mirror of
https://github.com/CherryHQ/cherry-studio.git
synced 2025-12-31 16:49:07 +08:00
424 lines
15 KiB
TypeScript
424 lines
15 KiB
TypeScript
import { getFileType } from '@main/utils/file'
|
||
import { FileType } from '@types'
|
||
import { app } from 'electron'
|
||
import logger from 'electron-log'
|
||
import * as fs from 'fs'
|
||
import * as path from 'path'
|
||
import { PDFDocument } from 'pdf-lib'
|
||
import { v4 as uuidv4 } from 'uuid'
|
||
|
||
export class PDFService {
|
||
// 使用方法而不是静态属性来获取目录路径
|
||
private static getTempDir(): string {
|
||
return path.join(app.getPath('temp'), 'CherryStudio')
|
||
}
|
||
|
||
private static getStorageDir(): string {
|
||
return path.join(app.getPath('userData'), 'files')
|
||
}
|
||
|
||
// 引入需要的模块
|
||
private static docx: any
|
||
private static pdfParse: any
|
||
|
||
// 懒加载模块
|
||
private static async loadModules() {
|
||
if (!this.docx) {
|
||
try {
|
||
this.docx = require('docx')
|
||
} catch (error) {
|
||
logger.error('[PDFService] Error loading docx module:', error)
|
||
throw new Error('无法加载docx模块，请确保已安装')
|
||
}
|
||
}
|
||
|
||
if (!this.pdfParse) {
|
||
try {
|
||
this.pdfParse = require('pdf-parse')
|
||
} catch (error) {
|
||
logger.error('[PDFService] Error loading pdf-parse module:', error)
|
||
throw new Error('无法加载pdf-parse模块，请确保已安装')
|
||
}
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 获取PDF文件的页数
|
||
* @param _ Electron IPC事件
|
||
* @param filePath PDF文件路径
|
||
* @returns PDF文件的页数
|
||
*/
|
||
static async getPDFPageCount(_: Electron.IpcMainInvokeEvent, filePath: string): Promise<number> {
|
||
try {
|
||
logger.info(`[PDFService] Getting page count for PDF: ${filePath}`)
|
||
const pdfBytes = fs.readFileSync(filePath)
|
||
const pdfDoc = await PDFDocument.load(pdfBytes)
|
||
const pageCount = pdfDoc.getPageCount()
|
||
logger.info(`[PDFService] PDF page count: ${pageCount}`)
|
||
return pageCount
|
||
} catch (error) {
|
||
logger.error('[PDFService] Error getting PDF page count:', error)
|
||
throw error
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 分割PDF文件
|
||
* @param _ Electron IPC事件
|
||
* @param file 原始PDF文件
|
||
* @param pageRange 页码范围，例如：1-5,8,10-15
|
||
* @returns 分割后的PDF文件信息
|
||
*/
|
||
static async splitPDF(_: Electron.IpcMainInvokeEvent, file: FileType, pageRange: string): Promise<FileType> {
|
||
try {
|
||
logger.info(`[PDFService] Splitting PDF: ${file.path}, page range: ${pageRange}`)
|
||
logger.info(`[PDFService] File details:`, JSON.stringify(file))
|
||
|
||
// 确保临时目录存在
|
||
const tempDir = PDFService.getTempDir()
|
||
if (!fs.existsSync(tempDir)) {
|
||
logger.info(`[PDFService] Creating temp directory: ${tempDir}`)
|
||
fs.mkdirSync(tempDir, { recursive: true })
|
||
}
|
||
|
||
// 确保存储目录存在
|
||
const storageDir = PDFService.getStorageDir()
|
||
if (!fs.existsSync(storageDir)) {
|
||
logger.info(`[PDFService] Creating storage directory: ${storageDir}`)
|
||
fs.mkdirSync(storageDir, { recursive: true })
|
||
}
|
||
|
||
// 读取原始PDF文件
|
||
logger.info(`[PDFService] Reading PDF file: ${file.path}`)
|
||
const pdfBytes = fs.readFileSync(file.path)
|
||
logger.info(`[PDFService] PDF file read, size: ${pdfBytes.length} bytes`)
|
||
const pdfDoc = await PDFDocument.load(pdfBytes)
|
||
logger.info(`[PDFService] PDF document loaded, page count: ${pdfDoc.getPageCount()}`)
|
||
|
||
// 创建新的PDF文档
