deffer.py

from datetime import datetime
import math
import os
import re
import shutil
import subprocess
import requests
import json
import base64
import random
import string
import time
from gtts import gTTS
from moviepy.editor import *
from pydub import AudioSegment
from PIL import Image
import edge_tts
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
import hashlib
from nltk import *
import subprocess
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize
# nltk.download('punkt')
# nltk.download('stopwords')
# nltk.download('averaged_perceptron_tagger')
from moviepy.editor import ImageClip, VideoFileClip
from config import *



# 分割内容
# 保存到文件
# 文本
# 
def spliter(novel_text):
    # 创建一个名为"txt文件夹"的文件夹（如果不存在）
    if not os.path.exists("split_txt"):
        os.makedirs("split_txt")

    novel_text=clip_to_whole(novel_text)
    # 按句子分割文本
    sentences = re.split(r'(?<=[.。！？])', novel_text)

    # 合并相邻的句子
    merged_sentences = []
    current_sentence = ""
    for sentence in sentences:
        current_sentence += sentence
        if len(current_sentence) > 50:  # 合并后的句子长度超过50个字符时分割
            merged_sentences.append(current_sentence)
            current_sentence = ""
    if current_sentence:
        merged_sentences.append(current_sentence)

    # 使用TF-IDF计算句子相似度
    tfidf_vectorizer = TfidfVectorizer()
    tfidf_matrix = tfidf_vectorizer.fit_transform(merged_sentences)
    similarities = cosine_similarity(tfidf_matrix)

    similarity_threshold=0.7

    # 将相似句子组合成场景
    scenes = []
    current_scene = []
    for i in range(len(merged_sentences)):
        if not current_scene:
            current_scene.append(i)
        else:
            if similarities[current_scene[0]][i] > similarity_threshold:
                current_scene.append(i)
            else:
                scenes.append(current_scene)
                current_scene = [i]

    # 保存每个段落到文件
    for i, scene in enumerate(scenes):
        scene_sentences = [merged_sentences[i] for i in scene]
        scene_text = "".join(scene_sentences)
        filename = f"split_txt/{i+1}.txt"
        with open(filename, 'w', encoding='utf-8') as scene_file:
            scene_file.write(scene_text)
        print(scene_text)

    # 处理最后一段文本,否则会有丢失的情况
    if current_scene:
        last_scene_text = "".join([merged_sentences[i] for i in current_scene])
        last_filename = f"split_txt/{len(scenes)+1}.txt"
        with open(last_filename, 'w', encoding='utf-8') as last_scene_file:
            last_scene_file.write(last_scene_text)
        print(last_scene_text)



# add
def clip_to_whole(content):
    res_content = content.replace('"', "'").strip().replace("\n", "")
    return res_content



# 只是去掉了换行,二维
def text_clip_to_whole(contents):
    res_contents = []
    for content in contents:
        res_content=clip_to_whole(content)
        res_contents.append(res_content)
    return res_contents







# 返回的是二维数组
def text_to_voice(contents):
    try:
        shutil.rmtree('./tts')
        os.mkdir('./tts')
    except:
        pass

    voice = []

    first=1
    for lines in contents:
        filenames = []
        second=1
        for line in lines:
            # 执行相应的处理操作，例如：
            # line = line.replace('"', "'").strip().replace("\n", "")
            # filename = gtts(line)
            filename = _edge_tts(first, second, line)
            savefile(line,'./tts/'+str(first)+'.'+str(second)+'.txt')
            filenames.append(filename)
            second+=1
        first += 1
        voice.append(filenames)


# 获取音频文件名
# 0
# 二维
# def loadvoice():
#     result = []
#     current_dir = []
#     base_path = "./tts"
#     with os.scandir(base_path) as entries:
#         for entry in entries:
#             if entry.is_file() and entry.name.endswith(".mp3"):
#                 current_dir.append(os.path.join("tts", entry.name))

#     current_dir.sort()
#     result.append(current_dir)
    
#     return result



#add
# 转换为音频
def _edge_tts(first,second,content):
    # voice = 'zh-CN-YunxiNeural'
    # rate = '-4%'
    # volume = '+0%'
    # filename = './tts/%s.mp3' % content
    # tts = edge_tts.Communicate(text=content, voice=voice, rate=rate, volume=volume)
    # tts.save(filename)
    filename = 'tts/'+str(first)+"."+str(second)+'.mp3'
    cmd = 'edge-tts --voice zh-CN-YunxiNeural --text "' + content + '" --write-media %s' % filename
    res = subprocess.call(cmd, shell=True)

    sound = AudioSegment.from_mp3(filename)
    duration = int(sound.duration_seconds)
    update_name = '%s.%s_duration_%f.mp3' % (filename.split('.')[0],filename.split('.')[1], duration)
    os.rename(filename, update_name)
    return update_name




