"""
 Copyright (c) 2023, salesforce.com, inc.
 All rights reserved.
 SPDX-License-Identifier: BSD-3-Clause
 For full license text, see the LICENSE file in the repo root or https://opensource.org/licenses/BSD-3-Clause
"""
import logging

import copy
import torch
import torch.nn as nn
from torch.cuda.amp import autocast as autocast
from transformers import T5TokenizerFast, BertTokenizer

from lavis.common.registry import registry
from lavis.models.blip2_models.blip2 import Blip2Base, disabled_train
from lavis.models.blip2_models.modeling_t5 import T5Config, T5ForConditionalGeneration

@registry.register_model("blip2_fmr") # frame-level moment retrieval
class Blip2FMR(Blip2Base):
    """
    BLIP2 T5 model.
    Supported model types:
        - pretrain_flant5xl: pretrained model with FlanT5-XL
        - pretrain_flant5xxl: pretrained model with FlanT5-XXL
        - caption_coco_flant5xl: fintuned image captioning model with FlanT5-XL
    Usage:
        >>> from lavis.models import load_model
        >>> model = load_model("blip2_t5", "pretrain_flant5xl")
    """

    PRETRAINED_MODEL_CONFIG_DICT = {
        "pretrain_flant5xl": "configs/models/blip2/blip2_pretrain_flant5xl.yaml",
        "pretrain_flant5xxl": "configs/models/blip2/blip2_pretrain_flant5xxl.yaml",
        "caption_coco_flant5xl": "configs/models/blip2/blip2_caption_flant5xl.yaml",
    }

    def __init__( self, img_size=224, drop_path_rate=0,
        use_grad_checkpoint=False, vit_precision="fp16", freeze_vit=True,
        num_query_token=32, t5_model="google/flan-t5-xl", prompt="",
        max_txt_len=32, frame_num=8, answer_num=5, apply_lemmatizer=False, task='qa'):
        """
        apply_lemmatizer: when set to True, postprocess predict_answers() result with lemmas.
        """
        super().__init__()
        
        self.task = task
        
        # vision backbone
        self.visual_encoder, self.ln_vision_loc, _ = self.init_vision_encoder(
        img_size, drop_path_rate, use_grad_checkpoint, vit_precision)
        # Freeze ViT
        if freeze_vit:
            for name, param in self.visual_encoder.named_parameters():
                param.requires_grad = False         
            self.visual_encoder = self.visual_encoder.eval()
            self.visual_encoder.train = disabled_train
            logging.info("freeze vision encoder")
            
        # text backbone
        self.t5_tokenizer = T5TokenizerFast.from_pretrained(t5_model)
        t5_config = T5Config.from_pretrained(t5_model)
        t5_config.dense_act_fn = "gelu"
        self.t5_model = T5ForConditionalGeneration.from_pretrained(
        t5_model, config=t5_config)
        # Freeze T5
        for name, param in self.t5_model.named_parameters():
            param.requires_grad = False
            param.data = param.data.bfloat16() 
        
        # Q-Former for Frame Localization
        self.Qformer_loc, self.query_tokens_loc = self.init_Qformer(
        num_query_token, self.visual_encoder.num_features)

        self.Qformer_loc.cls = None
        self.Qformer_loc.bert.embeddings.word_embeddings = None
        self.Qformer_loc.bert.embeddings.position_embeddings = None
        for layer in self.Qformer_loc.bert.encoder.layer:
            layer.output = None
            layer.intermediate = None
        self.t5_proj_loc = nn.Linear(
        self.Qformer_loc.config.hidden_size, self.t5_model.config.hidden_size
        )
            
        self.max_txt_len = 77
        #self.prompt = prompt
        answer_id = [71, 272, 205, 309, 262] # A B C D E
        self.answer_id = answer_id[:answer_num]
        # self.answer_id = [71, 272]
        self.yes_id, self.no_id = 4273, 150
        
        self._apply_lemmatizer = apply_lemmatizer
        self._lemmatizer = None
        
        self.frame_num = frame_num
        self.ANS_MAP = {'A':0, 'B':1, 'C':2, 'D':3, 'E':4}
        self.frame_prefix = ['Frame: ']
            
    def forward(self, samples):

        image = samples["video"]
        text_input = samples['loc_input'] # query + options + Prompt
        bs_answer = samples['qa_output'] # yes or no
        flat_answer = []
        for answer in bs_answer:
            answer = answer.split('_')
            for a in answer:
                flat_answer.append(a)
        
        b, t, c, w, h = image.shape 
        image = image.reshape(-1, c, w, h)
        image_embeds = self.ln_vision_loc(self.visual_encoder(image)) # bt, n, c
        _, n, _ = image_embeds.shape
        image_atts = torch.ones(image_embeds.size()[:-1], dtype=torch.long).to(image.device) # bt n c
        
