obo:
     """Get the OBO for a given graph.
 
     :param prefix: The prefix of the ontology to look up
     :param version: The pre-looked-up version of the ontology
     :param force: Download the data again
-    :param rewrite: Should the OBO cache be rewritten? Automatically set to true if ``force`` is true
-    :param strict: Should CURIEs be treated strictly? If true, raises exceptions on invalid/malformed
-    :param robot_check:
-        If set to false, will send the ``--check=false`` command to ROBOT to disregard
-        malformed ontology components. Necessary to load some ontologies like VO.
+    :param force_process: Should the OBO cache be rewritten? Automatically set to true
+        if ``force`` is true
+    :param strict: Should CURIEs be treated strictly? If true, raises exceptions on
+        invalid/malformed
+    :param robot_check: If set to false, will send the ``--check=false`` command to
+        ROBOT to disregard malformed ontology components. Necessary to load some
+        ontologies like VO.
+    :param upgrade: If set to true, will automatically upgrade relationships, such as
+        ``obo:chebi#part_of`` to ``BFO:0000051``
+    :param cache: Should cached objects be written? defaults to True
+
     :returns: An OBO object
 
     :raises OnlyOWLError: If the OBO foundry only has an OWL document for this resource.
 
-    Alternate usage if you have a custom url::
+    Alternate usage if you have a custom url
+
+    .. code-block:: python
 
-    >>> from pystow.utils import download
-    >>> from pyobo import Obo, from_obo_path
-    >>> url = ...
-    >>> obo_path = ...
-    >>> download(url=url, path=path)
-    >>> obo = from_obo_path(path)
+        from pystow.utils import download
+        from pyobo import Obo, from_obo_path
+
+        url = ...
+        obo_path = ...
+        download(url=url, path=path)
+        obo = from_obo_path(path)
     """
     if force:
-        rewrite = True
+        force_process = True
     if prefix == "uberon":
         logger.info("UBERON has so much garbage in it that defaulting to non-strict parsing")
         strict = False
 
-    obonet_json_gz_path = prefix_directory_join(
-        prefix, name=f"{prefix}.obonet.json.gz", ensure_exists=False, version=version
-    )
-    if obonet_json_gz_path.exists() and not force:
-        from .reader import from_obonet
-        from .utils.cache import get_gzipped_graph
-
-        logger.debug("[%s] using obonet cache at %s", prefix, obonet_json_gz_path)
-        return from_obonet(get_gzipped_graph(obonet_json_gz_path))
+    if not cache:
+        logger.debug("[%s] caching was turned off, so dont look for an obonet file", prefix)
+        obonet_json_gz_path = None
+    else:
+        obonet_json_gz_path = prefix_directory_join(
+            prefix, name=f"{prefix}.obonet.json.gz", ensure_exists=False, version=version
+        )
+        logger.debug(
+            "[%s] caching is turned on, so look for an obonet file at %s",
+            prefix,
+            obonet_json_gz_path,
+        )
+        if obonet_json_gz_path.exists() and not force:
+            from .utils.cache import get_gzipped_graph
+
+            logger.debug("[%s] using obonet cache at %s", prefix, obonet_json_gz_path)
+            return from_obonet(
+                get_gzipped_graph(obonet_json_gz_path),
+                strict=strict,
+                version=version,
+                upgrade=upgrade,
+                use_tqdm=use_tqdm,
+            )
+        else:
+            logger.debug("[%s] no obonet cache found at %s", prefix, obonet_json_gz_path)
 
     if has_nomenclature_plugin(prefix):
         obo = run_nomenclature_plugin(prefix, version=version)
-        logger.debug("[%s] caching nomenclature plugin", prefix)
-        obo.write_default(force=rewrite)
+        if cache:
+            logger.debug("[%s] caching nomenclature plugin", prefix)
+            obo.write_default(force=force_process)
         return obo
 
-    logger.debug("[%s] no obonet cache found at %s", prefix, obonet_json_gz_path)
-
     ontology_format, path = _ensure_ontology_path(prefix, force=force, version=version)
     if path is None:
         raise NoBuildError(prefix)
@@ -122,25 +166,23 @@ def get_ontology(
     else:
         raise UnhandledFormatError(f"[{prefix}] unhandled ontology file format: {path.suffix}")
 