# 文本生成图片
# 一维数组
# 
def text_to_pic(contents):
    i=1
    for content in contents:
        getTxt2Image(','.join(content),ADDER_PROMOTE,"./",i)
        i=i+1



#add
def getTxt2Image(text_prompt,add_prompt,localPythonExePath,num):#文生图核心方法#localPythonExePath是本地存储路径
    serverName = "server001"
    txt2img_url = Stable_diff_api+'/sdapi/v1/txt2img'
    print("提示词："+text_prompt)
    data = {'prompt': text_prompt+add_prompt+'(masterpiece:1,2), best quality, highres, original, extremely detailed wallpaper, perfect lighting,(extremely detailed CG:1.2),',
            'brach_size': 1,
        #    "sampler_name": "DPM++ 2m Karras",#可自行选择
           "steps": 20,
           "cfg_scale": 8, #提示词相关性
           "width" : 740,
           "height": 580,
           "restore_faces": "true",
           "negative_prompt": "NSFW, (worst quality:2), (low quality:2), (normal quality:2), lowres, normal quality, ((monochrome)), ((grayscale)), skin spots, acnes, skin blemishes, age spot, (ugly:1.331), (duplicate:1.331), (morbid:1.21), (mutilated:1.21), (tranny:1.331), mutated hands, (poorly drawn hands:1.5), blurry, (bad anatomy:1.21), (bad proportions:1.331), extra limbs, (disfigured:1.331), (missing arms:1.331), (extra legs:1.331), (fused fingers:1.61051), (too many fingers:1.61051), (unclear eyes:1.331), lowers, bad hands, missing fingers, extra digit,bad hands, missing fingers, (((extra arms and legs))),"
           }
    data.update(Stable_diff_conf)
    response = submit_post(txt2img_url, data)
    save_encoded_image(response.json()['images'][0], ".\pictures\image"+str(num)+".png")
    print("图片已经生成,并保存在pictures目录中")


#add
def save_encoded_image(b64_image: str, output_path: str):
   with open(output_path, "wb") as image_file:
       image_file.write(base64.b64decode(b64_image))

#add
def submit_post(url: str, data: dict):#发出Post请求
   return requests.post(url, data=json.dumps(data))




# 生成视频
# 传入两个二维数组
# 
def handle_movie(voice,txts):
    final_duration_clips = []
    txts=txts
    voice=voice
    print(len(voice))
    print(len(txts))

    for item in voice:
        duration_clips = []
        for line in item:
            filename = str(line.split('/')[-1])
            tmp = filename.split('_')
            # 获取语音文件名中的时长
            duration_clips.append(float(tmp[-1].replace('.mp3', '')))
        final_duration_clips.append(duration_clips)


    final_clip_all=None
    for i in range(len(txts)):

        # final_duration_clips[i]
        # txts[i]
        # voice[i]

        # 解说 + 字幕
        final_clip=None
        for j in range(len(txts[i])):

            # 计算当前行的起始时间和结束时间

            print("="*50)
            print("i, j ---> " + str(i) + ", " + str(j))
            print(voice[i][j])
            print(txts[i][j])
            print(final_duration_clips[i][j])
            print(str(i+1)+".png")


            start = 0
            end = start + final_duration_clips[i][j]

            print("strat:"+str(start))
            print("strat:"+str(end))

            print("="*50)


            # 添加语音剪辑
            voice_clip = AudioFileClip("./"+voice[i][j]).set_start(start).volumex(1.2)

            # 添加字幕剪辑
            screensize = (700,200)
            # txt_clip = TextClip(txts[i][j], fontsize=30, color='black', font='MaoKenWangXingYuan.ttf', stroke_color='black',stroke_width=0.3,size = screensize, method='caption')
            txt_clip = TextClip(txts[i][j], fontsize=30, color='white', font='MaoKenWangXingYuan.ttf',size = screensize, method='caption')
            txt_clip = txt_clip.set_pos('bottom').set_duration(final_duration_clips[i][j])
            txt_clip = txt_clip.set_start(start).set_end(end)

            # 计算视频中每张图片的剪辑长度
            # 定义镜头焦点效果
            def resize_func(t):
                # return 1 + 0.001*t
                return 1
            # 添加每张图片的剪辑，并设置位置、时长、帧率等属性
            image_clip = ImageClip("./pictures/image"+str(i+1)+".png").set_duration(final_duration_clips[i][j]).resize(resize_func).set_position(('center', 'center')).set_fps(25)
            # image_clip = image_clip.crossfadein(1).crossfadeout(1)


