From f1cfdb29f8343601190aaa9363d15cf32324bd34 Mon Sep 17 00:00:00 2001
From: kangfenmao <kangfenmao@qq.com>
Date: Sat, 12 Oct 2024 13:18:53 +0800
Subject: [PATCH] feat: add document files support

---
 package.json                                  |   1 +
 src/main/constant.ts                          |  94 ++++++
 src/main/services/FileManager.ts              |   9 +-
 src/main/utils/file.ts                        |  97 +-----
 src/renderer/src/config/constant.ts           |   1 +
 .../pages/home/Inputbar/AttachmentButton.tsx  |   4 +-
 .../src/pages/home/Inputbar/Inputbar.tsx      |   4 +-
 .../src/providers/AnthropicProvider.ts        |   2 +-
 src/renderer/src/providers/GeminiProvider.ts  |   2 +-
 src/renderer/src/providers/OpenAIProvider.ts  |   4 +-
 yarn.lock                                     | 279 +++++++++++++++++-
 11 files changed, 383 insertions(+), 114 deletions(-)
 create mode 100644 src/main/constant.ts

diff --git a/package.json b/package.json
index 5d9a8c4e..5177324d 100644
--- a/package.json
+++ b/package.json
@@ -39,6 +39,7 @@
     "electron-window-state": "^5.0.3",
     "fs-extra": "^11.2.0",
     "html2canvas": "^1.4.1",
+    "officeparser": "^4.1.1",
     "unzipper": "^0.12.3",
     "webdav": "4.11.4"
   },
diff --git a/src/main/constant.ts b/src/main/constant.ts
new file mode 100644
index 00000000..f9d52d74
--- /dev/null
+++ b/src/main/constant.ts
@@ -0,0 +1,94 @@
+export const imageExts = ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.webp']
+export const videoExts = ['.mp4', '.avi', '.mov', '.wmv', '.flv', '.mkv']
+export const audioExts = ['.mp3', '.wav', '.ogg', '.flac', '.aac']
+export const documentExts = ['.pdf', '.docx', '.pptx', '.xlsx', '.odt', '.odp', '.ods']
+export const textExts = [
+  '.txt', // 普通文本文件
+  '.md', // Markdown 文件
+  '.mdx', // Markdown 文件
+  '.html', // HTML 文件
+  '.htm', // HTML 文件的另一种扩展名
+  '.xml', // XML 文件
+  '.json', // JSON 文件
+  '.yaml', // YAML 文件
+  '.yml', // YAML 文件的另一种扩展名
+  '.csv', // 逗号分隔值文件
+  '.tsv', // 制表符分隔值文件
+  '.ini', // 配置文件
+  '.log', // 日志文件
+  '.rtf', // 富文本格式文件
+  '.tex', // LaTeX 文件
+  '.srt', // 字幕文件
+  '.xhtml', // XHTML 文件
+  '.nfo', // 信息文件（主要用于场景发布）
+  '.conf', // 配置文件
+  '.config', // 配置文件
+  '.env', // 环境变量文件
+  '.properties', // 配置属性文件
+  '.latex', // LaTeX 文档文件
+  '.rst', // reStructuredText 文件
+  '.php', // PHP 脚本文件，包含嵌入的 HTML
+  '.js', // JavaScript 文件（部分是文本，部分可能包含代码）
