{txt-to-video}——文本转视频思路

思路：

.txt -> 读取 -> 调用api合成语音 -> 合成语音 -> MP3+PNG/JPG -> 合成视频

目录结构

• new_video：存放最后生产的视频文件
• out_v：存放合成的MP3文件
• text：存放文本文件
• voices：存放api导出的MP3

代码结构：

import os
import time
import subprocess
from pydub import AudioSegment
from aip import AipSpeech
import pandas as pd
pd.set_option( 'display.max_columns', None)
pd.set_option( 'display.max_rows', None)
PATH = os.path.abspath(__file__).replace(os.path.abspath(__file__).split('\\')[-1], '')

def text_to_voice(text,file_name,r_num):
    print(len(text))
    text_list = []
    out_max = 500
    if len(text)> out_max:
        text_list.append(text[:out_max])
        text_list.append(text[out_max:])
        while True:
            if len(text_list[-1])>out_max:
                text_re = text_list[-1]
                text_list.remove(text_re)
                text_list.append(text_re[:out_max])
                text_list.append(text_re[out_max:])
            else:
                break
    print(len(text_list))
    file_name_list = []
    for i in range(0,len(text_list)):
        APP_ID = 'xxx'
        API_KEY = 'xxx'
        SECRET_KEY = 'xxx'
        try:
            client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)
            result = client.synthesis(text_list[i], 'zh', 1, {
                'vol': 5,
                'per': 3,
                'spd': 6
            })
            print(f"::{file_name}::{i}::{result}")
        except Exception as e:
            print(e)
        if not isinstance(result, dict):
            with open(f'voices/{r_num}_{i}.mp3', 'wb') as f:
                f.write(result)
        time.sleep(2)

        file_name_list.append(f'{r_num}_{i}.mp3')

    f = open(f'voices/{r_num}_ff.txt','a')
    for i in file_name_list:
        f.write(f'file {i}\n')
    f.close()
    return f'{r_num}_ff.txt'


def add_mp3(dir_name,file_name):
    f = open(f"{dir_name}/{file_name}",encoding='utf-8')
    ret = f.readlines()
    f.close()
    outfile_name = ret[0].replace('.mp3','_NEW.mp3').replace('file ','').replace('\n','')
    cmd = f'{PATH}{dir_name}\\ffmpeg.exe -f concat -safe 0'
    cmd+= f' -i {PATH}{dir_name}\\{file_name} -c copy {PATH}out_v\\{outfile_name}'
    print(cmd)
    subprocess.call(cmd,shell=True)


def filter_text(text_src):
    f= open(text_src, encoding='utf-8')
    ret = f.readlines()
    ret_out = []
    for i in ret:
        i = i.strip().replace('\n','')
        if i != '':
            ret_out.append(i)
    f.close()
    return ret_out


def filter_text_1(text_src):
    f = open(text_src, encoding='utf-8')
    ret = f.readlines()
    ret_out = []
    for i in ret:
        i = i.strip().replace('\n', '').replace('|','')
        if i != '':
            ret_out.append(i)
    f.close()
    ret_dict = {}
    str1 = ''
    for i in ret_out:
        if '===' in i:
            title = i.replace('===','').replace(' ','_')
            str1 = ''
            continue
        str1 += i
        ret_dict[title]=str1
    for i in ret_dict:
        print(i)
        print(len(ret_dict[i]))
    return ret_dict


def to_video(src):
    ff = ['ffmpeg.exe','back.png']
    l_r = [i for num, i in enumerate(os.listdir(src)) if num not in ff]
    print(l_r)
    for c in l_r:
        out_name = c.replace('.mp3','.mp4')
        cmd = f'{PATH}voices\\ffmpeg.exe -i {PATH}out_v\\{c} -i {PATH}out_v\\back.png' \
              f'  -acodec aac -strict -2 -vcodec libx264 -ar 22050 -ab 128k -ac 2 -y {PATH}new_video\\{out_name}'
        print(cmd)
        subprocess.call(cmd, shell=True)


def to_video_one(dir_name,vioce_name):
    out_name = vioce_name.replace('.mp3','.mp4')
    cmd = f'{PATH}{dir_name}\\ffmpeg.exe -i {PATH}{dir_name}\\{vioce_name} -i {PATH}{dir_name}\\back.png' \
          f'  -acodec aac -strict -2 -vcodec libx264 -ar 22050 -ab 128k -ac 2 -y {PATH}new_video\\{out_name}'
    print(cmd)
    subprocess.call(cmd, shell=True)

if __name__ == '__main__':
    # ret_dict = filter_text_1('text/1.txt')
    # k = 1
    # txt_file_list=[]
    # for i in ret_dict:
    #     if k>187:
    #         file_name = text_to_voice(ret_dict[i],i,k)
    #         txt_file_list.append(file_name)
    #     k +=1
    # print(txt_file_list)
    #---------------------------
    # f = open('out_v/full.txt','a')
    # for i in range(1,263):
    #     f.write(f'file {i}_0_NEW.mp3\n')
    # f.close()
    # add_mp3('out_v','full.txt')
    #-----------------------------
    # txt_file_list = []
    # for i in range(1,263):
    #     txt_file_list.append(f'{i}_ff.txt')
    # print(txt_file_list)
    # for i in txt_file_list:
    #     add_mp3('voices',i)
    # to_video('out_v/')
    #-------------------
    # to_video_one('out_v','1_0_NEW_NEW.mp3')
    #-------------------