|
||
const newPdfDoc = await PDFDocument.create()
|
||
logger.info(`[PDFService] New PDF document created`)
|
||
|
||
// 解析页码范围
|
||
const pageIndexes = this.parsePageRange(pageRange, pdfDoc.getPageCount())
|
||
logger.info(`[PDFService] Page range parsed, indexes: ${pageIndexes.join(', ')}`)
|
||
|
||
// 复制指定页面到新文档
|
||
const copiedPages = await newPdfDoc.copyPages(pdfDoc, pageIndexes)
|
||
logger.info(`[PDFService] Pages copied, count: ${copiedPages.length}`)
|
||
copiedPages.forEach((page, index) => {
|
||
logger.info(`[PDFService] Adding page ${index + 1} to new document`)
|
||
newPdfDoc.addPage(page)
|
||
})
|
||
|
||
// 保存新文档
|
||
logger.info(`[PDFService] Saving new PDF document`)
|
||
const newPdfBytes = await newPdfDoc.save()
|
||
logger.info(`[PDFService] New PDF document saved, size: ${newPdfBytes.length} bytes`)
|
||
|
||
// 生成新文件ID和路径
|
||
const uuid = uuidv4()
|
||
const ext = '.pdf'
|
||
// 使用之前已经声明的storageDir变量
|
||
const destPath = path.join(storageDir, uuid + ext)
|
||
logger.info(`[PDFService] Destination path: ${destPath}`)
|
||
|
||
// 写入新文件
|
||
logger.info(`[PDFService] Writing new PDF file`)
|
||
fs.writeFileSync(destPath, newPdfBytes)
|
||
logger.info(`[PDFService] New PDF file written`)
|
||
|
||
// 获取文件状态
|
||
const stats = fs.statSync(destPath)
|
||
logger.info(`[PDFService] File stats: size=${stats.size}, created=${stats.birthtime}`)
|
||
|
||
// 创建新文件信息
|
||
const newFile: FileType = {
|
||
id: uuid,
|
||
origin_name: `${path.basename(file.origin_name, '.pdf')}_pages_${pageRange}.pdf`,
|
||
name: uuid + ext,
|
||
path: destPath,
|
||
created_at: stats.birthtime.toISOString(),
|
||
size: stats.size,
|
||
ext: ext,
|
||
type: getFileType(ext),
|
||
count: 1,
|
||
pdf_page_range: pageRange
|
||
}
|
||
|
||
logger.info(`[PDFService] PDF split successful: ${newFile.path}`)
|
||
logger.info(`[PDFService] New file details:`, JSON.stringify(newFile))
|
||
return newFile
|
||
} catch (error) {
|
||
logger.error('[PDFService] Error splitting PDF:', error)
|
||
throw error
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 解析页码范围字符串为页码索引数组
|
||
* @param pageRange 页码范围字符串，例如：1-5,8,10-15
|
||
* @param totalPages PDF文档总页数
|
||
* @returns 页码索引数组（从0开始）
|
||
*/
|
||
private static parsePageRange(pageRange: string, totalPages: number): number[] {
|
||
logger.info(`[PDFService] Parsing page range: ${pageRange}, total pages: ${totalPages}`)
|
||
const pageIndexes: number[] = []
|
||
const parts = pageRange.split(',')
|
||
logger.info(`[PDFService] Page range parts: ${JSON.stringify(parts)}`)
|
||
|
||
try {
|
||
for (const part of parts) {
|
||
const trimmed = part.trim()
|
||
if (!trimmed) {
|
||
logger.info(`[PDFService] Empty part, skipping`)
|
||
continue
|
||
}
|
||
|
||
logger.info(`[PDFService] Processing part: ${trimmed}`)
|
||
|
||
if (trimmed.includes('-')) {
|
||
const [startStr, endStr] = trimmed.split('-')
|
||
const start = parseInt(startStr.trim())
|
||
const end = parseInt(endStr.trim())
|
||
logger.info(`[PDFService] Range part: ${trimmed}, start: ${start}, end: ${end}`)
|
||
|
||
if (isNaN(start) || isNaN(end)) {
|
||
logger.error(`[PDFService] Invalid range part (NaN): ${trimmed}`)
|
||
continue
|
||
}
|
||
|
||
if (start < 1 || end > totalPages || start > end) {
|
||
logger.warn(`[PDFService] Invalid range: ${start}-${end}, totalPages: ${totalPages}`)
|
||