+  '.ts', // TypeScript 文件
+  '.jsp', // JavaServer Pages 文件
+  '.aspx', // ASP.NET 文件
+  '.bat', // Windows 批处理文件
+  '.sh', // Unix/Linux Shell 脚本文件
+  '.py', // Python 脚本文件
+  '.rb', // Ruby 脚本文件
+  '.pl', // Perl 脚本文件
+  '.sql', // SQL 脚本文件
+  '.css', // Cascading Style Sheets 文件
+  '.less', // Less CSS 预处理器文件
+  '.scss', // Sass CSS 预处理器文件
+  '.sass', // Sass 文件
+  '.styl', // Stylus CSS 预处理器文件
+  '.coffee', // CoffeeScript 文件
+  '.ino', // Arduino 代码文件
+  '.ino', // Arduino 代码文件
+  '.asm', // Assembly 语言文件
+  '.go', // Go 语言文件
+  '.scala', // Scala 语言文件
+  '.swift', // Swift 语言文件
+  '.kt', // Kotlin 语言文件
+  '.rs', // Rust 语言文件
+  '.lua', // Lua 语言文件
+  '.groovy', // Groovy 语言文件
+  '.dart', // Dart 语言文件
+  '.hs', // Haskell 语言文件
+  '.clj', // Clojure 语言文件
+  '.cljs', // ClojureScript 语言文件
+  '.elm', // Elm 语言文件
+  '.erl', // Erlang 语言文件
+  '.ex', // Elixir 语言文件
+  '.exs', // Elixir 脚本文件
+  '.pug', // Pug (formerly Jade) 模板文件
+  '.haml', // Haml 模板文件
+  '.slim', // Slim 模板文件
+  '.tpl', // 模板文件（通用）
+  '.ejs', // Embedded JavaScript 模板文件
+  '.hbs', // Handlebars 模板文件
+  '.mustache', // Mustache 模板文件
+  '.jade', // Jade 模板文件 (已重命名为 Pug)
+  '.twig', // Twig 模板文件
+  '.blade', // Blade 模板文件 (Laravel)
+  '.vue', // Vue.js 单文件组件
+  '.jsx', // React JSX 文件
+  '.tsx', // React TSX 文件
+  '.graphql', // GraphQL 查询语言文件
+  '.gql', // GraphQL 查询语言文件
+  '.proto', // Protocol Buffers 文件
+  '.thrift', // Thrift 文件
+  '.toml', // TOML 配置文件
+  '.edn', // Clojure 数据表示文件
+  '.cake', // CakePHP 配置文件
+  '.ctp', // CakePHP 视图文件
+  '.cfm', // ColdFusion 标记语言文件
+  '.cfc', // ColdFusion 组件文件
+  '.m', // Objective-C 源文件
+  '.mm', // Objective-C++ 源文件
+  '.gradle', // Gradle 构建文件
+  '.groovy', // Gradle 构建文件
+  '.gradle', // Gradle 构建文件
+  '.kts' // Kotlin Script 文件
+]
diff --git a/src/main/services/FileManager.ts b/src/main/services/FileManager.ts
index eeca5d29..3ebc477b 100644
--- a/src/main/services/FileManager.ts
+++ b/src/main/services/FileManager.ts
@@ -1,3 +1,4 @@
+import { documentExts } from '@main/constant'
 import { getFileType } from '@main/utils/file'
 import { FileType } from '@types'
 import * as crypto from 'crypto'
@@ -13,6 +14,7 @@ import logger from 'electron-log'
 import * as fs from 'fs'
 import { writeFileSync } from 'fs'
 import { readFile } from 'fs/promises'
+import officeParser from 'officeparser'
 import * as path from 'path'
 import { v4 as uuidv4 } from 'uuid'
 