-    from .reader import from_obo_path
-
-    obo = from_obo_path(path, prefix=prefix, strict=strict)
-    if version is not None:
-        if obo.data_version is None:
-            logger.warning("[%s] did not have a version, overriding with %s", obo.ontology, version)
-            obo.data_version = version
-        elif obo.data_version != version:
-            logger.warning(
-                "[%s] had version %s, overriding with %s", obo.ontology, obo.data_version, version
-            )
-            obo.data_version = version
-    obo.write_default(force=rewrite)
+    obo = from_obo_path(
+        path,
+        prefix=prefix,
+        strict=strict,
+        version=version,
+        upgrade=upgrade,
+        use_tqdm=use_tqdm,
+        _cache_path=obonet_json_gz_path,
+    )
+    if cache:
+        obo.write_default(force=force_process)
     return obo
 
 
 def _ensure_ontology_path(
-    prefix: str, force, version
-) -> Union[tuple[str, Path], tuple[None, None]]:
+    prefix: str, force: bool, version: str | None
+) -> tuple[str, Path] | tuple[None, None]:
     for ontology_format, url in [
         ("obo", bioregistry.get_obo_download(prefix)),
         ("owl", bioregistry.get_owl_download(prefix)),
@@ -148,8 +190,8 @@ def _ensure_ontology_path(
     ]:
         if url is not None:
             try:
-                path = Path(ensure_path(prefix, url=url, force=force, version=version))
-            except urllib.error.HTTPError:
+                path = ensure_path(prefix, url=url, force=force, version=version)
+            except (urllib.error.HTTPError, pystow.utils.DownloadError):
                 continue
             else:
                 return ontology_format, path
@@ -215,33 +257,42 @@ CANT_PARSE = {
     "xl",
 }
 SKIP = {
-    "ncbigene",  # too big, refs acquired from other dbs
-    "pubchem.compound",  # to big, can't deal with this now
-    "gaz",  # Gazetteer is irrelevant for biology
-    "ma",  # yanked
-    "bila",  # yanked
-    # FIXME below
-    "emapa",  # recently changed with EMAP... not sure what the difference is anymore
-    "kegg.genes",
-    "kegg.genome",
-    "kegg.pathway",
-    # URL is wrong
-    "ensemblglossary",
-    # Too much junk
-    "biolink",
+    "ncbigene": "too big, refs acquired from other dbs",
+    "pubchem.compound": "top big, can't deal with this now",
+    "gaz": "Gazetteer is irrelevant for biology",
+    "ma": "yanked",
+    "bila": "yanked",
+    # Can't download",
+    "afpo": "unable to download",
+    "atol": "unable to download",
+    "eol": "unable to download, same source as atol",
+    "hog": "unable to download",
+    "vhog": "unable to download",
+    "gorel": "unable to download",
+    "dinto": "unable to download",
+    "gainesville.core": "unable to download",
+    "ato": "can't process",
+    "emapa": "recently changed with EMAP... not sure what the difference is anymore",
+    "kegg.genes": "needs fix",  # FIXME
+    "kegg.genome": "needs fix",  # FIXME
+    "kegg.pathway": "needs fix",  # FIXME
+    "ensemblglossary": "uri is wrong",
+    "epio": "content from fraunhofer is unreliable",
+    "epso": "content from fraunhofer is unreliable",
+    "gwascentral.phenotype": "website is down? or API changed?",  # FIXME
+    "gwascentral.study": "website is down? or API changed?",  # FIXME
 }
 
 X = TypeVar("X")
 
 
 def iter_helper(
-    f: Callable[[str], Mapping[str, X]],
+    f: Callable[[str, Unpack[GetOntologyKwargs]], Mapping[str, X]],
     leave: bool = False,
-    strict: bool = True,
-    **kwargs,
+    **kwargs: Unpack[IterHelperHelperDict],
 ) -> Iterable[tuple[str, str, X]]:
     """Yield all mappings extracted from each database given."""
-    for prefix, mapping in iter_helper_helper(f, strict=strict, **kwargs):
+    for prefix, mapping in iter_helper_helper(f, **kwargs):
         it = tqdm(
             mapping.items(),
             desc=f"iterating {prefix}",
@@ -250,22 +301,24 @@ def iter_helper(
             disable=None,
         )
         for key, value in it:
-            value = value.strip('"').replace("\n", " ").replace("\t", " ").replace("  ", " ")
+            if isinstance(value, str):
+                value = value.strip('"').replace("\n", " ").replace("\t", " ").replace("  ", " ")
+            # TODO deal with when this is not a string?
             if value:
                 yield prefix, key, value
 
 
 def _prefixes(
-    skip_below: Optional[str] = None,
+    skip_below: str | None = None,
     skip_below_inclusive: bool = True,
     skip_pyobo: bool = False,
-    skip_set: Optional[set[str]] = None,
+    skip_set: set[str] | None = None,
 ) -> Iterable[str]:
     for prefix, resource in sorted(bioregistry.read_registry().items()):
         if resource.no_own_terms:
             continue
         if prefix in SKIP:
-            tqdm.write(f"skipping {prefix} because in default skip set")
+            tqdm.write(f"skipping {prefix} because {SKIP[prefix]}")
             continue
         if skip_set and prefix in skip_set:
             tqdm.write(f"skipping {prefix} because in skip set")
@@ -287,37 +340,39 @@ def _prefixes(
 
