"""
This file contains primitives for multi-gpu communication.
This is useful when doing distributed training.
"""
import os
import pickle
import time
import functools
import sys
import json

import torch

def is_sm_dist():
    """Check if environment variables are set for Sagemaker Data Distributed
    This has not been tested
    """
    sm_framework_params = os.environ.get('SM_FRAMEWORK_PARAMS', None)
    if not sm_framework_params:
        return False
    sm_framework_params = json.loads(sm_framework_params)
    return bool(sm_framework_params.get('sagemaker_distributed_dataparallel_enabled', False))

def is_sm():
    """Check if running in a Sagemaker Training instance
    """
    sm_framework_params = os.environ.get('SM_FRAMEWORK_PARAMS', None)
    if sm_framework_params:
        return True
    return False

def is_herring():
    """When running on a Herring cluster, check if using smddprun
    """
    herring_env = os.environ.get('HERRING', None)
    if herring_env is None:
        return False
    return True

# Temporary solution to determine if we're using SM data parallel
use_herring = is_sm_dist() or is_herring()
if use_herring:
    import smdistributed.dataparallel.torch.distributed as dist
else:
    import torch.distributed as dist

def get_world_size():
    if not dist.is_available():
        return 1
    if not dist.is_initialized():
        return 1
    return dist.get_world_size()


def get_rank():
    if not dist.is_available():
        return 0
    if not dist.is_initialized():
        return 0
    return dist.get_rank()

def is_main_process():
    return get_rank() == 0

def master_only(func):
    @functools.wraps(func)
    def wrapper(*args, **kwargs):
        rank = get_rank()
        if rank == 0:
            return func(*args, **kwargs)
    return wrapper

def synchronize():
    """
    Helper function to synchronize (barrier) among all processes when
    using distributed training
    """
    '''if not dist.is_available():
        return
    if not dist.is_initialized():
        return'''
    world_size = get_world_size()
    if world_size == 1:
        return
    dist.barrier()

def all_gather(data):
    """
    Run all_gather on arbitrary picklable data (not necessarily tensors)
    Args:
        data: any picklable object
    Returns:
        list[data]: list of data gathered from each rank
    """
    world_size = get_world_size()
    if world_size == 1:
        return [data]

    # serialized to a Tensor
    buffer = pickle.dumps(data)
    storage = torch.ByteStorage.from_buffer(buffer)
    tensor = torch.ByteTensor(storage).to("cuda")

    # obtain Tensor size of each rank
    local_size = torch.IntTensor([tensor.numel()]).to("cuda")
    size_list = [torch.IntTensor([0]).to("cuda") for _ in range(world_size)]
    dist.all_gather(size_list, local_size)
    size_list = [int(size.item()) for size in size_list]
    max_size = max(size_list)

    # receiving Tensor from all ranks
    # we pad the tensor because torch all_gather does not support
    # gathering tensors of different shapes
    tensor_list = []
    for _ in size_list:
        tensor_list.append(torch.ByteTensor(size=(max_size,)).to("cuda"))
    if local_size != max_size:
        padding = torch.ByteTensor(size=(max_size - local_size,)).to("cuda")
        tensor = torch.cat((tensor, padding), dim=0)
    dist.all_gather(tensor_list, tensor)

    data_list = []
    for size, tensor in zip(size_list, tensor_list):
        buffer = tensor.cpu().numpy().tobytes()[:size]
        data_list.append(pickle.loads(buffer))

    return data_list

def reduce_dict(input_dict, average=True):
    """
    Args:
        input_dict (dict): all the values will be reduced
        average (bool): whether to do average or sum
    Reduce the values in the dictionary from all processes so that process with rank
    0 has the averaged results. Returns a dict with the same fields as
    input_dict, after reduction.
    TODO: Test this
    """
    world_size = get_world_size()
    if world_size < 2:
        return input_dict
    with torch.no_grad():
        names = []
        values = []
        # sort the keys so that they are consistent across processes
        for k in sorted(input_dict.keys()):
            names.append(k)
            values.append(input_dict[k])
        values = torch.stack(values, dim=0)
        dist.reduce(values, dst=0)
        if get_rank() == 0 and average:
            # only main process gets accumulated, so only divide by
            # world_size in this case
            values /= world_size
        reduced_dict = {k: v for k, v in zip(names, values)}
    return reduced_dict