@@ -172,7 +174,12 @@ class FileManager {
   }
 
   public readFile = async (_: Electron.IpcMainInvokeEvent, id: string): Promise<string> => {
-    const filePath = path.join(this.storageDir, id)
+    const filePath = id.includes('/') ? id : path.join(this.storageDir, id)
+
+    if (documentExts.includes(path.extname(filePath))) {
+      return await officeParser.parseOfficeAsync(filePath)
+    }
+
     return fs.readFileSync(filePath, 'utf8')
   }
 
diff --git a/src/main/utils/file.ts b/src/main/utils/file.ts
index f2b1d626..fa8c0ef9 100644
--- a/src/main/utils/file.ts
+++ b/src/main/utils/file.ts
@@ -1,101 +1,8 @@
+import { audioExts, documentExts, imageExts, textExts, videoExts } from '@main/constant'
+
 import { FileTypes } from '../../renderer/src/types'
 
 export function getFileType(ext: string): FileTypes {
-  const imageExts = ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.webp']
-  const videoExts = ['.mp4', '.avi', '.mov', '.wmv', '.flv', '.mkv']
-  const audioExts = ['.mp3', '.wav', '.ogg', '.flac', '.aac']
-  const documentExts = ['.pdf', '.doc', '.docx', '.xls', '.xlsx', '.ppt', '.pptx']
-  const textExts = [
-    '.txt', // 普通文本文件
-    '.md', // Markdown 文件
-    '.mdx', // Markdown 文件
-    '.html', // HTML 文件
-    '.htm', // HTML 文件的另一种扩展名
-    '.xml', // XML 文件
-    '.json', // JSON 文件
-    '.yaml', // YAML 文件
-    '.yml', // YAML 文件的另一种扩展名
-    '.csv', // 逗号分隔值文件
-    '.tsv', // 制表符分隔值文件
-    '.ini', // 配置文件
-    '.log', // 日志文件
-    '.rtf', // 富文本格式文件
-    '.tex', // LaTeX 文件
-    '.srt', // 字幕文件
-    '.xhtml', // XHTML 文件
-    '.nfo', // 信息文件（主要用于场景发布）
-    '.conf', // 配置文件
-    '.config', // 配置文件
-    '.env', // 环境变量文件
-    '.properties', // 配置属性文件
-    '.latex', // LaTeX 文档文件
-    '.rst', // reStructuredText 文件
-    '.php', // PHP 脚本文件，包含嵌入的 HTML
-    '.js', // JavaScript 文件（部分是文本，部分可能包含代码）
-    '.ts', // TypeScript 文件
-    '.jsp', // JavaServer Pages 文件
-    '.aspx', // ASP.NET 文件
-    '.bat', // Windows 批处理文件
-    '.sh', // Unix/Linux Shell 脚本文件
-    '.py', // Python 脚本文件
-    '.rb', // Ruby 脚本文件
-    '.pl', // Perl 脚本文件
-    '.sql', // SQL 脚本文件
-    '.css', // Cascading Style Sheets 文件
-    '.less', // Less CSS 预处理器文件
-    '.scss', // Sass CSS 预处理器文件
-    '.sass', // Sass 文件
-    '.styl', // Stylus CSS 预处理器文件
-    '.coffee', // CoffeeScript 文件
-    '.ino', // Arduino 代码文件
-    '.ino', // Arduino 代码文件
-    '.asm', // Assembly 语言文件
-    '.go', // Go 语言文件
-    '.scala', // Scala 语言文件
-    '.swift', // Swift 语言文件
-    '.kt', // Kotlin 语言文件
-    '.rs', // Rust 语言文件
-    '.lua', // Lua 语言文件
-    '.groovy', // Groovy 语言文件
-    '.dart', // Dart 语言文件
-    '.hs', // Haskell 语言文件
-    '.clj', // Clojure 语言文件
-    '.cljs', // ClojureScript 语言文件
-    '.elm', // Elm 语言文件
-    '.erl', // Erlang 语言文件
-    '.ex', // Elixir 语言文件
-    '.exs', // Elixir 脚本文件
-    '.pug', // Pug (formerly Jade) 模板文件
-    '.haml', // Haml 模板文件
-    '.slim', // Slim 模板文件
-    '.tpl', // 模板文件（通用）
-    '.ejs', // Embedded JavaScript 模板文件
-    '.hbs', // Handlebars 模板文件
-    '.mustache', // Mustache 模板文件
-    '.jade', // Jade 模板文件 (已重命名为 Pug)
-    '.twig', // Twig 模板文件
-    '.blade', // Blade 模板文件 (Laravel)
-    '.vue', // Vue.js 单文件组件
-    '.jsx', // React JSX 文件
-    '.tsx', // React TSX 文件
-    '.graphql', // GraphQL 查询语言文件
-    '.gql', // GraphQL 查询语言文件
-    '.proto', // Protocol Buffers 文件
-    '.thrift', // Thrift 文件
-    '.toml', // TOML 配置文件
-    '.edn', // Clojure 数据表示文件
-    '.cake', // CakePHP 配置文件
-    '.ctp', // CakePHP 视图文件
-    '.cfm', // ColdFusion 标记语言文件
-    '.cfc', // ColdFusion 组件文件
-    '.m', // Objective-C 源文件
-    '.mm', // Objective-C++ 源文件
-    '.gradle', // Gradle 构建文件
-    '.groovy', // Gradle 构建文件
-    '.gradle', // Gradle 构建文件
-    '.kts' // Kotlin Script 文件
-  ]
-
   ext = ext.toLowerCase()
   if (imageExts.includes(ext)) return FileTypes.IMAGE
   if (videoExts.includes(ext)) return FileTypes.VIDEO
diff --git a/src/renderer/src/config/constant.ts b/src/renderer/src/config/constant.ts
index 8174f5c5..607e70ce 100644
--- a/src/renderer/src/config/constant.ts
+++ b/src/renderer/src/config/constant.ts
@@ -9,6 +9,7 @@ export const isWindows = platform === 'win32' || platform === 'win64'
 export const isLinux = platform === 'linux'
 
