rag-rfb.py

# coding=utf-8
# Copyright 2020 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
#
# 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.

# Lint as: python3
"""rag-rfb dataset."""

import datasets
import json
import numpy as np
import glob
import os


_CITATION = """
place holder
"""

_URL = "https://github.com/unicamp-dl/rag-rfb"

_DESCRIPTION = """
Retrieval Augmented Generation (RAG) dataset for Brazilian Federal Revenue Service (Receita Federal do Brasil ― RFB).
"""

_URLS = {
    "2024-questions": "https://huggingface.co/datasets/unicamp-dl/rag-rfb/resolve/main/questions_QA_2024_v1.0.json",
    "2024-sources": "https://huggingface.co/datasets/unicamp-dl/rag-rfb/resolve/main/2024-sources",
}



def generate_examples_questions(filepath):

    with open(filepath, encoding="utf-8") as input_file:
        questions = json.load(input_file)

    for (idx, question) in enumerate(questions):
        
        # Convert the "all_formatted_references" dictionary to a list to avoid multiple nulled rows

        all_formatted_references = []

        for reference in np.sort(list(question['all_formatted_references'].keys())):
            all_formatted_references += question['all_formatted_references'][reference]

        question['all_formatted_references'] = all_formatted_references

        yield idx, question



def generate_examples_sources(filepath):

    all_files = np.sort(glob.glob(filepath + "/*.txt"))

    for idx, which_file in enumerate(all_files):
        with open(which_file, encoding="utf-8") as input_file:
            file_data = input_file.read()

        features = {"file": os.path.basename(which_file),
                    "text": file_data}

        yield idx, features



class RAG_RFB(datasets.GeneratorBasedBuilder):

    BUILDER_CONFIGS = (
        [
            datasets.BuilderConfig(
                name="2024-questions",
                description="Questions from 2024 Questions & Answers document.",
                version=datasets.Version("1.0.0"),
            ),

            datasets.BuilderConfig(
                name="2024-sources",
                description="Legal documents referred by the 2024 Questions & Answers document.",
                version=datasets.Version("1.0.0"),
            )
        ]
    )

    DEFAULT_CONFIG_NAME = "2024-questions"


    def _info(self):
        name = self.config.name
        if "questions" in name:
            features = {
                "question_number": datasets.Value("string"),
                "question_summary": datasets.Value("string"),
                "question_text": datasets.Value("string"),
                "answer": datasets.Sequence(datasets.Value("string"), length=-1), 
                "answer_cleaned": datasets.Sequence(datasets.Value("string"), length=-1), 
                "references": datasets.Sequence(datasets.Value("string"), length=-1), 
                "linked_questions": datasets.Sequence(datasets.Value("string"), length=-1), 

                "formatted_references": datasets.Sequence({"título": datasets.Value("string"),
                                                           "artigos": datasets.Sequence(datasets.Value("string"), length=-1),
                                                           "anexos": datasets.Sequence(datasets.Value("string"), length=-1),
                                                           "file": datasets.Value("string")}),

                "embedded_references": datasets.Sequence(datasets.Value("string"), length=-1), 
                
                "formatted_embedded_references":  datasets.Sequence({"título": datasets.Value("string"),
                                                           "artigos": datasets.Sequence(datasets.Value("string"), length=-1),
                                                           "anexos": datasets.Sequence(datasets.Value("string"), length=-1),
                                                           "file": datasets.Value("string")}),

                "all_formatted_references":  datasets.Sequence({"título": datasets.Value("string"),
                                                           "artigos": datasets.Sequence(datasets.Value("string"), length=-1),
                                                           "anexos": datasets.Sequence(datasets.Value("string"), length=-1),
                                                           "file": datasets.Value("string")})
            }
        else:
            features = {
                "file": datasets.Value("string"),
                "text": datasets.Value("string"),
            }

        return datasets.DatasetInfo(
            description=f"{_DESCRIPTION}\n{self.config.description}",
            features=datasets.Features(features),
            supervised_keys=None,
            homepage=_URL,
            citation=_CITATION,
        )


    def _split_generators(self, dl_manager):
        url = _URLS[self.config.name]
        dl_path = dl_manager.download_and_extract(url)

        return (datasets.SplitGenerator(name=self.config.name, gen_kwargs={"filepath": dl_path}),)


    def _generate_examples(self, filepath, args=None):
        """Yields examples."""

        if "questions" in self.config.name:
            return generate_examples_questions(filepath)
        else:
            return generate_examples_sources(filepath)