Python实现剪映草稿

​
首先，剪映官网并没有提供相应的API和SDK等解决方案，于是乎去找有没有对应的开源项目，找过csdn、gitee、github等平台，最终在gitee上找到了自己想要的解决方案：剪映草稿生成: 一个把数据保存到剪映草稿的小插件
在github上也有相关的解决方案，但是没有添加字幕，这个作者对代码进行了重写，实现了添加字幕的功能，博主本人主要是写Java代码的，本来想看看有没有对应的Java解决方案，但是找遍了全网都没有，看来只能通过Java去调用Python的接口去实现这个功能了，所以针对这个源码需要改写

如何把项目跑起来

下载完源码之后会看到一个 pyproject.toml 文件

[tool.poetry]
name = "JianYingDraft.PY"
version = "1.0.3"
description = "帮助剪映快速生成草稿的方案"
authors = ["聂长安 <1351261434@qq.com>"]
readme = "README.md"
packages = [{ include = "JianYingDraft" }]
include = ["README.md", "LICENSE"]
license = "MIT"

[tool.poetry.dependencies]
python = "^3.12"
pymediainfo = "^6.1.0"
basiclibrary-py = "^0.6.10"
pytest = "^8.1.1"


[tool.poetry.group.dev.dependencies]
pytest = "^8.1.1"
pylint = "^3.1.0"

[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"

虽然我不知道这个文件是啥，但是我看着有点类似Java的项目的pom.xml文件或者Vue前端的package.json文件,这里是用来管理依赖包版本的,所以这个包肯定有用，顺着这个思路去百度，找到了这样的一段话

越来越多的依赖管理工具开始倾向于使用pyproject.toml。例如poetry，它会把项目依赖信息写在此文件中，取代传统的requirements.txt，让依赖管理更加规范统一

看来这个依赖管理工具可能是poetry，于是乎去查询poetry如何使用的，命令如下：

pip install poetry
poetry install

执行完这个之后发现缺少BasicLibrary，继续安装

pip install BasicLibrary==1.2.6
pip install BasicLibrary.PY==0.6.43
pip install pymediainfo==6.1.0

反正缺少什么安装什么就行了，直到项目不报错能跑起来就算解决了，启动之前记得配置_projectConfig.ini文件，主要配置的是你剪映的草稿位置，然后执行main.py就会生成两个json文件,这样你去剪映草稿打开就可以进行编辑了，效果如下

改造

因为自己是以Python接口的方式提供给Java那边去调用,所以需要引入Flask，直接贴代码，有问题后面再私聊

import os
import shutil
import zipfile
from flask import Flask, jsonify, request, send_file
from JianYingDraft.core.draft import Draft
from projectHelper import ProjectHelper

app = Flask(__name__)


def estimate_reading_time(text, speed_multiplier=1.0, words_per_minute=275):
    num_characters = len(text)
    estimated_time_seconds = (num_characters / words_per_minute) * 60 / speed_multiplier
    return estimated_time_seconds


def zip_folder(folder_path, zip_path):
    """将指定文件夹打包为 ZIP 文件"""
    with zipfile.ZipFile(zip_path, 'w', zipfile.ZIP_DEFLATED) as zipf:
        for root, dirs, files in os.walk(folder_path):
            for file in files:
                file_path = os.path.join(root, file)
                arcname = os.path.relpath(file_path, folder_path)
                zipf.write(file_path, arcname)


@app.route('/basic_using', methods=['POST'])
def basic_using_api():
    try:
        data = request.get_json()
        story_name = data.get('storyName')
        base_path = data.get('basePath')
        real_path = data.get('realPath')
        jian_ying_path = data.get('jianYingPath')
        jian_ying_path = jian_ying_path.replace("\\", "\\\\")

        # 验证参数
        if not base_path.endswith('material'):
            raise ValueError('资源文件夹必须以 "material" 结尾')
        if not os.path.exists(base_path):
            raise FileNotFoundError('指定的资源文件夹不存在')

        # 获取 material 的上级目录（即“重生的复仇”目录）
        parent_path = os.path.dirname(base_path)

        # 确保 material 文件夹不会被修改：复制草稿保存路径到临时路径
        draft_temp_folder = os.path.join(parent_path, "draft_temp")
        os.makedirs(draft_temp_folder, exist_ok=True)

        # 初始化草稿对象
        draft = Draft(jian_ying_path, story_name)

        # 收集图片和文本文件
        image_files = []
        text_files = []
        all_files = sorted(os.listdir(base_path), key=lambda x: int(os.path.splitext(x)[0]))
        for filename in all_files:
            file_path = os.path.join(base_path, filename)
            if os.path.isfile(file_path):
                if filename.endswith(('.jpg', '.png')):
                    image_files.append(file_path)
                elif filename.endswith('.txt'):
                    text_files.append(file_path)

        # 将图片和字幕添加到草稿中
        for img_path, txt_path in zip(image_files, text_files):
            with open(txt_path, 'r', encoding='utf-8') as f:
                text = f.read()
                duration = estimate_reading_time(text) * 1000000
                draft.add_media(img_path, duration=duration)
                draft.add_subtitle(text, duration=duration, color="#550000")