 export const imageExts = ['.jpg', '.png', '.jpeg']
+export const documentExts = ['.pdf', '.docx', '.pptx', '.xlsx', '.odt', '.odp', '.ods']
 export const textExts = [
   '.txt', // 普通文本文件
   '.md', // Markdown 文件
diff --git a/src/renderer/src/pages/home/Inputbar/AttachmentButton.tsx b/src/renderer/src/pages/home/Inputbar/AttachmentButton.tsx
index 34e7df5d..3986bd89 100644
--- a/src/renderer/src/pages/home/Inputbar/AttachmentButton.tsx
+++ b/src/renderer/src/pages/home/Inputbar/AttachmentButton.tsx
@@ -1,5 +1,5 @@
 import { PaperClipOutlined } from '@ant-design/icons'
-import { imageExts, textExts } from '@renderer/config/constant'
+import { documentExts, imageExts, textExts } from '@renderer/config/constant'
 import { isVisionModel } from '@renderer/config/models'
 import { FileType, Model } from '@renderer/types'
 import { Tooltip } from 'antd'
@@ -15,7 +15,7 @@ interface Props {
 
 const AttachmentButton: FC<Props> = ({ model, files, setFiles, ToolbarButton }) => {
   const { t } = useTranslation()
-  const extensions = isVisionModel(model) ? [...imageExts, ...textExts] : [...textExts]
+  const extensions = isVisionModel(model) ? [...imageExts, ...textExts, ...documentExts] : [...textExts]
 
   const onSelectFile = async () => {
     if (files.length > 0) {
diff --git a/src/renderer/src/pages/home/Inputbar/Inputbar.tsx b/src/renderer/src/pages/home/Inputbar/Inputbar.tsx
index 3e86975d..e15d3024 100644
--- a/src/renderer/src/pages/home/Inputbar/Inputbar.tsx
+++ b/src/renderer/src/pages/home/Inputbar/Inputbar.tsx
@@ -7,7 +7,7 @@ import {
   PauseCircleOutlined,
   QuestionCircleOutlined
 } from '@ant-design/icons'
-import { imageExts, textExts } from '@renderer/config/constant'
+import { documentExts, imageExts, textExts } from '@renderer/config/constant'
 import { isVisionModel } from '@renderer/config/models'
 import db from '@renderer/databases'
 import { useAssistant } from '@renderer/hooks/useAssistant'
@@ -60,7 +60,7 @@ const Inputbar: FC<Props> = ({ assistant, setActiveTopic }) => {
   const dispatch = useAppDispatch()
 
   const isVision = useMemo(() => isVisionModel(model), [model])
-  const supportExts = useMemo(() => [...textExts, ...(isVision ? imageExts : [])], [isVision])
+  const supportExts = useMemo(() => [...textExts, ...documentExts, ...(isVision ? imageExts : [])], [isVision])
   const inputTokenCount = useMemo(() => estimateTextTokens(text), [text])
 