 
 def iter_helper_helper(
-    f: Callable[[str], X],
+    f: Callable[[str, Unpack[GetOntologyKwargs]], X],
     use_tqdm: bool = True,
-    skip_below: Optional[str] = None,
-    skip_below_inclusive: bool = True,
+    skip_below: str | None = None,
     skip_pyobo: bool = False,
-    skip_set: Optional[set[str]] = None,
-    strict: bool = True,
-    **kwargs,
+    skip_set: set[str] | None = None,
+    **kwargs: Unpack[SlimGetOntologyKwargs],
 ) -> Iterable[tuple[str, X]]:
     """Yield all mappings extracted from each database given.
 
-    :param f: A function that takes a prefix and gives back something that will be used by an outer function.
+    :param f: A function that takes a prefix and gives back something that will be used
+        by an outer function.
     :param use_tqdm: If true, use the tqdm progress bar
-    :param skip_below: If true, skip sources whose names are less than this (used for iterative curation
+    :param skip_below: If true, skip sources whose names are less than this (used for
+        iterative curation
     :param skip_pyobo: If true, skip sources implemented in PyOBO
     :param skip_set: A pre-defined blacklist to skip
-    :param strict: If true, will raise exceptions and crash the program instead of logging them.
+    :param strict: If true, will raise exceptions and crash the program instead of
+        logging them.
     :param kwargs: Keyword arguments passed to ``f``.
-    :yields: A prefix and the result of the callable ``f``
 
     :raises TypeError: If a type error is raised, it gets re-raised
     :raises urllib.error.HTTPError: If the resource could not be downloaded
     :raises urllib.error.URLError: If another problem was encountered during download
     :raises ValueError: If the data was not in the format that was expected (e.g., OWL)
+
+    :yields: A prefix and the result of the callable ``f``
     """
+    strict = kwargs.get("strict", True)
     prefixes = list(
         _prefixes(
             skip_set=skip_set,
             skip_below=skip_below,
             skip_pyobo=skip_pyobo,
-            skip_below_inclusive=skip_below_inclusive,
         )
     )
     prefix_it = tqdm(
@@ -325,24 +380,35 @@ def iter_helper_helper(
     )
     for prefix in prefix_it:
         prefix_it.set_postfix(prefix=prefix)
+        tqdm.write(
+            click.style(f"\n{prefix} - {bioregistry.get_name(prefix)}", fg="green", bold=True)
+        )
         try:
             yv = f(prefix, **kwargs)  # type:ignore
+        except (UnhandledFormatError, NoBuildError) as e:
+            # make sure this comes before the other runtimeerror catch
+            logger.warning("[%s] %s", prefix, e)
         except urllib.error.HTTPError as e:
             logger.warning("[%s] HTTP %s: unable to download %s", prefix, e.getcode(), e.geturl())
             if strict and not bioregistry.is_deprecated(prefix):
                 raise
-        except urllib.error.URLError:
-            logger.warning("[%s] unable to download", prefix)
+        except urllib.error.URLError as e:
+            logger.warning("[%s] unable to download - %s", prefix, e.reason)
             if strict and not bioregistry.is_deprecated(prefix):
                 raise
-        except MissingPrefixError as e:
-            logger.warning("[%s] missing prefix: %s", prefix, e)
+        except ParseError as e:
+            if not e.node:
+                logger.warning("[%s] %s", prefix, e)
+            else:
+                logger.warning(str(e))
             if strict and not bioregistry.is_deprecated(prefix):
                 raise e
+        except RuntimeError as e:
+            if "DrugBank" not in str(e):
+                raise
+            logger.warning("[drugbank] invalid credentials")
         except subprocess.CalledProcessError:
             logger.warning("[%s] ROBOT was unable to convert OWL to OBO", prefix)
-        except UnhandledFormatError as e:
-            logger.warning("[%s] %s", prefix, e)
         except ValueError as e:
             if _is_xml(e):
                 # this means that it tried doing parsing on an xml page
@@ -355,6 +421,9 @@ def iter_helper_helper(
                 logger.exception(
                     "[%s] got exception %s while parsing", prefix, e.__class__.__name__
                 )
+        except zipfile.BadZipFile as e:
+            # This can happen if there's an error on UMLS
+            logger.exception("[%s] got exception %s while parsing", prefix, e.__class__.__name__)
         except TypeError as e:
             logger.exception("[%s] got exception %s while parsing", prefix, e.__class__.__name__)
             if strict:
@@ -369,7 +438,7 @@ def _is_xml(e) -> bool:
     )
 