        # 设置草稿保存路径为临时路径
        draft._draft_folder = draft_temp_folder
        draft.save()

        # 将生成的草稿文件移动到“重生的复仇”目录
        for file_name in os.listdir(draft_temp_folder):
            file_path = os.path.join(draft_temp_folder, file_name)
            if os.path.isfile(file_path):
                with open(file_path, 'r', encoding='utf-8') as f:
                    content = f.read()
                    updated_content = content.replace(r"/opt/app/chatgpt/JianYingDraft-master/chaonao", jian_ying_path)
                with open(file_path, 'w', encoding='utf-8') as f:
                    f.write(updated_content)
            shutil.move(os.path.join(draft_temp_folder, file_name), os.path.join(parent_path, file_name))

        # 删除临时目录
        shutil.rmtree(draft_temp_folder)

        # 打包文件夹为 ZIP 文件
        zip_file_path = os.path.join(parent_path, f"{story_name}.zip")
        zip_folder(parent_path, zip_file_path)

        # 返回下载链接
        return send_file(zip_file_path, as_attachment=True)

    except Exception as e:
        return jsonify({"error": str(e)}), 500


if __name__ == '__main__':
    app.run(debug=True, host='0.0.0.0', port=7158)

Java调用方的代码：

@Override
    public void getJianYingDraft(String storyId, String jianYingPath, HttpServletResponse res) {
        try {
            jianYingPath = URLDecoder.decode(jianYingPath, StandardCharsets.UTF_8.name());
        } catch (UnsupportedEncodingException e) {
            throw new RuntimeException(e);
        }
        String BaseURL = "/opt/app/chatgpt/JianYingDraft-master/chaonao";
        ChatStory chatStory = chatStoryService.getById(storyId);
        String storyName = chatStory.getStoryName();

        // 1. 创建故事名称的文件夹，地址为 BaseURL\\storyName\\material
        File storyFolder = new File(BaseURL + File.separator + storyName + File.separator + "material");
        if (!storyFolder.exists()) {
            storyFolder.mkdirs();
        }

        // 2. 下载图片和字幕文件到 material 文件夹
        List<JianYingDraftVO> jianYingDraftVOList = baseMapper.getJianYingDraftList(storyId);
        for (int i = 0; i < jianYingDraftVOList.size(); i++) {
            JianYingDraftVO draftVO = jianYingDraftVOList.get(i);
            if (draftVO != null) {
                try {
                    // 下载图片
                    if (org.springframework.util.StringUtils.hasText(draftVO.getImageUrl())) {
                        URL imageUrl = new URL(draftVO.getImageUrl());
                        String imageFileName = i + "." + getFileExtension(draftVO.getImageUrl());
                        FileUtils.copyURLToFile(imageUrl, new File(storyFolder + File.separator + imageFileName));
                    }
                    // 创建字幕文件
                    if (org.springframework.util.StringUtils.hasText(draftVO.getContent())) {
                        File subtitleFile = new File(storyFolder + File.separator + i + ".txt");
                        FileUtils.writeStringToFile(subtitleFile, draftVO.getContent(), "UTF-8");
                    }
                } catch (IOException e) {
                    log.error("Error downloading material: " + e.getMessage());
                }
            }
        }

        // 3. 调用 Python 接口
        Map<String, Object> requestBody = new HashMap<>();
        requestBody.put("storyName", storyName);
        requestBody.put("basePath", storyFolder.getAbsolutePath());
        requestBody.put("realPath", BaseURL);
        requestBody.put("jianYingPath", jianYingPath);

        OkHttpClient client = OkHttpUtils.okHttpClient;
        okhttp3.RequestBody body = okhttp3.RequestBody
                .create(okhttp3.MediaType.parse("application/json; charset=utf-8"), JSON.toJSONString(requestBody));
        Request request = new Request.Builder()
                .url("http://xxx.xxx.xxx.xxx:7158/basic_using")
                .post(body)
                .addHeader("content-type", "application/json")
                .build();

        try (Response response = client.newCall(request).execute()) {
            if (response.code() == 200 && response.body() != null) {
                // 获取返回的 ZIP 文件流
                InputStream inputStream = response.body().byteStream();

                // 定义下载文件的路径
                String zipFilePath = BaseURL + File.separator + storyName + ".zip";
                File zipFile = new File(zipFilePath);

                // 保存 ZIP 文件到指定路径
                try (FileOutputStream outputStream = new FileOutputStream(zipFile)) {
                    byte[] buffer = new byte[1024];
                    int bytesRead;
                    while ((bytesRead = inputStream.read(buffer)) != -1) {
                        outputStream.write(buffer, 0, bytesRead);
                    }
                }

                // 设置 HTTP 响应，返回 ZIP 文件
                res.setContentType("application/zip");
                res.setHeader("Content-Disposition", "attachment; filename=" + storyName + ".zip");
                try (InputStream fileInputStream = new FileInputStream(zipFile)) {
                    org.apache.commons.io.IOUtils.copy(fileInputStream, res.getOutputStream());
                    res.flushBuffer();
                }
            } else {
                log.error("Error calling basic_using API: " + response.message());
            }
        } catch (IOException e) {
            log.error("Error calling Python API: " + e.getMessage());
        }
    }

这样就完成了草稿导出的功能，浏览器直接会下载一个zip文件，导入到剪映草稿就可以使用了，中间其实还是有很多细节的，也花了不少时间，不过都在代码中体现了，注释也写好了