convert_text_for_ner.py

import datasets
# coding=utf-8
# Copyright 2024 HuggingFace Datasets Authors.
# Lint as: python3
"""The Shipping label Dataset."""

logger = datasets.logging.get_logger(__name__)


_CITATION = """
"""

_DESCRIPTION = """
The goal of this task is to provide a dataset for name entity recognition."""

_URL = "https://raw.githubusercontent.com/SanghaviHarshPankajkumar/shipping_label_project/main/NER/data/"
_TRAINING_FILE = "train.txt"
_VAL_FILE = "val.txt"
_TEST_FILE = "test.txt"


class shipping_labels_Config(datasets.BuilderConfig):
    """Shipping Label Dataset for ner"""

    def __init__(self, **kwargs):
        """BuilderConfig for Shipping Label data.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(shipping_labels_Config, self).__init__(**kwargs)


class shiping_label_ner(datasets.GeneratorBasedBuilder):
    """Shipping Label Dataset for ner"""

    BUILDER_CONFIGS = [
        shipping_labels_Config(
            name="shipping_label_ner", version=datasets.Version("1.0.0"), description="Shipping Label Dataset for ner"
        ),
    ]

    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-GCNUM",
                                "I-GCNUM",
                                "B-BGNUM",
                                "I-BGNUM",
                                "B-DATE",
                                "I-DATE",
                                "B-ORG",
                                "I-ORG",
                                "B-LOCATION",
                                "I-LOCATION",
                                "B-NAME",
                                "I-NAME",
                                "B-BARCODE",
                                "I-BARCODE",
                            ]
                        )
                    ),
                }
            ),
            supervised_keys=None,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        urls_to_download = {
            "train": f"{_URL}{_TRAINING_FILE}",
            "test": f"{_URL}{_TEST_FILE}",
            "val": f"{_URL}{_VAL_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.TEST, gen_kwargs={"filepath": downloaded_files["test"]}),
            datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={"filepath": downloaded_files["val"]}),
        ]

    def _generate_examples(self, filepath):
        logger.info("⏳ Generating examples from = %s", filepath)
        with open(filepath, encoding="utf-8") as f:
            current_tokens = []
            current_labels = []
            sentence_counter = 0
            for row in f:
                row = row.rstrip()
                if row:
                    token, label = row.split(" ")
                    current_tokens.append(token)
                    current_labels.append(label)
                else:
                    # New sentence
                    if not current_tokens:
                        # Consecutive empty lines will cause empty sentences
                        continue
                    assert len(current_tokens) == len(current_labels), "💔 between len of tokens & labels"
                    sentence = (
                        sentence_counter,
                        {
                            "id": str(sentence_counter),
                            "tokens": current_tokens,
                            "ner_tags": current_labels,
                        },
                    )
                    sentence_counter += 1
                    current_tokens = []
                    current_labels = []
                    yield sentence
            # Don't forget last sentence in dataset 🧐
            if current_tokens:
                yield sentence_counter, {
                    "id": str(sentence_counter),
                    "tokens": current_tokens,
                    "ner_tags": current_labels,
                }