feat: Initial commit of PDF Tools project

This commit is contained in:
2025-08-25 02:29:48 +08:00
parent af6827cd9e
commit 30180e50a2
48 changed files with 36364 additions and 1 deletions

View File

@@ -0,0 +1,426 @@
const mongoose = require('mongoose');
const conversionTaskSchema = new mongoose.Schema({
taskId: {
type: String,
required: true,
unique: true,
index: true
},
batchId: {
type: String,
index: true // 用于批量转换任务分组
},
userId: {
type: String,
required: false, // 允许匿名用户
index: true
},
fileId: {
type: String,
required: true,
index: true
},
sourceFile: {
name: String,
size: Number,
type: String
},
outputFormat: {
type: String,
required: true,
enum: ['docx', 'html', 'txt', 'png', 'jpg', 'xlsx', 'pptx']
},
options: {
// Word转换选项
preserveLayout: Boolean,
includeImages: Boolean,
imageQuality: {
type: String,
enum: ['low', 'medium', 'high']
},
ocrEnabled: Boolean,
// HTML转换选项
responsive: Boolean,
embedImages: Boolean,
cssFramework: {
type: String,
enum: ['none', 'bootstrap', 'tailwind']
},
includeMetadata: Boolean,
// 图片转换选项
resolution: {
type: Number,
min: 72,
max: 300
},
format: {
type: String,
enum: ['png', 'jpg', 'webp']
},
quality: {
type: Number,
min: 1,
max: 100
},
pageRange: {
type: String,
enum: ['all', 'first', 'custom']
},
customRange: {
start: Number,
end: Number
},
// 文本转换选项
preserveLineBreaks: Boolean,
encoding: {
type: String,
enum: ['utf8', 'gbk', 'ascii']
},
// 批量处理选项
mergeOutput: Boolean,
zipResults: Boolean,
namePattern: String
},
status: {
type: String,
enum: ['pending', 'processing', 'completed', 'failed', 'cancelled'],
default: 'pending',
index: true
},
progress: {
type: Number,
min: 0,
max: 100,
default: 0
},
priority: {
type: Number,
default: 0,
index: true // 用于优先级队列
},
resultFile: {
fileName: String,
filePath: String,
fileSize: Number,
downloadUrl: String
},
processingMetrics: {
startTime: Date,
endTime: Date,
duration: Number, // 毫秒
peakMemoryUsage: Number, // 字节
cpuTime: Number // 毫秒
},
errorInfo: {
code: String,
message: String,
stack: String,
step: String // 失败的转换步骤
},
retryCount: {
type: Number,
default: 0,
max: 3
},
queuePosition: Number,
estimatedCompletionTime: Date,
actualCompletionTime: Date,
qualityScore: {
type: Number,
min: 0,
max: 100
},
userRating: {
score: {
type: Number,
min: 1,
max: 5
},
feedback: String,
ratedAt: Date
}
}, {
timestamps: true
});
// 复合索引
conversionTaskSchema.index({ userId: 1, createdAt: -1 });
conversionTaskSchema.index({ status: 1, priority: -1, createdAt: 1 });
conversionTaskSchema.index({ batchId: 1, status: 1 });
conversionTaskSchema.index({ fileId: 1, outputFormat: 1 });
// 虚拟字段:转换时长
conversionTaskSchema.virtual('durationFormatted').get(function() {
if (!this.processingMetrics.duration) return null;
const seconds = Math.floor(this.processingMetrics.duration / 1000);
const minutes = Math.floor(seconds / 60);
const hours = Math.floor(minutes / 60);
if (hours > 0) {
return `${hours}h ${minutes % 60}m ${seconds % 60}s`;
} else if (minutes > 0) {
return `${minutes}m ${seconds % 60}s`;
} else {
return `${seconds}s`;
}
});
// 虚拟字段:是否可重试
conversionTaskSchema.virtual('canRetry').get(function() {
return this.status === 'failed' && this.retryCount < 3;
});
// 虚拟字段:是否可取消
conversionTaskSchema.virtual('canCancel').get(function() {
return ['pending', 'processing'].includes(this.status);
});
// 实例方法:开始处理
conversionTaskSchema.methods.startProcessing = function() {
this.status = 'processing';
this.progress = 0;
this.processingMetrics.startTime = new Date();
return this.save();
};
// 实例方法:更新进度
conversionTaskSchema.methods.updateProgress = function(progress, step) {
this.progress = Math.min(100, Math.max(0, progress));
if (step) {
this.errorInfo.step = step;
}
return this.save();
};
// 实例方法:标记完成
conversionTaskSchema.methods.markCompleted = function(resultFile) {