        #pass
        query_tokens = self.query_tokens_loc.expand(image_embeds.shape[0], -1, -1)
        query_output = self.Qformer_loc.bert(
            query_embeds=query_tokens, encoder_hidden_states=image_embeds,
            encoder_attention_mask=image_atts, return_dict=True)
        inputs_t5 = self.t5_proj_loc(query_output.last_hidden_state)
        atts_t5 = torch.ones(inputs_t5.size()[:-1], dtype=torch.long).to(image.device) 

        with torch.cuda.amp.autocast(dtype=torch.bfloat16):
            # Frame Prefix
            frame_prefix = self.t5_tokenizer(
                self.frame_prefix, padding="longest", add_special_tokens=False,
                truncation=True, max_length=self.max_txt_len, return_tensors="pt",
            ).to(image.device) # 
            # print('frame_prefix 1', frame_prefix.input_ids.shape) 8, 4
            frame_prefix_id = torch.repeat_interleave(frame_prefix.input_ids, b*t, 0)
            frame_prefix_mask = torch.repeat_interleave(frame_prefix.attention_mask, b*t, 0)
            # Question, Options input
            input_tokens = self.t5_tokenizer(
                text_input, padding="longest", truncation=True,
                max_length=self.max_txt_len, return_tensors="pt").to(image.device)
            input_ids = torch.repeat_interleave(input_tokens.input_ids, t, 0)
            input_attention_mask = torch.repeat_interleave(input_tokens.attention_mask, t, 0)

            # Output target
            output_tokens = self.t5_tokenizer(
                flat_answer, padding="longest", truncation=True,
                max_length=self.max_txt_len, return_tensors="pt").to(image.device)
            targets = output_tokens.input_ids.masked_fill(
                output_tokens.input_ids == self.t5_tokenizer.pad_token_id, -100)
            output_tokens_mask = output_tokens.attention_mask #torch.repeat_interleave(output_tokens.attention_mask, t, dim=0)
            #targets = torch.repeat_interleave(targets, t, dim=0)
            # input for QA
            frame_predix_embed = self.t5_model.encoder.embed_tokens(frame_prefix_id)
            inputs_embeds = self.t5_model.encoder.embed_tokens(input_ids)
            inputs_embeds = torch.cat([frame_predix_embed, inputs_t5, inputs_embeds], dim=1)
            encoder_atts = torch.cat([frame_prefix_mask, atts_t5, input_attention_mask], dim=1)

            outputs = self.t5_model(
                inputs_embeds=inputs_embeds, attention_mask=encoder_atts,
                decoder_attention_mask=output_tokens_mask, return_dict=True, labels=targets)
            loss = outputs.loss
                
        return {"loss": loss}
        
    @torch.no_grad()
    def generate(self,
        samples,
        use_nucleus_sampling=False,
        num_beams=5, max_length=30,
        min_length=1, top_p=0.9,
        repetition_penalty=1.0, length_penalty=1.0,
        num_captions=1, temperature=1,):
        """
        Args:
            samples (dict): A dictionary containing the following keys:
                - image (torch.Tensor): A tensor of shape (batch_size, 3, H, W)
            use_nucleus_sampling (bool): Whether to use nucleus sampling. If False, use top-k sampling.
            num_beams (int): Number of beams for beam search. 1 means no beam search.
            max_length (int): The maximum length of the sequence to be generated.
            min_length (int): The minimum length of the sequence to be generated.
            top_p (float): The cumulative probability for nucleus sampling.
            repetition_penalty (float): The parameter for repetition penalty. 1.0 means no penalty.
            num_captions (int): Number of captions to be generated for each image.
        Returns:
            captions (list): A list of strings of length batch_size * num_captions.
        """
        out = {}
        image, qid = samples["video"], samples['question_id']
        text_input, bs_answer = samples['loc_input'], samples['qa_output']  # Q + Options + Prompt: Choose an answer from options based on the frame.
        # print('text_input', text_input)
        flat_answer = []
        # print('bs_answer', bs_answer)
        for answer in bs_answer:
            answer = answer.split('_')
            for a in answer:
                flat_answer.append(a)
        # print('flat_answer', flat_answer)
        
