# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.


import json
import os
import subprocess

from loguru import logger as log

# pyrefly: ignore  # import-error
from cosmos_gradio.gradio_app.gradio_util import (
    get_output_folder,
    get_outputs,
)


class GradioApp2Cli:
    """Most basic server using an existing CLI model.
    The parameter validation is left to the CLI app.
    This server has no communication channel with the workers, so no errors are reported.
    """

    def __init__(
        self,
        cli_cmd: str,
        num_workers: int = 8,
        checkpoint_dir: str = "checkpoints",
        output_dir: str = "outputs",
    ):
        self.cli_cmd = cli_cmd
        self.num_workers = num_workers
        self.checkpoint_dir = checkpoint_dir
        self.process = None
        self.output_dir = output_dir
        self._setup_environment()

    def _setup_environment(self):
        self.env = os.environ.copy()

    def infer_dict(self, args: dict, output_dir=None):
        output_dir = get_output_folder(self.output_dir)

        log.info(f"torchrun nprocs={self.num_workers} with CLI command={self.cli_cmd}")

        log.debug(json.dumps(args, indent=2))
        # Save arguments to JSON file in output_dir
        os.makedirs(output_dir, exist_ok=True)
        args_file = os.path.join(output_dir, "inference_args.json")
        with open(args_file, "w") as f:
            json.dump(args, f, indent=2)
        log.info(f"Saved inference arguments to {args_file}")

        torchrun_cmd = [
            "torchrun",
            f"--nproc_per_node={self.num_workers}",
            "--nnodes=1",
            "--node_rank=0",
            self.cli_cmd,
            f"--num_gpus={self.num_workers}",
            f"--controlnet_specs={args_file}",
            f"--checkpoint_dir={self.checkpoint_dir}",  # specific to transfer1
            f"--video_save_folder={output_dir}",
        ]

        log.info(f"Running command: {' '.join(torchrun_cmd)}")

        # Launch worker processes
        try:
            # pyrefly: ignore  # bad-assignment
            self.process = subprocess.Popen(
                torchrun_cmd,
                env=self.env,
                text=True,
                bufsize=1,
                universal_newlines=True,
            )

            # Wait for the process to complete
            # pyrefly: ignore  # missing-attribute
            return_code = self.process.wait()

            if return_code == 0:
                log.info("Inference completed successfully")
            else:
                log.error(f"Inference failed with return code: {return_code}")
                raise subprocess.CalledProcessError(return_code, torchrun_cmd)

        except Exception as e:
            log.error(f"Error running inference: {e}")
            raise e

        return get_outputs(output_dir)

    def infer(
        self,
        request_text,
        output_folder=None,
    ):
        try:
            request_data = json.loads(request_text)
        except json.JSONDecodeError as e:
            return (
                None,
                f"Error parsing request JSON: {e}\nPlease ensure your request is valid JSON.",
            )

        return self.infer_dict(request_data, output_folder)