File size: 5,842 Bytes
04a41df
 
 
 
 
 
 
5f92742
2d5f16d
04a41df
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2d5f16d
04a41df
 
c3bcc10
806a8ab
04a41df
10630a1
 
04a41df
 
 
185f86f
 
 
 
 
 
 
 
 
 
 
04a41df
2d5f16d
04a41df
2d5f16d
c3bcc10
 
 
 
04a41df
c3bcc10
2d5f16d
04a41df
 
2d5f16d
c3bcc10
04a41df
 
 
 
 
 
 
 
 
 
2cafdd0
04a41df
 
 
2cafdd0
 
 
 
 
 
2d5f16d
 
 
04a41df
 
2cafdd0
04a41df
 
 
 
 
2d5f16d
04a41df
2d5f16d
04a41df
 
 
 
 
f0bc1ac
2934055
2cafdd0
2d5f16d
c3bcc10
 
 
 
2d5f16d
c3bcc10
 
 
 
10630a1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2cafdd0
04a41df
2cafdd0
 
 
 
2d5f16d
 
 
04a41df
 
c3bcc10
04a41df
2cafdd0
 
04a41df
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
# coding=utf-8

"""COSUJU: The Court Summaries and Judgements Dataset."""

from __future__ import absolute_import, division, print_function

import json
import os
import re

import datasets


logger = datasets.logging.get_logger(__name__)


_CITATION = """\
@InProceedings{huggingface:dataset,
title   = {CoSuJu 500+ Court Judegements and Summaries for Machine Text Summarization},
authors = {Busani Ndlovu, Luke Jordan},
year    = {2021}
}
"""

# TODO: Complete description
_DESCRIPTION = """\
Court Summaries and Judgements (CoSuJu) Dataset
"""

# Redundant but may useful in future.
_URL = "https://frtnx.github.io/cosuju-extractor/dataset/"
_URLS = {
    # 'train': _URL + 'train-v1.1.json',
    'train': _URL + 'train-v1.0.json'
}


class CosujuConfig(datasets.BuilderConfig):
    """BuilderConfig for COSUJU."""

    def __init__(self, **kwargs):
        """BuilderConfig for COSUJU.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(CosujuConfig, self).__init__(**kwargs)


class Cosuju(datasets.GeneratorBasedBuilder):
    """COSUJU: The Court Summaries and Judgements Dataset. Version 1.1.0"""

    VERSION = datasets.Version("1.1.0")

    # Allows configuration to be chosen at run time
    # data = datasets.load_dataset('my_dataset', 'include_no_summary')
    # data = datasets.load_dataset('my_dataset', 'exclude_no_summary')
    BUILDER_CONFIGS = [
        datasets.BuilderConfig(name="include_no_summary", version=VERSION, description="Includes rows with no summary documents"),
        datasets.BuilderConfig(name='exclude_no_summary', version=VERSION, description="Excludes rows with no summary documents")
    ]

    DEFAULT_CONFIG_NAME = 'exclude_no_summary'

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    'id': datasets.Value('string'),
                    'title': datasets.Value('string'),
                    'url': datasets.Value('string'),
                    'year': datasets.Value('string'),
                    'update_date': datasets.Value('string'),
                    'summary_document': {
                            'filename': datasets.Value('string'),
                            'file_url': datasets.Value('string'),
                            'file_content': datasets.Value('string')
                        },
                    'judgement_document': {
                        'filename': datasets.Value('string'),
                        'file_url': datasets.Value('string'),
                        'file_content': datasets.Value('string')
                    },
                    'judgement_paragraphs': [datasets.Value('string')],
                    'max_length_judgement_paras': datasets.Value('int64')
                }
            ),
            supervised_keys=None,
            homepage='https://huggingface.co/datasets/FRTNX/cosuju',
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        downloaded_files = dl_manager.download_and_extract(_URLS)
        
        return [
            datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={'filepath': downloaded_files['train']})
        ]

    def _generate_examples(self, filepath):
        """This function returns the examples in the raw (text) form."""
        logger.info('generating examples from = %s', filepath)
        with open(filepath, encoding="utf-8") as f:
            cosuju = json.load(f)
            for row in cosuju['data']:
                # exclude cases with no summaries when configuration demands it
                if self.config.name == 'exclude_no_summary':
                    if not row['summary_document']:
                        continue

                    # edge case sanitization
                    if type(row['summary_document']) == dict and \
                        row['summary_document']['file_content'] == '':
                        continue

                if 'judgement_document' not in list(row):
                    continue

                if not row['judgement_document']:
                    continue
            
                judgement_text = row['judgement_document']['file_content']
            
                if not judgement_text:
                    continue
            
                judgement_paragraphs = re.split('\\n\\n\[\d+\]', judgement_text)
                clean_paragraphs = [paragraph.replace('\n', ' ').strip() for paragraph in judgement_paragraphs]
                squeaky_clean_paragraphs = [re.sub(' +', ' ', paragraph) for paragraph in clean_paragraphs]
                ultimate_paragraphs = [paragraph.replace('\x0c2', '').replace('  ', ' ') for paragraph in squeaky_clean_paragraphs]
            
                row['judgement_paragraphs'] = ultimate_paragraphs
                row['max_length_judgement_paras'] = max([len(paragraph.split(' ')) for paragraph in ultimate_paragraphs])

                id_ = row['id']
                result = {
                    'id': row['id'],
                    'title': row['title'],
                    'url': row['url'],
                    'year': row['year'],
                    'update_date': row['update_date'],
                    'judgement_paragraphs': row['judgement_paragraphs'],
                    'max_length_judgement_paras': row['max_length_judgement_paras']
                }

                # This is to keep things feature compliant
                for prop in ['summary_document', 'judgement_document']:
                    if row[prop]:
                        result[prop] = row[prop]
                    else:
                        result[prop] = { 'filename': '', 'file_url': '', 'file_content': '' }
                    
                yield id_, result