this.status = 'completed';
this.progress = 100;
this.resultFile = resultFile;
this.processingMetrics.endTime = new Date();
this.actualCompletionTime = new Date();
if (this.processingMetrics.startTime) {
this.processingMetrics.duration =
this.processingMetrics.endTime - this.processingMetrics.startTime;
}
return this.save();
};
// 实例方法:标记失败
conversionTaskSchema.methods.markFailed = function(error) {
this.status = 'failed';
this.errorInfo = {
code: error.code || 'UNKNOWN_ERROR',
message: error.message,
stack: error.stack,
step: error.step || 'unknown'
};
this.processingMetrics.endTime = new Date();
if (this.processingMetrics.startTime) {
this.processingMetrics.duration =
this.processingMetrics.endTime - this.processingMetrics.startTime;
}
return this.save();
};
// 实例方法:取消任务
conversionTaskSchema.methods.cancel = function() {
if (this.canCancel) {
this.status = 'cancelled';
this.processingMetrics.endTime = new Date();
return this.save();
}
throw new Error('任务无法取消');
};
// 实例方法:重试任务
conversionTaskSchema.methods.retry = function() {
if (this.canRetry) {
this.retryCount += 1;
this.status = 'pending';
this.progress = 0;
this.errorInfo = {};
this.processingMetrics = {};
return this.save();
}
throw new Error('任务无法重试');
};
// 实例方法:设置用户评分
conversionTaskSchema.methods.setUserRating = function(score, feedback) {
this.userRating = {
score,
feedback,
ratedAt: new Date()
};
return this.save();
};
// 静态方法:获取队列中的任务
conversionTaskSchema.statics.getQueuedTasks = function(limit = 10) {
return this.find({ status: 'pending' })
.sort({ priority: -1, createdAt: 1 })
.limit(limit);
};
// 静态方法:获取用户任务历史
conversionTaskSchema.statics.getUserTasks = function(userId, options = {}) {
const {
page = 1,
limit = 10,
status,
outputFormat,
sortBy = 'createdAt',
sortOrder = -1
} = options;
const query = { userId };
if (status) query.status = status;
if (outputFormat) query.outputFormat = outputFormat;
const skip = (page - 1) * limit;
const sort = { [sortBy]: sortOrder };
return this.find(query)
.sort(sort)
.skip(skip)
.limit(limit)
.populate('fileId')
.lean();
};
// 静态方法:获取转换统计
conversionTaskSchema.statics.getConversionStats = async function(timeRange = '7d') {
const cutoffDate = new Date();
switch (timeRange) {
case '1d':
cutoffDate.setDate(cutoffDate.getDate() - 1);
break;
case '7d':
cutoffDate.setDate(cutoffDate.getDate() - 7);
break;
case '30d':
cutoffDate.setDate(cutoffDate.getDate() - 30);
break;
case '90d':
cutoffDate.setDate(cutoffDate.getDate() - 90);
break;
}
const pipeline = [
{ $match: { createdAt: { $gte: cutoffDate } } },
{
$group: {
_id: '$status',
count: { $sum: 1 },
avgDuration: { $avg: '$processingMetrics.duration' },
avgRating: { $avg: '$userRating.score' }
}
}
];
const statusStats = await this.aggregate(pipeline);
const formatStats = await this.aggregate([
{ $match: { createdAt: { $gte: cutoffDate } } },
{
$group: {
_id: '$outputFormat',
count: { $sum: 1 },
successRate: {
$avg: {
$cond: [{ $eq: ['$status', 'completed'] }, 1, 0]
}
}
}
}
]);
return {
byStatus: statusStats,
byFormat: formatStats,
timeRange
};
};
// 静态方法:清理旧任务
conversionTaskSchema.statics.cleanupOldTasks = async function(daysOld = 30) {
const cutoffDate = new Date();
cutoffDate.setDate(cutoffDate.getDate() - daysOld);
const oldTasks = await this.find({
createdAt: { $lt: cutoffDate },
status: { $in: ['completed', 'failed', 'cancelled'] }
});
// 删除关联的结果文件
const fs = require('fs');
let cleanedFiles = 0;
for (const task of oldTasks) {
if (task.resultFile && task.resultFile.filePath) {
try {
if (fs.existsSync(task.resultFile.filePath)) {
fs.unlinkSync(task.resultFile.filePath);
cleanedFiles++;
}
} catch (error) {
console.error(`删除结果文件失败: ${task.resultFile.filePath}`, error);
}
}
}
// 删除数据库记录
const deleteResult = await this.deleteMany({
