# Copyright (c) Microsoft
#               2022 Chengdong Liang (liangchengdong@mail.nwpu.edu.cn)
#
# 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 gradio as gr
import wespeaker
# import wespeakerruntime as wespeaker
# from sklearn.metrics.pairwise import cosine_similarity

STYLE = """
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bootstrap@5.1.3/dist/css/bootstrap.min.css" integrity="sha256-YvdLHPgkqJ8DVUxjjnGVlMMJtNimJ6dYkowFFvp4kKs=" crossorigin="anonymous">
"""
OUTPUT_OK = (STYLE + """
    <div class="container">
        <div class="row"><h1 style="text-align: center">The speakers are</h1></div>
        <div class="row"><h1 class="display-1 text-success" style="text-align: center">{:.1f}%</h1></div>
        <div class="row"><h1 style="text-align: center">similar</h1></div>
        <div class="row"><h1 class="text-success" style="text-align: center">Welcome, human!</h1></div>
        <div class="row"><small style="text-align: center">(You must get at least 70% to be considered the same person)</small></div>
    </div>
""")
OUTPUT_FAIL = (STYLE + """
    <div class="container">
        <div class="row"><h1 style="text-align: center">The speakers are</h1></div>
        <div class="row"><h1 class="display-1 text-danger" style="text-align: center">{:.1f}%</h1></div>
        <div class="row"><h1 style="text-align: center">similar</h1></div>
        <div class="row"><h1 class="text-danger" style="text-align: center">Warning! stranger!</h1></div>
        <div class="row"><small style="text-align: center">(You must get at least 70% to be considered the same person)</small></div>
    </div>
""")

OUTPUT_ERROR = (STYLE + """
    <div class="container">
        <div class="row"><h1 style="text-align: center">Input Error</h1></div>
        <div class="row"><h1 class="text-danger" style="text-align: center">{}!</h1></div>
    </div>
""")

cn_model = wespeaker.load_model("chinese")
en_model = wespeaker.load_model("english")


def speaker_verification(audio_path1, audio_path2, lang='CN'):
    if audio_path1 == None or audio_path2 == None:
        output = OUTPUT_ERROR.format('Please enter two audios')
        return output
    if lang == 'EN':
        model = cn_model
    elif lang == 'CN':
        model = en_model
    else:
        output = OUTPUT_ERROR.format('Please select a language')
        return output
    # emb1 = model.extract_embedding(audio_path1)
    # emb2 = model.extract_embedding(audio_path2)
    # cos_score = model.compute_cosine_score(emb1[0], emb2[0])
    # cos_score = (cos_score + 1) / 2.0
    cos_score = wespeaker.compute_similarity(audio_path1, audio_path2)

    if cos_score >= 0.70:
        output = OUTPUT_OK.format(cos_score * 100)
    else:
        output = OUTPUT_FAIL.format(cos_score * 100)

    return output


# input
inputs = [
    gr.inputs.Audio(source="microphone",
                    type="filepath",
                    optional=True,
                    label='Speaker#1'),
    gr.inputs.Audio(source="microphone",
                    type="filepath",
                    optional=True,
                    label='Speaker#2'),
    gr.Radio(['EN', 'CN'], label='Language'),
]

output = gr.outputs.HTML(label="")

# description
description = ("<p>WeSpeaker Demo ! Try it with your own voice ! Note: We recommend that the audio length be greater than 5s !</p>"
               "<p>CN: cnceleb_resnet34.onnx. EN: voxceleb_resnet34.onnx</p>")

article = (
    "<p style='text-align: center'>"
    "<a href='https://github.com/wenet-e2e/wespeaker' target='_blank'>Github: Learn more about WeSpeaker</a>"
    "</p>")

examples = [
    ['examples/BAC009S0764W0228.wav', 'examples/BAC009S0764W0328.wav', 'CN'],
    ['examples/BAC009S0913W0133.wav', 'examples/BAC009S0764W0228.wav', 'CN'],
    ['examples/00001_spk1.wav', 'examples/00003_spk2.wav', 'EN'],
    ['examples/00010_spk2.wav', 'examples/00024_spk1.wav', 'EN'],
    ['examples/00001_spk1.wav', 'examples/00024_spk1.wav', 'EN'],
    ['examples/00010_spk2.wav', 'examples/00003_spk2.wav', 'EN'],
]

interface = gr.Interface(
    fn=speaker_verification,
    inputs=inputs,
    outputs=output,
    title="Speaker Verification in WeSpeaker : 基于 WeSpeaker 的说话人确认",
    description=description,
    article=article,
    examples=examples,
    theme="huggingface",
)

interface.launch(enable_queue=True)