            # 将图像、字幕和音频剪辑合并为一个剪辑，并添加淡入淡出效果
            video_clip = CompositeVideoClip([image_clip,txt_clip])
            video_clip = video_clip.set_audio(voice_clip)
            if final_clip is None:  # 添加此行代码
                final_clip = video_clip  # 添加此行代码
            else:
                final_clip = concatenate_videoclips([final_clip,video_clip])


        if final_clip_all is None:  # 添加此行代码
            final_clip_all = final_clip  # 添加此行代码
        # 这一行使后来加的
        else:
            final_clip_all = concatenate_videoclips([final_clip_all,final_clip])

        # 导出视频剪辑
        # final_clip.write_videofile("video/"+str(i+1)+".mp4", fps=25)


    current_time = datetime.now()
    time_string = current_time.strftime("%Y_%m_%d_%H_%M_%S")
    print(time_string)

    # 这一行使后来加的
    final_clip_all.write_videofile("video/"+time_string+".mp4", fps=25)
    print("*"*50)
    print("完成！")




















# 内容翻译
# 一维数组
# 一维数组
def translate(fileContents):
    resultered=[]
    for fileContent in fileContents:
        textTranslatedtoEnglish=json.loads(translater(fileContent))
        resulter=textTranslatedtoEnglish['trans_result']
        resulter=handleResult(resulter)
        resultered.append(resulter)
    return resultered

# 
# 
# 
def translater(query):
    # Set your own appid/appkey.
    appid = '20230925001829106'
    appkey = 'GibMVhrk0dwgvV6Lt_iD'

    # For list of language codes, please refer to `https://api.fanyi.baidu.com/doc/21`
    from_lang = 'zh'
    to_lang =  'en'

    endpoint = 'http://api.fanyi.baidu.com'
    path = '/api/trans/vip/translate'
    url = endpoint + path

    # Generate salt and sign
    def make_md5(s, encoding='utf-8'):
        return hashlib.md5(s.encode(encoding)).hexdigest()

    salt = random.randint(32768, 65536)
    sign = make_md5(appid + query + str(salt) + appkey)

    # Build request
    headers = {'Content-Type': 'application/x-www-form-urlencoded'}
    payload = {'appid': appid, 'q': query, 'from': from_lang, 'to': to_lang, 'salt': salt, 'sign': sign}

    # Send request
    r = requests.post(url, params=payload, headers=headers)
    result = r.json()

    time.sleep(0.2)

    # Show response
    return (json.dumps(result, indent=4, ensure_ascii=False))

# add
def handleResult(resulter):
    returned = ""  # 初始化一个空字符串用于存储结果
    for i in resulter:
        # resulter中的每个元素（假设是一个字典）中的dst值取出来，用逗号连接起来，最终返回一个字符串。
        returned += i['dst']+","  # 修正这里的索引
    return returned






# 提取关键词
# 一维数组
# 二位数组
def use_extract_keywords(texts):
    re_texts=[]
    for text in texts:
        re_text=extract_keywords(text)
        re_texts.append(re_text)
    return re_texts

#add
def extract_keywords(text):
    # 将文本拆分成单词
    words = text.split()

    # 创建一个用于存储短语的列表
    phrases = []

    # 创建一个集合用于跟踪已经出现过的单词
    seen_words = set()
    n=3
    # 生成n-gram短语
    for i in range(len(words) - n + 1):
        phrase = ' '.join(words[i:i + n])
        # 检查短语中是否有已经出现的单词
        if not any(word in seen_words for word in phrase.split()):
            phrases.append(phrase)
            # 将短语中的单词添加到已经出现过的单词集合中
            seen_words.update(phrase.split())

    return phrases



# 将每个段落拆分成句子,一维变二维
# 一维数组
# 二维
def dozen_txt_double(contents):
    doublelist=[]
    for content in contents:
        innerlist = []
        lines=split_sentences(content)
        for line in lines:
            innerlist.append(line)
        doublelist.append(innerlist)
    return doublelist

# add
# 将每个段落拆分成句子

def split_sentences(text):
    # 去除文本中的引号
    text = text.replace('"', '').replace('“', '').replace('”', '').replace('‘', '').replace('’', '')
    # 将文本按照换行符进行切割
    lines = text.split('\n')
    # 去除每行开头和结尾的空格，并去除空行
    lines = [line.strip() for line in lines if line.strip()]
    # 将每个段落拆分成句子
    sentences = []
    for line in lines:
        # 按照句号、问号、感叹号等标点符号进行拆分
        sentences.extend(re.split(r'[.!?。？！]+', line))

