Create CRAFT-Chem.py

CRAFT-Chem.py

@@ -0,0 +1,114 @@
+import datasets
+
+
+logger = datasets.logging.get_logger(__name__)
+
+
+_CITATION = """\
+@article{krallinger2015chemdner,
+  title={The CHEMDNER corpus of chemicals and drugs and its annotation principles},
+  author={Krallinger, Martin and Rabal, Obdulia and Leitner, Florian and Vazquez, Miguel and Salgado, David and Lu, Zhiyong and Leaman, Robert and Lu, Yanan and Ji, Donghong and Lowe, Daniel M and others},
+  journal={Journal of cheminformatics},
+  volume={7},
+  number={1},
+  pages={1--17},
+  year={2015},
+  publisher={BioMed Central}
+}
+"""
+
+_DESCRIPTION = """\
+"""
+
+_HOMEPAGE = ""
+_URL = "https://github.com/cambridgeltl/MTL-Bioinformatics-2016/raw/master/data/BC5CDR-chem-IOB/"
+_TRAINING_FILE = "train.tsv"
+_DEV_FILE = "devel.tsv"
+_TEST_FILE = "test.tsv"
+
+
+class BC4CHEMDConfig(datasets.BuilderConfig):
+    """BuilderConfig for  BC4CHEMD"""
+
+    def __init__(self, **kwargs):
+        """BuilderConfig for  BC4CHEMD.
+        Args:
+          **kwargs: keyword arguments forwarded to super.
+        """
+        super(BC4CHEMDConfig, self).__init__(**kwargs)
+
+
+class  BC4CHEMD(datasets.GeneratorBasedBuilder):
+    """ BC4CHEMD dataset."""
+
+    BUILDER_CONFIGS = [
+        BC4CHEMDConfig(name="BC5CDR", version=datasets.Version("1.0.0"), description=" BC5CDR dataset"),
+    ]
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "id": datasets.Value("string"),
+                    "tokens": datasets.Sequence(datasets.Value("string")),
+                    "ner_tags": datasets.Sequence(
+                        datasets.features.ClassLabel(
+                            names=[
+                                "O",
+                                "B-CHEBI",
+                                "I-CHEBI",
+                            ]
+                        )
+                    ),
+                }
+            ),
+            supervised_keys=None,
+            homepage=_HOMEPAGE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+        urls_to_download = {
+            "train": f"{_URL}{_TRAINING_FILE}",
+            "dev": f"{_URL}{_DEV_FILE}",
+            "test": f"{_URL}{_TEST_FILE}",
+        }
+        downloaded_files = dl_manager.download_and_extract(urls_to_download)
+
+        return [
+            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloaded_files["train"]}),
+            datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["dev"]}),
+            datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": downloaded_files["test"]}),
+        ]
+
+    def _generate_examples(self, filepath):
+        logger.info("⏳ Generating examples from = %s", filepath)
+        with open(filepath, encoding="utf-8") as f:
+            guid = 0
+            tokens = []
+            ner_tags = []
+            for line in f:
+                if line == "" or line == "\n":
+                    if tokens:
+                        print(tokens)
+                        yield guid, {
+                            "id": str(guid),
+                            "tokens": tokens,
+                            "ner_tags": ner_tags,
+                        }
+                        guid += 1
+                        tokens = []
+                        ner_tags = []
+                else:
+                    # tokens are tab separated
+                    splits = line.split("\t")
+                    tokens.append(splits[0])
+                    ner_tags.append(splits[1].rstrip())
+            # last example
+            yield guid, {
+                "id": str(guid),
+                "tokens": tokens,
+                "ner_tags": ner_tags,
+            }