import os
import copy
import json
import logging

import torch
from torch.utils.data import TensorDataset

from bert_finetune_cls.utils import get_intent_labels

# 日志对象初始化
logger = logging.getLogger(__name__)


class InputExample(object):
    """
    定义InputExample类数据
    """

    def __init__(self, guid, words, intent_label=None, ):
        # 每个样本的独特序号
        self.guid = guid
        # 原文本
        self.words = words
        # 意图标签
        self.intent_label = intent_label

    def __repr__(self):
        """
         这里重写我们的输入信息
        """
        return str(self.to_json_string())

    def to_dict(self):
        """
         将此实例序列化到Python字典中
        __dict__:类的静态函数、类函数、普通函数、全局变量以及一些内置的属性都是放在类__dict__里的
        对象实例的__dict__中存储了一些self.xxx的一些东西
        """
        output = copy.deepcopy(self.__dict__)
        return output

    def to_json_string(self):
        """
        类的属性等信息(字典格式)dumps进入json string
        json.dumps()函数将python对象编码成JSON字符串
        indent=2.文件格式中加入了换行与缩进
        """
        return json.dumps(self.to_dict(), indent=2, sort_keys=True) + "\n"


class InputFeatures(object):
    """
    定义输入到BERT模型的InputFeatures类数据
    """

    def __init__(self, input_ids, attention_mask, token_type_ids, intent_label_id):
        # 输入样本序列在bert词表里的索引，可以直接喂给nn.embedding
        self.input_ids = input_ids
        # 注意力mask，padding的部分为0，其他为1
        self.attention_mask = attention_mask
        # 表示每个token属于句子1还是句子2
        self.token_type_ids = token_type_ids
        # 意图标签
        self.intent_label_id = intent_label_id

    def __repr__(self):
        """
        这里重写我们的输入信息
        """
        return str(self.to_json_string())

    def to_dict(self):
        """
         将此实例序列化到Python字典中
        __dict__:类的静态函数、类函数、普通函数、全局变量以及一些内置的属性都是放在类__dict__里的
        对象实例的__dict__中存储了一些self.xxx的一些东西
        """
        output = copy.deepcopy(self.__dict__)
        return output

    def to_json_string(self):
        """
        类的属性等信息(字典格式)dump进入json string
        json.dumps()函数将python对象编码成JSON字符串
        indent=2.文件格式中加入了换行与缩进
        """
        return json.dumps(self.to_dict(), indent=2, sort_keys=True) + "\n"


class JointProcessor(object):
    """
    意图分类任务的数据处理器
    """

    def __init__(self, args):
        # 参数
        self.args = args
        # 获取意图标签
        self.intent_labels = get_intent_labels(args)
        # 输入文本文件
        self.input_text_file = 'seq.in'
        # 意图标签文件
        self.intent_label_file = 'label'

    @classmethod
    def _read_file(cls, input_file, quotechar=None):
        """
        逐行读取输入文件
        :param input_file: 输入文件路径
        :param quotechar:
        :return: 句子列表
        """
        with open(input_file, "r", encoding="utf-8") as f:
            lines = []
            for line in f:
                lines.append(line.strip())
            return lines

    def _create_examples(self, texts, intents, set_type):
        """
        为训练集与验证集构建example
        :param lines: 句子列表
        :param set_type: 区分训练集与验证集
        :return: 处理后的InputExample类数据
        """
        examples = []
        for i, (text, intent) in enumerate(zip(texts, intents)):
            # 每个样本的独特序号
            guid = "%s-%s" % (set_type, i)
            # 去掉首尾空格
            words = text.split()  # Some are spaced twice
            # 意图标签索引
            intent_label = self.intent_labels.index(
                intent) if intent in self.intent_labels else self.intent_labels.index("UNK")
            # 将每一行JSON文件转化成InputExample类数据
            examples.append(InputExample(guid=guid,
                                         words=words,
                                         intent_label=intent_label))
        return examples

    def get_examples(self, mode):
        """
        获得样例数据
        Args:
            mode: train, dev, test
        """
        # 拼接数据路径
        data_path = os.path.join(self.args.data_dir, self.args.task, mode)
        logger.info("LOOKING AT {}".format(data_path))
        # 构建InputExample类样例数据
        return self._create_examples(texts=self._read_file(os.path.join(data_path, self.input_text_file)),
                                     intents=self._read_file(os.path.join(data_path, self.intent_label_file)),
                                     set_type=mode)


# 如果有多个数据集，则数据集的processor可以通过映射得到
processors = {
    "atis": JointProcessor,
    "snips": JointProcessor
}


def convert_examples_to_features(examples, max_seq_len, tokenizer,
                                 cls_token_segment_id=0,
                                 pad_token_segment_id=0,
                                 sequence_a_segment_id=0,
                                 mask_padding_with_zero=True):
    """
    将example数据转化为BERT模型需要的输入格式
    :param examples: 样例数据
    :param max_seq_len: 最大序列长度
    :param tokenizer: 分词
    :param cls_token_segment_id: 0
    :param pad_token_segment_id: 0
    :param sequence_a_segment_id: 0
    :param mask_padding_with_zero: True
    :return:
    """
    # 基于当前分词模型的设置
    cls_token = tokenizer.cls_token
    sep_token = tokenizer.sep_token
    unk_token = tokenizer.unk_token
    pad_token_id = tokenizer.pad_token_id

    features = []
    # 循环遍历每一个样例数据
    for (ex_index, example) in enumerate(examples):
        # 每隔1000个数据，写入日志
        if ex_index % 5000 == 0:
            logger.info("Writing example %d of %d" % (ex_index, len(examples)))

