import json import gzip import datasets from datasets import Value, Sequence logger = datasets.logging.get_logger(__name__) _DESCRIPTION = """\ V2EX is a dataset curated by https://www.v2ex.com/ open data. """ _HOMEPAGE = "https://huggingface.co/datasets/Dialogue-Model-Research-Group/v2ex" _DATA_URLS = { "topic": "v2ex_topic.jsonl.gz", "replies": "v2ex_replies.jsonl.gz", "conversation": "v2ex_conversation.jsonl.gz", "context-response": "v2ex_context_response.jsonl.gz", } class V2EXConfig(datasets.BuilderConfig): """BuilderConfig for shu.""" def __init__(self, *args, subsets, **kwargs) -> None: """BuilderConfig for shu. Args: **kwargs: keyword arguments forwarded to super. """ super(V2EXConfig, self).__init__( *args, name="+".join(subsets), **kwargs ) self.subsets = subsets class V2EX(datasets.GeneratorBasedBuilder): """V2EX: Chinese book dataset.""" VERSION = datasets.Version("0.1.0") BUILDER_CONFIG_CLASS = V2EXConfig BUILDER_CONFIGS = [V2EXConfig(subsets=[subset]) for subset in _DATA_URLS] DEFAULT_CONFIG_NAME = "topic" def _info(self): if self.config.name == "topic": features = datasets.Features({ "id": Value("int64"), "title": Value("string"), "content": Value("string"), "content_rendered": Value("string"), "syntax": Value("int64"), "url": Value("string"), "replies": Value("int64"), "last_reply_by": Value("string"), "created": Value("int64"), "last_modified": Value("int64"), "last_touched": Value("int64"), "member": { "id": Value("int64"), "username": Value("string"), "bio": Value("string"), "website": Value("string"), "github": Value("string"), "url": Value("string"), "avatar": Value("string"), "created": Value("int64"), }, "node": { "id": Value("int64"), "url": Value("string"), "name": Value("string"), "title": Value("string"), "header": Value("string"), "footer": Value("string"), "avatar": Value("string"), "topics": Value("int64"), "created": Value("int64"), "last_modified": Value("int64"), }, "supplements": Sequence({ "id": Value("int64"), "content": Value("string"), "content_rendered": Value("string"), "syntax": Value("int64"), "created": Value("int64"), }), }) elif self.config.name == "replies": features = datasets.Features({ "id": Value(dtype="int64", id=None), "content": Value(dtype="string", id=None), "content_rendered": Value(dtype="string", id=None), "created": Value(dtype="int64", id=None), "member": { "id": Value(dtype="int64", id=None), "username": Value(dtype="string", id=None), "bio": Value(dtype="string", id=None), "website": Value(dtype="string", id=None), "github": Value(dtype="string", id=None), "url": Value(dtype="string", id=None), "avatar": Value(dtype="string", id=None), "created": Value(dtype="int64", id=None) }, "topic_id": Value(dtype="int64", id=None), }) elif self.config.name == "conversation": features = datasets.Features({ "conversation": Sequence({ "username": Value("string"), "post": Value("string"), }), "url": Value(dtype="string"), }) elif self.config.name == "context-response": features = datasets.Features({ "context": Value("string"), "response": Value("string"), }) else: raise NotImplementedError( "This dataset is not implemented." ) return datasets.DatasetInfo( description=_DESCRIPTION, features=features, homepage=_HOMEPAGE, ) def _split_generators(self, dl_manager): data_urls = {subset: _DATA_URLS[subset] for subset in self.config.subsets} archive = dl_manager.download(data_urls) return [ datasets.SplitGenerator( name=datasets.Split.TRAIN, gen_kwargs={ "files": { subset: archive[subset] for subset in self.config.subsets }, }, ), ] def _generate_examples(self, files): key = 0 for subset in files: filepath = files[subset] for line in gzip.open(filepath, "rt", encoding="utf-8"): j = json.loads(line) yield key, j key += 1