createdAt: { $lt: cutoffDate },
status: { $in: ['completed', 'failed', 'cancelled'] }
});
return {
deletedTasks: deleteResult.deletedCount,
cleanedFiles
};
};
// 中间件:保存后更新用户统计
conversionTaskSchema.post('save', async function(doc) {
if (doc.userId && doc.isModified('status')) {
try {
const User = require('./User');
const updateData = {};
if (doc.status === 'completed') {
updateData['$inc'] = {
'statistics.totalConversions': 1,
'statistics.successfulConversions': 1
};
} else if (doc.status === 'failed') {
updateData['$inc'] = {
'statistics.totalConversions': 1,
'statistics.failedConversions': 1
};
}
if (Object.keys(updateData).length > 0) {
await User.findOneAndUpdate({ userId: doc.userId }, updateData);
}
} catch (error) {
console.error('更新用户统计失败:', error);
}
}
});
const ConversionTask = mongoose.model('ConversionTask', conversionTaskSchema);
module.exports = ConversionTask;

307
server/models/File.js Normal file
View File

@@ -0,0 +1,307 @@
const mongoose = require('mongoose');
const fileSchema = new mongoose.Schema({
fileId: {
type: String,
required: true,
unique: true,
index: true
},
userId: {
type: String,
required: false, // 允许匿名用户上传
index: true
},
originalName: {
type: String,
required: true,
trim: true
},
fileName: {
type: String,
required: true,
unique: true
},
fileSize: {
type: Number,
required: true,
min: 0
},
mimeType: {
type: String,
required: true,
enum: [
'application/pdf',
'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
'text/html',
'text/plain',
'image/png',
'image/jpeg'
]
},
filePath: {
type: String,
required: true
},
status: {
type: String,
enum: ['uploading', 'ready', 'processing', 'completed', 'failed', 'deleted'],
default: 'ready'
},
metadata: {
pageCount: Number,
width: Number,
height: Number,
hasImages: Boolean,
hasText: Boolean,
isEncrypted: Boolean,
pdfVersion: String,
creationDate: Date,
modificationDate: Date,
author: String,
title: String,
subject: String,
keywords: String
},
checksum: {
type: String,
index: true // 用于去重
},
downloadCount: {
type: Number,
default: 0
},
lastDownloaded: Date,
tags: [String],
isPublic: {
type: Boolean,
default: false
},
expiresAt: {
type: Date,
index: { expireAfterSeconds: 0 } // MongoDB TTL索引自动删除过期文件
}
}, {
timestamps: true
});
// 复合索引
fileSchema.index({ userId: 1, createdAt: -1 });
fileSchema.index({ status: 1, createdAt: -1 });
fileSchema.index({ mimeType: 1, status: 1 });
fileSchema.index({ fileSize: 1 });
// 虚拟字段:格式化文件大小
fileSchema.virtual('formattedFileSize').get(function() {
const sizes = ['Bytes', 'KB', 'MB', 'GB', 'TB'];
if (this.fileSize === 0) return '0 Bytes';
const i = Math.floor(Math.log(this.fileSize) / Math.log(1024));
const size = (this.fileSize / Math.pow(1024, i)).toFixed(1);
return size + ' ' + sizes[i];
});
// 虚拟字段:文件扩展名
fileSchema.virtual('fileExtension').get(function() {
const mimeToExt = {
'application/pdf': 'pdf',
'application/vnd.openxmlformats-officedocument.wordprocessingml.document': 'docx',
'text/html': 'html',
'text/plain': 'txt',
'image/png': 'png',
'image/jpeg': 'jpg'
};
return mimeToExt[this.mimeType] || 'unknown';
});
// 虚拟字段:是否已过期
fileSchema.virtual('isExpired').get(function() {
if (!this.expiresAt) return false;
return new Date() > this.expiresAt;
});
// 实例方法:设置过期时间
fileSchema.methods.setExpiration = function(hours = 24) {
this.expiresAt = new Date(Date.now() + hours * 60 * 60 * 1000);
return this.save();
};
// 实例方法:增加下载次数
fileSchema.methods.incrementDownload = function() {
this.downloadCount += 1;
this.lastDownloaded = new Date();
return this.save();
};
// 实例方法:更新状态
fileSchema.methods.updateStatus = function(status) {
this.status = status;
return this.save();
};
// 静态方法:根据校验和查找重复文件
fileSchema.statics.findByChecksum = function(checksum) {
return this.findOne({ checksum, status: { $ne: 'deleted' } });
};