        # Tokenize word by word (for cls)
        tokens = []
        # 遍历每一个样例数据中的每一个单词
        for word in example.words:
            # 对token进行分词
            word_tokens = tokenizer.tokenize(word)
            # 如果存在错误编码
            if not word_tokens:
                word_tokens = [unk_token]  # For handling the bad-encoded word
            tokens.extend(word_tokens)

        # Account for [CLS] and [SEP]
        special_tokens_count = 2
        # 如果句子过长，则进行截断
        if len(tokens) > max_seq_len - special_tokens_count:
            # token词表也进行相应的截断
            tokens = tokens[:(max_seq_len - special_tokens_count)]

        # Add [SEP] token
        tokens += [sep_token]
        token_type_ids = [sequence_a_segment_id] * len(tokens)

        # Add [CLS] token
        tokens = [cls_token] + tokens
        token_type_ids = [cls_token_segment_id] + token_type_ids

        # 将tokens转化为bert词表中对应的id
        input_ids = tokenizer.convert_tokens_to_ids(tokens)

        # The mask has 1 for real tokens and 0 for padding tokens. Only real
        # tokens are attended to.
        attention_mask = [1 if mask_padding_with_zero else 0] * len(input_ids)

        # 需要填充的序列长度
        padding_length = max_seq_len - len(input_ids)
        # 输入样本序列在bert词表里的索引
        input_ids = input_ids + ([pad_token_id] * padding_length)
        # 注意力mask，padding的部分为0，其他为1
        attention_mask = attention_mask + ([0 if mask_padding_with_zero else 1] * padding_length)
        # token_type_ids表示每个token属于句子1还是句子2
        token_type_ids = token_type_ids + ([pad_token_segment_id] * padding_length)

        # 验证长度是否填充至最长序列
        assert len(input_ids) == max_seq_len, "Error with input length {} vs {}".format(len(input_ids), max_seq_len)
        assert len(attention_mask) == max_seq_len, "Error with attention mask length {} vs {}".format(
            len(attention_mask), max_seq_len)
        assert len(token_type_ids) == max_seq_len, "Error with token type length {} vs {}".format(len(token_type_ids),
                                                                                                  max_seq_len)

        # 意图标签取整
        intent_label_id = int(example.intent_label)

        # 如果是前5个数据，则记录日志
        if ex_index < 5:
            logger.info("*** Example ***")
            logger.info("guid: %s" % example.guid)
            logger.info("tokens: %s" % " ".join([str(x) for x in tokens]))
            logger.info("input_ids: %s" % " ".join([str(x) for x in input_ids]))
            logger.info("attention_mask: %s" % " ".join([str(x) for x in attention_mask]))
            logger.info("token_type_ids: %s" % " ".join([str(x) for x in token_type_ids]))
            logger.info("intent_label: %s (id = %d)" % (example.intent_label, intent_label_id))

        # 构造InputFeatures类BERT模型输入数据
        features.append(
            InputFeatures(input_ids=input_ids,
                          attention_mask=attention_mask,
                          token_type_ids=token_type_ids,
                          intent_label_id=intent_label_id,
                          ))

    return features


def load_and_cache_examples(args, tokenizer, mode):
    """
    将数据转化为cache文件，方便下一次快速加载
    :param args: 参数
    :param tokenizer: 分词
    :param mode: 区分训练、验证、测试
    :return:
    """
    # 加载数据处理器
    processor = processors[args.task](args)

    # 拼接cach文件路径
    cached_features_file = os.path.join(
        args.data_dir,
        'cached_{}_{}_{}_{}'.format(
            mode,
            args.task,
            list(filter(None, args.model_name_or_path.split("/"))).pop(),
            args.max_seq_len
        )
    )
    # 如果路径存在，则加载cach文件
    if os.path.exists(cached_features_file) and False:
        logger.info("Loading features from cached file %s", cached_features_file)
        features = torch.load(cached_features_file)
    # 如果路径不存在
    else:
        # Load data features from dataset file
        logger.info("Creating features from dataset file at %s", args.data_dir)
        # 区分数据集
        if mode == "train":
            examples = processor.get_examples("train")
        elif mode == "dev":
            examples = processor.get_examples("dev")
        elif mode == "test":
            examples = processor.get_examples("test")
        else:
            raise Exception("For mode, Only train, dev, test is available")

        # 将example数据转化为features数据
        features = convert_examples_to_features(examples,
                                                args.max_seq_len,
                                                tokenizer,
                                                )
        logger.info("Saving features into cached file %s", cached_features_file)
        # 将数据保存至cach路径中
        torch.save(features, cached_features_file)

    # Convert to Tensors and build dataset
    all_input_ids = torch.tensor(
        [f.input_ids for f in features],
        dtype=torch.long
    )
    all_attention_mask = torch.tensor(
        [f.attention_mask for f in features],
        dtype=torch.long
    )
    all_token_type_ids = torch.tensor(
        [f.token_type_ids for f in features],
        dtype=torch.long
    )
    all_intent_label_ids = torch.tensor(
        [f.intent_label_id for f in features],
        dtype=torch.long
    )
    # 构造dataset
    dataset = TensorDataset(all_input_ids, all_attention_mask,
                            all_token_type_ids, all_intent_label_ids)
    return dataset