# coding=utf-8 # Copyright 2022 The HuggingFace Datasets Authors and the current dataset script contributor. # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. import os from typing import List import datasets import xml.etree.ElementTree as ET import uuid import html from .bigbiohub import kb_features from .bigbiohub import BigBioConfig from .bigbiohub import Tasks _LANGUAGES = ['English'] _PUBMED = True _LOCAL = False _CITATION = """\ @article{Wei2015, title = { {GNormPlus}: An Integrative Approach for Tagging Genes, Gene Families, and Protein Domains }, author = {Chih-Hsuan Wei and Hung-Yu Kao and Zhiyong Lu}, year = 2015, journal = {{BioMed} Research International}, publisher = {Hindawi Limited}, volume = 2015, pages = {1--7}, doi = {10.1155/2015/918710}, url = {https://doi.org/10.1155/2015/918710} } """ _DATASETNAME = "citation_gia_test_collection" _DISPLAYNAME = "Citation GIA Test Collection" _DESCRIPTION = """\ The Citation GIA Test Collection was recently created for gene indexing at the NLM and includes 151 PubMed abstracts with both mention-level and document-level annotations. They are selected because both have a focus on human genes. """ _HOMEPAGE = "https://www.ncbi.nlm.nih.gov/research/bionlp/Tools/gnormplus/" _LICENSE = 'License information unavailable' _URLS = { _DATASETNAME: [ "https://www.ncbi.nlm.nih.gov/CBBresearch/Lu/Demo/tmTools/download/GNormPlus/GNormPlusCorpus.zip" ] } _SUPPORTED_TASKS = [Tasks.NAMED_ENTITY_RECOGNITION, Tasks.NAMED_ENTITY_DISAMBIGUATION] _SOURCE_VERSION = "1.0.0" _BIGBIO_VERSION = "1.0.0" class CitationGIATestCollection(datasets.GeneratorBasedBuilder): """ The Citation GIA Test Collection was recently created for gene indexing at the NLM and includes 151 PubMed abstracts with both mention-level and document-level annotations. They are selected because both have a focus on human genes. """ SOURCE_VERSION = datasets.Version(_SOURCE_VERSION) BIGBIO_VERSION = datasets.Version(_BIGBIO_VERSION) BUILDER_CONFIGS = [ BigBioConfig( name="citation_gia_test_collection_source", version=SOURCE_VERSION, description="citation_gia_test_collection source schema", schema="source", subset_id="citation_gia_test_collection", ), BigBioConfig( name="citation_gia_test_collection_bigbio_kb", version=BIGBIO_VERSION, description="citation_gia_test_collection BigBio schema", schema="bigbio_kb", subset_id="citation_gia_test_collection", ), ] DEFAULT_CONFIG_NAME = "citation_gia_test_collection_source" def _info(self) -> datasets.DatasetInfo: if self.config.schema == "source": features = datasets.Features( { "id": datasets.Value("string"), "passages": [ { "id": datasets.Value("string"), "type": datasets.Value("string"), "text": datasets.Sequence(datasets.Value("string")), "offsets": datasets.Sequence([datasets.Value("int32")]), } ], "entities": [ { "id": datasets.Value("string"), "type": datasets.Value("string"), "text": datasets.Sequence(datasets.Value("string")), "offsets": datasets.Sequence([datasets.Value("int32")]), "normalized": [ { "db_name": datasets.Value("string"), "db_id": datasets.Value("string"), } ], } ], } ) elif self.config.schema == "bigbio_kb": features = kb_features return datasets.DatasetInfo( description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, license=str(_LICENSE), citation=_CITATION, ) def _split_generators(self, dl_manager) -> List[datasets.SplitGenerator]: urls = _URLS[_DATASETNAME] data_dir = dl_manager.download_and_extract(urls) return [ datasets.SplitGenerator( name=datasets.Split.TEST, gen_kwargs={ "filepath": os.path.join( data_dir[0], "GNormPlusCorpus/NLMIAT.BioC.xml" ), "split": "NLMIAT", }, ), ] def _get_entities(self, annot_d: dict) -> dict: """' Converts annotation dict to entity dict. """ ent = { "id": str(uuid.uuid4()), "type": annot_d["type"], "text": [annot_d["text"]], "offsets": [annot_d["offsets"]], "normalized": [ { "db_name": "NCBI Gene" if annot_d["type"].isdigit() else "", "db_id": annot_d["type"] if annot_d["type"].isdigit() else "", } ], } return ent def _get_offsets_entities( child, parent_text: str, child_text: str, offset: int ) -> List[int]: """ Extracts child text offsets from parent text for entities. Some offsets that were present in the datset were wrong mainly because of string encodings. Also a little fraction of parent strings doesn't contain its respective child strings. Hence few assertion errors in the entitity offsets checking test. """ if child_text in parent_text: index = parent_text.index(child_text) start = index + offset else: start = offset end = start + len(child_text) return [start, end] def _process_annot(self, annot: ET.Element, passages: dict) -> dict: """' Converts annotation XML Element to Python dict. """ parent_text = " ".join([p["text"] for p in passages.values()]) annot_d = dict() a_d = {a.tag: a.text for a in annot} for a in list(annot): if a.tag == "location": offset = int(a.attrib["offset"]) annot_d["offsets"] = self._get_offsets_entities( html.escape(parent_text[offset:]), html.escape(a_d["text"]), offset ) elif a.tag != "infon": annot_d[a.tag] = html.escape(a.text) else: annot_d[a.attrib["key"]] = html.escape(a.text) return annot_d def _parse_elem(self, elem: ET.Element) -> dict: """' Converts document XML Element to Python dict. """ elem_d = dict() passages = dict() annotations = elem.findall(".//annotation") elem_d["entities"] = [] for child in elem: elem_d[child.tag] = [] for child in elem: if child.tag == "passage": elem_d[child.tag].append( { c.tag: html.escape( " ".join( list( filter( lambda item: item, [t.strip("\n") for t in c.itertext()], ) ) ) ) for c in child } ) elif child.tag == "id": elem_d[child.tag] = html.escape(child.text) for passage in elem_d["passage"]: infon = passage["infon"] passage.pop("infon", None) passages[infon] = passage elem_d["passages"] = passages elem_d.pop("passage", None) for a in annotations: elem_d["entities"].append(self._process_annot(a, elem_d["passages"])) return elem_d def _generate_examples(self, filepath, split): root = ET.parse(filepath).getroot() if self.config.schema == "source": uid = 0 for elem in root.findall("document"): row = self._parse_elem(elem) uid += 1 passages = row["passages"] yield uid, { "id": str(uid), "passages": [ { "id": str(uuid.uuid4()), "type": "title", "text": [passages["title"]["text"]], "offsets": [ [ int(passages["title"]["offset"]), int(passages["title"]["offset"]) + len(passages["title"]["text"]), ] ], }, { "id": str(uuid.uuid4()), "type": "abstract", "text": [passages["abstract"]["text"]], "offsets": [ [ int(passages["abstract"]["offset"]), int(passages["abstract"]["offset"]) + len(passages["abstract"]["text"]), ] ], }, ], "entities": [self._get_entities(a) for a in row["entities"]], } elif self.config.schema == "bigbio_kb": uid = 0 for elem in root.findall("document"): row = self._parse_elem(elem) uid += 1 passages = row["passages"] yield uid, { "id": str(uid), "document_id": str(uuid.uuid4()), "passages": [ { "id": str(uuid.uuid4()), "type": "title", "text": [passages["title"]["text"]], "offsets": [ [ int(passages["title"]["offset"]), int(passages["title"]["offset"]) + len(passages["title"]["text"]), ] ], }, { "id": str(uuid.uuid4()), "type": "abstract", "text": [passages["abstract"]["text"]], "offsets": [ [ int(passages["abstract"]["offset"]), int(passages["abstract"]["offset"]) + len(passages["abstract"]["text"]), ] ], }, ], "entities": [self._get_entities(a) for a in row["entities"]], "relations": [], "events": [], "coreferences": [], }