the_legal_pile_preprocessed.py

# coding=utf-8
# Copyright 2020 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.

"""The Legal Pile (Preprocessed)"""


import datasets
import json

try:
    import lzma as xz
except ImportError:
    import pylzma as xz

logger = datasets.logging.get_logger(__name__)

_DESCRIPTION = """The Legal Pile (Preprocessed)"""

_HOMEPAGE = "https://iliaschalkidis.github.io/"

_VERSION = "1.0.0"


class LegalPileConfig(datasets.BuilderConfig):
    """BuilderConfig for Legal Pile Corpora"""
    def __init__(self, name, n_files, **kwargs):
        super(LegalPileConfig, self).__init__(**kwargs)
        self.name = name
        self.filepath = f'https://huggingface.co/datasets/lexlms/the_legal_pile_preprocessed/' \
                        f'resolve/main/data_files/{self.name}.jsonl.tar.xz'
        self.n_files = n_files


class LegalPile(datasets.GeneratorBasedBuilder):
    """
    The Legal Pile
    """

    VERSION = datasets.Version(_VERSION)
    BUILDER_CONFIG_CLASS = LegalPileConfig
    BUILDER_CONFIGS = [
        LegalPileConfig(
            name=f"eu-legislation",
            n_files=1,
            description="EU Legislation",
        ),
        LegalPileConfig(
            name=f"eu-court-cases",
            n_files=1,
            description="EU Court cases",
        ),
        LegalPileConfig(
            name=f"ecthr-cases",
            n_files=1,
            description="ECtHR cases",
        ),
        LegalPileConfig(
            name=f"uk-legislation",
            n_files=1,
            description="UK Legislation",
        ),
        LegalPileConfig(
            name=f"uk-court-cases",
            n_files=1,
            description="UK Court cases",
        ),
        LegalPileConfig(
            name=f"indian-court-cases",
            n_files=1,
            description="Indian Court cases",
        ),
        LegalPileConfig(
            name=f"canadian-legislation",
            n_files=1,
            description="Canadian Legislation",
        ),
        LegalPileConfig(
            name=f"canadian-court-cases",
            n_files=1,
            description="Canadian Legislation",
        ),
        LegalPileConfig(
            name=f"us-legislation",
            n_files=1,
            description="US Legislation",
        ),
        LegalPileConfig(
            name=f"us-contracts",
            n_files=7,
            description="US Contracts",
        ),
        LegalPileConfig(
            name=f"us-court-cases",
            n_files=16,
            description="US Court cases",
        )
    ]

    def _info(self):
        features = datasets.Features(
            {
                "text": datasets.Value("string")
            }
        )
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=features,
            supervised_keys=None,
            homepage=_HOMEPAGE,
        )

    def _split_generators(self, dl_manager):
        train_data_dir = dl_manager.download([self.config.filepath.replace('.jsonl.tar.xz', f'_train_{idx+1}.jsonl.tar.xz')
                                              for idx in range(self.config.n_files)])
        test_data_dir = dl_manager.download([self.config.filepath.replace('.jsonl.tar.xz', '_test_1.jsonl.tar.xz')])

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={"filepaths": train_data_dir, "split": 'train'},
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={"filepaths": test_data_dir, "split": 'test'},
            )
        ]

    def _generate_examples(self, filepaths, split):
        """
        Reads line by line samples and generates examples.
        :param filepath: Path to jsonl files with line by line examples.
        """
        id_ = 0
        for filepath in filepaths:
            logger.info("⏳ Generating examples from = %s", filepath)
            with xz.open(open(filepath, "rb"), "rt", encoding='utf-8') as f:
                for row in f:
                    if row:
                         try:
                              data = json.loads(row)
                              yield id_, {
                                  "text": data["text"]
                              }
                              id_ += 1
                         except:
                              continue