continue
|
||
}
|
||
|
||
for (let i = start; i <= end; i++) {
|
||
pageIndexes.push(i - 1) // PDF页码从0开始，但用户输入从1开始
|
||
logger.info(`[PDFService] Added page index: ${i - 1} (page ${i})`)
|
||
}
|
||
} else {
|
||
const page = parseInt(trimmed)
|
||
logger.info(`[PDFService] Single page: ${page}`)
|
||
|
||
if (isNaN(page)) {
|
||
logger.error(`[PDFService] Invalid page number (NaN): ${trimmed}`)
|
||
continue
|
||
}
|
||
|
||
if (page < 1 || page > totalPages) {
|
||
logger.warn(`[PDFService] Page ${page} out of range, totalPages: ${totalPages}`)
|
||
continue
|
||
}
|
||
|
||
pageIndexes.push(page - 1) // PDF页码从0开始，但用户输入从1开始
|
||
logger.info(`[PDFService] Added page index: ${page - 1} (page ${page})`)
|
||
}
|
||
}
|
||
|
||
// 去重并排序
|
||
const result = [...new Set(pageIndexes)].sort((a, b) => a - b)
|
||
logger.info(`[PDFService] Final page indexes: ${result.join(', ')}`)
|
||
return result
|
||
} catch (error) {
|
||
logger.error(`[PDFService] Error parsing page range: ${error}`)
|
||
// 如果解析出错，返回空数组
|
||
return []
|
||
}
|
||
}
|
||
|
||
/**
|
||
* 将PDF转换为Word文档
|
||
* @param _ Electron IPC事件
|
||
* @param params 包含PDF缓冲区和输出路径的参数
|
||
* @returns 转换结果
|
||
*/
|
||
static async toWord(_: Electron.IpcMainInvokeEvent, params: { pdfBuffer: ArrayBuffer, outputPath?: string }): Promise<{ success: boolean, path?: string, error?: string }> {
|
||
try {
|
||
logger.info('[PDFService] Starting PDF to Word conversion')
|
||
|
||
// 加载必要的模块
|
||
await this.loadModules()
|
||
|
||
const { pdfBuffer, outputPath } = params
|
||
|
||
// 将ArrayBuffer转换为Buffer
|
||
const buffer = Buffer.from(pdfBuffer)
|
||
|
||
// 解析PDF文本
|
||
logger.info('[PDFService] Parsing PDF text')
|
||
|
||
// 使用更高级的选项来提取文本
|
||
const pdfData = await this.pdfParse(buffer, {
|
||
// 使用自定义的渲染器来提取文本
|
||
pagerender: function(pageData) {
|
||
// 检查页面数据是否有效
|
||
if (!pageData || !pageData.getTextContent) {
|
||
logger.warn('[PDFService] Invalid page data or getTextContent method not available');
|
||
return Promise.resolve('');
|
||
}
|
||
|
||
// 记录页面信息
|
||
logger.info(`[PDFService] Processing page ${pageData.pageIndex + 1}`);
|
||
|
||
return pageData.getTextContent({
|
||
// 启用更多选项以获取更好的文本提取结果
|
||
normalizeWhitespace: true,
|
||
disableCombineTextItems: false
|
||
}).then(function(textContent: { items: Array<{ str?: string, transform?: number[], width?: number }> }) {
|
||
// 记录文本内容项目数
|
||
logger.info(`[PDFService] Page ${pageData.pageIndex + 1} has ${textContent.items.length} text items`);
|
||
|
||
if (textContent.items.length === 0) {
|
||
logger.warn(`[PDFService] No text items found on page ${pageData.pageIndex + 1}`);
|
||
return '';
|
||
}
|
||
|
||
let lastY: number | null = null;
|
||
let lastX: number | null = null;
|
||
let text = '';
|
||
|
||
// 处理文本内容
|
||
for (const item of textContent.items) {
|
||
// 确保 item 有 str 属性且不为空
|
||
if ('str' in item && item.str && item.str.trim().length > 0 && item.transform && item.transform.length >= 6) {
|
||
const x = item.transform[4];
|
||
const y = item.transform[5];
|
||
|
||
// 根据Y坐标判断是否是新行
|
||
if (lastY !== null && Math.abs(lastY - y) > 5) {
|
||
text += '\n';
|
||
lastX = null; // 重置X坐标
|
||
}
|
||
// 根据X坐标判断是否需要添加空格
|
||
else if (lastX !== null && text.length > 0 &&
|
||
text[text.length - 1] !== ' ' &&
|
||
item.str[0] !== ' ' &&
|
||
x - lastX > 10) { // 如果X坐标差距较大，添加空格
|
||