// 静态方法:获取用户文件
fileSchema.statics.getUserFiles = function(userId, options = {}) {
const {
page = 1,
limit = 10,
status,
mimeType,
sortBy = 'createdAt',
sortOrder = -1
} = options;
const query = { userId };
if (status) query.status = status;
if (mimeType) query.mimeType = mimeType;
const skip = (page - 1) * limit;
const sort = { [sortBy]: sortOrder };
return this.find(query)
.sort(sort)
.skip(skip)
.limit(limit)
.lean();
};
// 静态方法:清理过期文件
fileSchema.statics.cleanupExpiredFiles = async function() {
const expiredFiles = await this.find({
expiresAt: { $lt: new Date() },
status: { $ne: 'deleted' }
});
const fs = require('fs');
const path = require('path');
let cleanedCount = 0;
let freedSpace = 0;
for (const file of expiredFiles) {
try {
// 删除物理文件
if (fs.existsSync(file.filePath)) {
freedSpace += file.fileSize;
fs.unlinkSync(file.filePath);
}
// 更新数据库记录
await this.findByIdAndUpdate(file._id, { status: 'deleted' });
cleanedCount++;
} catch (error) {
console.error(`清理文件失败: ${file.fileName}`, error);
}
}
return { cleanedCount, freedSpace };
};
// 静态方法:获取文件统计
fileSchema.statics.getFileStats = async function() {
const pipeline = [
{
$group: {
_id: '$status',
count: { $sum: 1 },
totalSize: { $sum: '$fileSize' }
}
}
];
const statusStats = await this.aggregate(pipeline);
const typeStats = await this.aggregate([
{
$group: {
_id: '$mimeType',
count: { $sum: 1 },
totalSize: { $sum: '$fileSize' }
}
}
]);
const totalStats = await this.aggregate([
{
$group: {
_id: null,
totalFiles: { $sum: 1 },
totalSize: { $sum: '$fileSize' },
totalDownloads: { $sum: '$downloadCount' }
}
}
]);
return {
byStatus: statusStats,
byType: typeStats,
total: totalStats[0] || { totalFiles: 0, totalSize: 0, totalDownloads: 0 }
};
};
// 静态方法:搜索文件
fileSchema.statics.searchFiles = function(searchTerm, options = {}) {
const {
userId,
page = 1,
limit = 10,
status = { $ne: 'deleted' }
} = options;
const query = {
status,
$or: [
{ originalName: { $regex: searchTerm, $options: 'i' } },
{ tags: { $in: [new RegExp(searchTerm, 'i')] } },
{ 'metadata.title': { $regex: searchTerm, $options: 'i' } },
{ 'metadata.subject': { $regex: searchTerm, $options: 'i' } }
]
};
if (userId) query.userId = userId;
const skip = (page - 1) * limit;
return this.find(query)
.sort({ createdAt: -1 })
.skip(skip)
.limit(limit)
.lean();
};
// 中间件:删除前清理物理文件
fileSchema.pre('remove', async function(next) {
try {
const fs = require('fs');
if (fs.existsSync(this.filePath)) {
fs.unlinkSync(this.filePath);
}
next();
} catch (error) {
console.error('删除物理文件失败:', error);
next(error);
}
});
// 中间件:保存后更新用户统计
fileSchema.post('save', async function(doc) {
if (doc.userId && doc.isNew) {
try {
const User = require('./User');
await User.findOneAndUpdate(
{ userId: doc.userId },
{ $inc: { 'statistics.totalFileSize': doc.fileSize } }
);
} catch (error) {
console.error('更新用户统计失败:', error);
}
}
});
const File = mongoose.model('File', fileSchema);
module.exports = File;

249
server/models/User.js Normal file
View File

@@ -0,0 +1,249 @@
const mongoose = require('mongoose');
const bcrypt = require('bcryptjs');
const userSchema = new mongoose.Schema({
userId: {
type: String,
required: true,
unique: true,
index: true
},
email: {
type: String,
required: true,
unique: true,
lowercase: true,
trim: true,
match: [/^\w+([.-]?\w+)*@\w+([.-]?\w+)*(\.\w{2,3})+$/, '请输入有效的邮箱地址']
},
username: {
type: String,
required: true,
trim: true,
minlength: 2,
maxlength: 50
},
password: {
type: String,
required: true,
minlength: 6
},
settings: {
defaultOutputFormat: {
type: String,
enum: ['docx', 'html', 'txt', 'png', 'jpg'],
default: 'docx'
},
imageQuality: {
type: String,
enum: ['low', 'medium', 'high'],
default: 'medium'
},
autoDownload: {
type: Boolean,
default: true
},
language: {
type: String,
enum: ['zh-CN', 'en-US'],