        b, t, c, w, h = image.shape       
        image = image.reshape(-1, c, w, h)
        with torch.cuda.amp.autocast(enabled=(self.device != torch.device("cpu"))):
            image_embeds = self.ln_vision_loc(self.visual_encoder(image)) # bt, n, c
        _, n, _ = image_embeds.shape
        image_atts = torch.ones(image_embeds.size()[:-1], dtype=torch.long).to(image.device) # bt n c
        
        query_tokens = self.query_tokens_loc.expand(image_embeds.shape[0], -1, -1)
        query_output = self.Qformer_loc.bert(
            query_embeds=query_tokens, encoder_hidden_states=image_embeds,
            encoder_attention_mask=image_atts, return_dict=True)
        inputs_t5 = self.t5_proj_loc(query_output.last_hidden_state)
        atts_t5 = torch.ones(inputs_t5.size()[:-1], dtype=torch.long).to(image.device) 
            
        with torch.cuda.amp.autocast(dtype=torch.bfloat16):
            
            frame_prefix = self.t5_tokenizer(
                self.frame_prefix, padding="longest", add_special_tokens=False,
                truncation=True, max_length=self.max_txt_len, return_tensors="pt",
                ).to(image.device) # 
            #print('frame_prefix 1', frame_prefix.input_ids.shape) 8, 4
            frame_prefix_id = torch.repeat_interleave(frame_prefix.input_ids, b*t, 0)
            frame_prefix_mask = torch.repeat_interleave(frame_prefix.attention_mask, b*t, 0)
            # Question, Options input
            input_tokens = self.t5_tokenizer(
                text_input, padding="longest", truncation=True,
                max_length=self.max_txt_len, return_tensors="pt").to(image.device)
            input_ids = torch.repeat_interleave(input_tokens.input_ids, t, 0)
            input_attention_mask = torch.repeat_interleave(input_tokens.attention_mask, t, 0)

            frame_predix_embed = self.t5_model.encoder.embed_tokens(frame_prefix_id)
            inputs_embeds = self.t5_model.encoder.embed_tokens(input_ids)
            inputs_embeds = torch.cat([frame_predix_embed, inputs_t5, inputs_embeds], dim=1)
            encoder_atts = torch.cat([frame_prefix_mask, atts_t5, input_attention_mask], dim=1)

            outputs = self.t5_model.generate(
                inputs_embeds=inputs_embeds, attention_mask=encoder_atts,
                do_sample=use_nucleus_sampling, top_p=top_p,
                temperature=temperature, num_beams=1,
                max_new_tokens=max_length, min_length=min_length,
                repetition_penalty=repetition_penalty, length_penalty=length_penalty,
                num_return_sequences=num_captions, return_dict_in_generate=True,
                output_hidden_states=True, output_scores=True)
                # print('answer', answer)
            pred_logits = outputs.scores[0] #outputs_embed_qa.logits.detach()
            pred_logits = pred_logits[:, [self.no_id, self.yes_id]] # b, 5
            pred_yes_score = pred_logits[:, 1].cpu().tolist() 
            pred_ans = torch.argmax(pred_logits, dim=-1).cpu().tolist()
                     
        out['answer'] = flat_answer
        multiframe_qid = []
        for q in qid:
            for i in range(t):
                multiframe_qid.append(q)
                
        out['qid'] = multiframe_qid
        out['yes_score'] = pred_yes_score
        out['pred_ans'] = pred_ans 
        
        return out

    def predict_answers(
        self,
        samples,
        num_beams=5,
        inference_method="generate",
        max_len=10,
        min_len=1,
        num_ans_candidates=128,
        answer_list=None,
        prompt="",
        length_penalty=-1,
        **kwargs
    ):
        image = samples["image"]
        with torch.cuda.amp.autocast(enabled=(self.device != torch.device("cpu"))):
            image_embeds = self.ln_vision_loc(self.visual_encoder(image))
        image_atts = torch.ones(image_embeds.size()[:-1], dtype=torch.long).to(
            image.device
        )