text += ' ';
|
||
}
|
||
|
||
text += item.str;
|
||
lastY = y;
|
||
lastX = x + (item.width || 0);
|
||
}
|
||
}
|
||
|
||
// 记录提取的文本长度
|
||
logger.info(`[PDFService] Extracted ${text.length} characters from page ${pageData.pageIndex + 1}`);
|
||
|
||
return text;
|
||
}).catch(function(error: Error) {
|
||
logger.error(`[PDFService] Error extracting text from page ${pageData.pageIndex + 1}:`, error);
|
||
return '';
|
||
});
|
||
}
|
||
});
|
||
|
||
// 获取提取的文本
|
||
const pdfText = pdfData.text;
|
||
|
||
// 记录提取的文本长度
|
||
logger.info(`[PDFService] Extracted text length: ${pdfText.length}`);
|
||
|
||
// 如果文本太短，可能是提取失败
|
||
if (pdfText.length < 10) {
|
||
logger.warn(`[PDFService] Extracted text is too short: "${pdfText}"`);
|
||
} else {
|
||
// 记录提取的文本的前100个字符，用于调试
|
||
logger.info(`[PDFService] First 100 characters of extracted text: "${pdfText.substring(0, 100).replace(/\n/g, '\\n')}..."`);
|
||
}
|
||
|
||
// 记录PDF的元数据
|
||
if (pdfData.info) {
|
||
logger.info(`[PDFService] PDF metadata:`, JSON.stringify(pdfData.info));
|
||
}
|
||
|
||
// 记录PDF的页数
|
||
logger.info(`[PDFService] PDF page count: ${pdfData.numpages}`);
|
||
|
||
// 记录PDF的版本
|
||
if (pdfData.pdfInfo && pdfData.pdfInfo.version) {
|
||
logger.info(`[PDFService] PDF version: ${pdfData.pdfInfo.version}`);
|
||
}
|
||
|
||
// 创建Word文档
|
||
logger.info('[PDFService] Creating Word document')
|
||
const { Document, Packer, Paragraph, TextRun } = this.docx
|
||
|
||
// 将PDF文本分割成段落，使用更智能的分割方法
|
||
const paragraphs = pdfText
|
||
.split(/\n{2,}/) // 使用两个或更多换行符分割段落
|
||
.map((p: string) => p.replace(/\n/g, ' ').trim()) // 将段落内的换行符替换为空格
|
||
.filter((p: string) => p.length > 0); // 过滤掉空段落
|
||
|
||
// 创建Word文档对象
|
||
const doc = new Document({
|
||
sections: [
|
||
{
|
||
properties: {},
|
||
children: paragraphs.map((text: string) =>
|
||
new Paragraph({
|
||
children: [
|
||
new TextRun({
|
||
text: text.trim(),
|
||
size: 24 // 12pt
|
||
})
|
||
]
|
||
})
|
||
)
|
||
}
|
||
]
|
||
})
|
||
|
||
// 生成Word文档
|
||
logger.info('[PDFService] Generating Word document buffer')
|
||
const docBuffer = await Packer.toBuffer(doc)
|
||
|
||
// 确定输出路径
|
||
let finalOutputPath = outputPath
|
||
|
||
// 如果没有提供输出路径，则使用对话框让用户选择保存位置
|
||
if (!finalOutputPath) {
|
||
const { dialog } = require('electron')
|
||
const result = await dialog.showSaveDialog({
|
||
title: '保存Word文档',
|
||
defaultPath: 'converted-document.docx',
|
||
filters: [{ name: 'Word文档', extensions: ['docx'] }]
|
||
})
|
||
|
||
if (result.canceled) {
|
||
logger.info('[PDFService] User canceled save dialog')
|
||
return { success: false, error: '用户取消了保存' }
|
||
}
|
||
|
||
finalOutputPath = result.filePath
|
||
}
|
||
|
||
// 保存Word文档
|
||
if (!finalOutputPath) {
|
||
logger.error('[PDFService] Output path is undefined')
|
||
return {
|
||
success: false,
|
||
error: '输出路径未定义'
|
||
}
|
||
}
|
||
|
||
logger.info(`[PDFService] Saving Word document to: ${finalOutputPath}`)
|
||
fs.writeFileSync(finalOutputPath, docBuffer)
|
||
|
||
logger.info('[PDFService] PDF to Word conversion completed successfully')
|
||
return {
|
||
success: true,
|
||
path: finalOutputPath
|
||
}
|
||
} catch (error) {
|
||
logger.error('[PDFService] Error converting PDF to Word:', error)
|
||
return {
|
||
success: false,
|
||
error: error instanceof Error ? error.message : String(error)
|
||
}
|
||
}
|
||
}
|
||
}
|