default: 'zh-CN'
},
autoDelete: {
type: Boolean,
default: true
},
deleteDelay: {
type: Number,
default: 24, // 小时
min: 1,
max: 168 // 7天
},
maxConcurrentTasks: {
type: Number,
default: 3,
min: 1,
max: 10
},
conversionTimeout: {
type: Number,
default: 10, // 分钟
min: 5,
max: 60
}
},
profile: {
avatar: String,
bio: String,
company: String,
website: String
},
statistics: {
totalConversions: {
type: Number,
default: 0
},
successfulConversions: {
type: Number,
default: 0
},
failedConversions: {
type: Number,
default: 0
},
totalFileSize: {
type: Number,
default: 0 // 字节
}
},
lastLoginAt: {
type: Date,
default: null
},
lastActiveAt: {
type: Date,
default: Date.now
},
isActive: {
type: Boolean,
default: true
},
isEmailVerified: {
type: Boolean,
default: false
},
emailVerificationToken: String,
passwordResetToken: String,
passwordResetExpires: Date
}, {
timestamps: true,
toJSON: {
transform: function(doc, ret) {
delete ret.password;
delete ret.emailVerificationToken;
delete ret.passwordResetToken;
delete ret.passwordResetExpires;
delete ret.__v;
return ret;
}
}
});
// 索引
userSchema.index({ email: 1 });
userSchema.index({ username: 1 });
userSchema.index({ createdAt: -1 });
userSchema.index({ lastActiveAt: -1 });
// 密码加密中间件
userSchema.pre('save', async function(next) {
if (!this.isModified('password')) return next();
try {
const salt = await bcrypt.genSalt(12);
this.password = await bcrypt.hash(this.password, salt);
next();
} catch (error) {
next(error);
}
});
// 实例方法:验证密码
userSchema.methods.comparePassword = async function(candidatePassword) {
try {
return await bcrypt.compare(candidatePassword, this.password);
} catch (error) {
throw error;
}
};
// 实例方法:更新最后活动时间
userSchema.methods.updateLastActive = function() {
this.lastActiveAt = new Date();
return this.save();
};
// 实例方法:更新统计信息
userSchema.methods.updateStatistics = function(update) {
Object.assign(this.statistics, update);
return this.save();
};
// 静态方法:根据邮箱查找用户
userSchema.statics.findByEmail = function(email) {
return this.findOne({ email: email.toLowerCase() });
};
// 静态方法:获取活跃用户
userSchema.statics.getActiveUsers = function(days = 30) {
const cutoffDate = new Date();
cutoffDate.setDate(cutoffDate.getDate() - days);
return this.find({
lastActiveAt: { $gte: cutoffDate },
isActive: true
});
};
// 静态方法:获取用户统计
userSchema.statics.getUserStats = async function() {
const pipeline = [
{
$group: {
_id: null,
totalUsers: { $sum: 1 },
activeUsers: {
$sum: {
$cond: [{ $eq: ['$isActive', true] }, 1, 0]
}
},
verifiedUsers: {
$sum: {
$cond: [{ $eq: ['$isEmailVerified', true] }, 1, 0]
}
},
totalConversions: { $sum: '$statistics.totalConversions' },
totalFileSize: { $sum: '$statistics.totalFileSize' }
}
}
];
const result = await this.aggregate(pipeline);
return result[0] || {
totalUsers: 0,
activeUsers: 0,
verifiedUsers: 0,
totalConversions: 0,
totalFileSize: 0
};
};
// 虚拟字段:转换成功率
userSchema.virtual('conversionSuccessRate').get(function() {
if (this.statistics.totalConversions === 0) return 0;
return (this.statistics.successfulConversions / this.statistics.totalConversions * 100).toFixed(1);
});
// 虚拟字段:格式化文件大小
userSchema.virtual('formattedFileSize').get(function() {
const sizes = ['Bytes', 'KB', 'MB', 'GB'];
if (this.statistics.totalFileSize === 0) return '0 Bytes';
const i = Math.floor(Math.log(this.statistics.totalFileSize) / Math.log(1024));
return (this.statistics.totalFileSize / Math.pow(1024, i)).toFixed(1) + ' ' + sizes[i];
});
// 中间件:删除前清理关联数据
userSchema.pre('remove', async function(next) {
try {
// 这里可以添加删除用户相关文件和转换记录的逻辑
// await FileModel.deleteMany({ userId: this.userId });
// await ConversionTaskModel.deleteMany({ userId: this.userId });
next();
} catch (error) {
next(error);
}
});
const User = mongoose.model('User', userSchema);
module.exports = User;