        query_tokens = self.query_tokens.expand(image_embeds.shape[0], -1, -1)
        query_output = self.Qformer.bert(
            query_embeds=query_tokens,
            encoder_hidden_states=image_embeds,
            encoder_attention_mask=image_atts,
            return_dict=True,
        )

        inputs_t5 = self.t5_proj(query_output.last_hidden_state)
        atts_t5 = torch.ones(inputs_t5.size()[:-1], dtype=torch.long).to(image.device)

        if isinstance(samples["text_input"], str):
            samples["text_input"] = [samples["text_input"]]
        if prompt:
            text_input = [prompt.format(question) for question in samples["text_input"]]
        else:
            text_input = samples["text_input"]

        input_tokens = self.t5_tokenizer(
            text_input, padding="longest", return_tensors="pt"
        ).to(image.device)

        encoder_atts = torch.cat([atts_t5, input_tokens.attention_mask], dim=1)

        device_type = "cuda" if "cuda" in str(self.device) else "cpu"
        with torch.amp.autocast(device_type=device_type, dtype=torch.bfloat16):
            inputs_embeds = self.t5_model.encoder.embed_tokens(input_tokens.input_ids)
            inputs_embeds = torch.cat([inputs_t5, inputs_embeds], dim=1)

            outputs = self.t5_model.generate(
                inputs_embeds=inputs_embeds,
                attention_mask=encoder_atts,
                do_sample=False,
                num_beams=num_beams,
                max_new_tokens=max_len,
                min_length=min_len,
                length_penalty=length_penalty,
            )
            output_text = self.t5_tokenizer.batch_decode(
                outputs, skip_special_tokens=True
            )

        if self._apply_lemmatizer:
            output_text = self._lemmatize(output_text)

        return output_text

    def _lemmatize(self, answers):
        def apply(answer):
            doc = self.lemmatizer(answer)

            words = []
            for token in doc:
                if token.pos_ in ["NOUN", "VERB"]:
                    words.append(token.lemma_)
                else:
                    words.append(token.text)
            answer = " ".join(words)

            return answer

        return [apply(answer) for answer in answers]

    @property
    def lemmatizer(self):
        if self._lemmatizer is None:
            try:
                import spacy

                self._lemmatizer = spacy.load("en_core_web_sm")
            except ImportError:
                logging.error(
                    """
                    Please install spacy and en_core_web_sm model to apply lemmatization.
                    python -m spacy download en_core_web_sm
                    OR
                    import spacy.cli
                    spacy.cli.download("en_core_web_sm")
                    """
                )
                exit(1)

        return self._lemmatizer

    @classmethod
    def from_config(cls, cfg):
        img_size = cfg.get("image_size")
        num_query_token = cfg.get("num_query_token")
        t5_model = cfg.get("t5_model")

        drop_path_rate = cfg.get("drop_path_rate", 0)
        use_grad_checkpoint = cfg.get("use_grad_checkpoint", False)
        vit_precision = cfg.get("vit_precision", "fp16")
        freeze_vit = cfg.get("freeze_vit", True)

        prompt = cfg.get("prompt", "")
        max_txt_len = cfg.get("max_txt_len", 32)
        frame_num = cfg.get("frame_num", 8)
        answer_num = cfg.get("answer_num", 5) 
        apply_lemmatizer = cfg.get("apply_lemmatizer", False)
        task = cfg.get("task", 'train_loc_freeze_qa')

        model = cls(
            img_size=img_size,
            drop_path_rate=drop_path_rate,
            use_grad_checkpoint=use_grad_checkpoint,
            vit_precision=vit_precision,
            freeze_vit=freeze_vit,
            num_query_token=num_query_token,
            t5_model=t5_model,
            prompt=prompt,
            max_txt_len=max_txt_len,
            apply_lemmatizer=apply_lemmatizer,
            frame_num=frame_num,
            answer_num=answer_num,
            task=task,
        )
        model.load_checkpoint_from_config(cfg)
        # if 'pretrain_loc' in task:
        # model.load_qformer_loc()

        return model