    # 去除句子开头和结尾的空格，并去除空句子
    sentences = [sentence.strip() for sentence in sentences if sentence.strip()]
    # 返回结果
    return sentences
















def delete_files_in_folder(folder_path):
    try:
        # 获取文件夹中的所有文件列表
        file_list = os.listdir(folder_path)

        # 遍历文件列表并删除每个文件
        for filename in file_list:
            file_path = os.path.join(folder_path, filename)
            if os.path.isfile(file_path):
                os.remove(file_path)

        print("所有文件已成功删除。")
    except Exception as e:
        print("发生错误：{str(e)}")


def delete_file(file_name):
    if os.path.exists(file_name):
        os.remove(file_name)
        print("成功删除文件"+file_name)
    else:
        print("文件\""+file_name+"\"不存在，已经忽略")



def get_all_pic_files(path):
    files = []
    for i in range(1, get_cont()+1):
        filename = f"{i}.png"
        file_path = os.path.join(path, filename)
        if os.path.exists(file_path):
            files.append(file_path)
    return files




def savefile(content,name):
    # 打开文件，如果文件不存在将会创建新文件
    file = open(name, "w",encoding='utf-8')
    # 写入内容到文件
    file.write(content)
    # 关闭文件
    file.close()



# def load_voice():

#     # 指定文件夹路径
#     folder_path = './tts'

#     # 获取文件夹中所有以.mp3结尾的文件
#     mp3_files = [filename for filename in os.listdir(folder_path) if filename.endswith('.mp3')]

#     # 定义一个自定义排序函数，按照文件名中的数字部分进行排序
#     def custom_sort(filename):
#         # 提取文件名中的数字部分作为排序关键字
#         match = re.search(r'\d+', filename)
#         if match:
#             return int(match.group())
#         else:
#             return 0  # 如果没有找到数字部分，默认为0

#     # 对文件名进行排序
#     sorted_files = sorted(mp3_files, key=custom_sort)

#     return sorted_files





def load_voice():
    # 指定文件夹路径
    folder_path = './tts'

    # 获取文件夹中所有以.mp3结尾的文件
    mp3_files = [filename for filename in os.listdir(folder_path) if filename.endswith('.mp3')]

    # 定义一个自定义排序函数，按照文件名中的数字部分进行排序
    def custom_sort(filename):
        # 提取文件名中的数字部分作为排序关键字
        match = re.search(r'\d+\.\d+', filename)
        if match:
            return tuple(map(int, match.group().split('.')))
        else:
            return (0, 0)  # 如果没有找到数字部分，默认为(0,0)

    # 对文件名进行排序
    sorted_files = sorted(mp3_files, key=custom_sort)

    # 将文件名按照行号分组，存储在二维列表中
    voice = []
    for filename in sorted_files:
        match = re.search(r'(\d+)\.\d+', filename)
        if match:
            line_num = int(match.group(1))
            while line_num > len(voice):
                voice.append([])
            voice[-1].append(folder_path+'/'+filename)
    return voice






# 获取文本
# 0
# 一维
def loadtxt():
    contents=[]
    for i in range(1,get_cont()+1):
        helloFile=open("split_txt/"+str(i)+".txt",encoding='utf-8')
        fileContent=helloFile.read()
        contents.append(fileContent)
        helloFile.close()
    return contents


# 获取章节数
def get_cont():
    cont=0
    with os.scandir('split_txt') as entries:
        for entry in entries:
            if entry.is_file():
                cont=cont+1
    return cont



# 读取小说文本
# 文件名
# 文本
def get_whole_file(filename):
    with open(filename, 'r', encoding='utf-8') as file:
        novel_text = file.read()
    return novel_text


# def get_cont2():
#     cont=0
#     with os.scandir('video') as entries:
#         for entry in entries:
#             if entry.is_file():
#                 cont=cont+1
#     return cont



# 二维list


def double_txt(contents):
    doublelist=[]
    for content in contents:
        lines=split_sentences(content)
        innerlist = []
        for line in lines:
            innerlist.append(line)
        doublelist.append(innerlist)
    return doublelist






# 重构图片
# 序号 附加的关键词
# 0
def re_create_pic(num,prom):
    delete_file('image'+str(num)+".png")
    getTxt2Image(translater(get_txt_by_num(num)),prom,"./",num)
    # getTxt2Image(translater(clip_to_whole(get_txt_by_num(num))),prom,"./",num)

#add
def get_txt_by_num(num):
    helloFile=open("split_txt/"+str(num)+".txt",encoding='utf-8')
    fileContent=helloFile.read()
    helloFile.close()
    return fileContent




def open_file_content():
    os.startfile('content.txt')



def open_floder_video():
    os.system(f'start explorer "video"')



def open_floder_pic():
    os.system(f'start explorer "pictures"')