 
-def _prep_dir(directory: Union[None, str, pathlib.Path]) -> pathlib.Path:
+def _prep_dir(directory: None | str | pathlib.Path) -> pathlib.Path:
     if directory is None:
         rv = DATABASE_DIRECTORY
     elif isinstance(directory, str):
@@ -383,26 +452,28 @@ def _prep_dir(directory: Union[None, str, pathlib.Path]) -> pathlib.Path:
 
 
 def db_output_helper(
-    f: Callable[..., Iterable[tuple[str, ...]]],
+    it: Iterable[tuple[str, ...]],
     db_name: str,
     columns: Sequence[str],
     *,
-    directory: Union[None, str, pathlib.Path] = None,
-    strict: bool = True,
+    directory: None | str | pathlib.Path = None,
+    strict: bool = False,
     use_gzip: bool = True,
-    summary_detailed: Optional[Sequence[int]] = None,
-    **kwargs,
+    summary_detailed: Sequence[int] | None = None,
 ) -> list[pathlib.Path]:
     """Help output database builds.
 
-    :param f: A function that takes a prefix and gives back something that will be used by an outer function.
+    :param f: A function that takes a prefix and gives back something that will be used
+        by an outer function.
     :param db_name: name of the output resource (e.g., "alts", "names")
     :param columns: The names of the columns
-    :param directory: The directory to output everything, or defaults to :data:`pyobo.constants.DATABASE_DIRECTORY`.
+    :param directory: The directory to output everything, or defaults to
+        :data:`pyobo.constants.DATABASE_DIRECTORY`.
     :param strict: Passed to ``f`` by keyword
-    :param kwargs: Passed to ``f`` by splat
+
     :returns: A sequence of paths that got created.
     """
+    start = time.time()
     directory = _prep_dir(directory)
 
     c: typing.Counter[str] = Counter()
@@ -415,10 +486,17 @@ def db_output_helper(
     db_sample_path = directory.joinpath(f"{db_name}_sample.tsv")
     db_summary_path = directory.joinpath(f"{db_name}_summary.tsv")
     db_summary_detailed_path = directory.joinpath(f"{db_name}_summary_detailed.tsv")
+    db_metadata_path = directory.joinpath(f"{db_name}_metadata.json")
+    rv: list[tuple[str, pathlib.Path]] = [
+        ("Metadata", db_metadata_path),
+        ("Data", db_path),
+        ("Sample", db_sample_path),
+        ("Summary", db_summary_path),
+    ]
 
     logger.info("writing %s to %s", db_name, db_path)
     logger.info("writing %s sample to %s", db_name, db_sample_path)
-    it = f(strict=strict, **kwargs)
+    sample_rows = []
     with gzip.open(db_path, mode="wt") if use_gzip else open(db_path, "w") as gzipped_file:
         writer = get_writer(gzipped_file)
 
@@ -430,12 +508,13 @@ def db_output_helper(
             writer.writerow(columns)
             sample_writer.writerow(columns)
 
-            for row, _ in zip(it, range(10)):
+            for row, _ in zip(it, range(10), strict=False):
                 c[row[0]] += 1
                 if summary_detailed is not None:
                     c_detailed[tuple(row[i] for i in summary_detailed)] += 1
                 writer.writerow(row)
                 sample_writer.writerow(row)
+                sample_rows.append(row)
 
         # continue just in the gzipped one
         for row in it:
@@ -444,7 +523,6 @@ def db_output_helper(
                 c_detailed[tuple(row[i] for i in summary_detailed)] += 1
             writer.writerow(row)
 
-    logger.info(f"writing {db_name} summary to {db_summary_path}")
     with open(db_summary_path, "w") as file:
         writer = get_writer(file)
         writer.writerows(c.most_common())
@@ -454,8 +532,8 @@ def db_output_helper(
         with open(db_summary_detailed_path, "w") as file:
             writer = get_writer(file)
             writer.writerows((*keys, v) for keys, v in c_detailed.most_common())
+        rv.append(("Summary (Detailed)", db_summary_detailed_path))
 
-    db_metadata_path = directory.joinpath(f"{db_name}_metadata.json")
     with open(db_metadata_path, "w") as file:
         json.dump(
             {
@@ -468,12 +546,12 @@ def db_output_helper(
             indent=2,
         )
 
-    rv: list[pathlib.Path] = [
-        db_metadata_path,
-        db_path,
-        db_sample_path,
-        db_summary_path,
-    ]
-    if summary_detailed:
-        rv.append(db_summary_detailed_path)
-    return rv
+    elapsed = time.time() - start
+    click.secho(f"\nWrote the following files in {elapsed:.1f} seconds\n", fg="green")
+    click.secho(indent(tabulate(rv), " "), fg="green")
+
+    click.secho("\nSample rows:\n", fg="green")
+    click.secho(indent(tabulate(sample_rows, headers=columns), " "), fg="green")
+    click.echo()
+
+    return [path for _, path in rv]