   _text = text
diff --git a/src/renderer/src/providers/AnthropicProvider.ts b/src/renderer/src/providers/AnthropicProvider.ts
index 8b48a686..390582e7 100644
--- a/src/renderer/src/providers/AnthropicProvider.ts
+++ b/src/renderer/src/providers/AnthropicProvider.ts
@@ -37,7 +37,7 @@ export default class AnthropicProvider extends BaseProvider {
           }
         })
       }
-      if (file.type === FileTypes.TEXT) {
+      if ([FileTypes.TEXT, FileTypes.DOCUMENT].includes(file.type)) {
         const fileContent = await (await window.api.file.read(file.id + file.ext)).trim()
         parts.push({
           type: 'text',
diff --git a/src/renderer/src/providers/GeminiProvider.ts b/src/renderer/src/providers/GeminiProvider.ts
index a98f67da..600f239d 100644
--- a/src/renderer/src/providers/GeminiProvider.ts
+++ b/src/renderer/src/providers/GeminiProvider.ts
@@ -40,7 +40,7 @@ export default class GeminiProvider extends BaseProvider {
           }
         } as InlineDataPart)
       }
-      if (file.type === FileTypes.TEXT) {
+      if ([FileTypes.TEXT, FileTypes.DOCUMENT].includes(file.type)) {
         const fileContent = await (await window.api.file.read(file.id + file.ext)).trim()
         parts.push({
           text: file.origin_name + '\n' + fileContent
diff --git a/src/renderer/src/providers/OpenAIProvider.ts b/src/renderer/src/providers/OpenAIProvider.ts
index 0f4ba21a..a007f087 100644
--- a/src/renderer/src/providers/OpenAIProvider.ts
+++ b/src/renderer/src/providers/OpenAIProvider.ts
@@ -64,7 +64,7 @@ export default class OpenAIProvider extends BaseProvider {
 
     if (this.isNotSupportFiles) {
       if (message.files) {
-        const textFiles = message.files.filter((file) => file.type === FileTypes.TEXT)
+        const textFiles = message.files.filter((file) => [FileTypes.TEXT, FileTypes.DOCUMENT].includes(file.type))
 
         if (textFiles.length > 0) {
           let text = ''
@@ -104,7 +104,7 @@ export default class OpenAIProvider extends BaseProvider {
           image_url: { url: image.data }
         })
       }
-      if (file.type === FileTypes.TEXT) {
+      if ([FileTypes.TEXT, FileTypes.DOCUMENT].includes(file.type)) {
         const fileContent = await (await window.api.file.read(file.id + file.ext)).trim()
         parts.push({
           type: 'text',
diff --git a/yarn.lock b/yarn.lock
index fbbec603..f7499179 100644
--- a/yarn.lock
+++ b/yarn.lock
@@ -2227,7 +2227,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"@xmldom/xmldom@npm:^0.8.8":
+"@xmldom/xmldom@npm:^0.8.10, @xmldom/xmldom@npm:^0.8.8":
   version: 0.8.10
   resolution: "@xmldom/xmldom@npm:0.8.10"
   checksum: 10c0/c7647c442502720182b0d65b17d45d2d95317c1c8c497626fe524bda79b4fb768a9aa4fae2da919f308e7abcff7d67c058b102a9d641097e9a57f0b80187851f
@@ -2287,6 +2287,7 @@ __metadata:
     localforage: "npm:^1.10.0"
     lodash: "npm:^4.17.21"
     mime: "npm:^4.0.4"
+    officeparser: "npm:^4.1.1"
     openai: "npm:^4.52.1"
     prettier: "npm:^3.2.4"
     react: "npm:^18.2.0"
@@ -2941,6 +2942,16 @@ __metadata:
   languageName: node
   linkType: hard
 
+"bl@npm:^1.0.0":
+  version: 1.2.3
+  resolution: "bl@npm:1.2.3"
+  dependencies:
+    readable-stream: "npm:^2.3.5"
+    safe-buffer: "npm:^5.1.1"
+  checksum: 10c0/ee6478864d3b1295614f269f3fbabeb2362a2f2fc7f8dc2f6c1f944a278d84e0572ecefd6d0b0736d7418763f98dc3b2738253191ea9e98e4b08de211cfac0a6
+  languageName: node
+  linkType: hard
+
 "bluebird-lst@npm:^1.0.9":
   version: 1.0.9
   resolution: "bluebird-lst@npm:1.0.9"
@@ -3022,6 +3033,23 @@ __metadata:
   languageName: node
   linkType: hard
 
+"buffer-alloc-unsafe@npm:^1.1.0":
+  version: 1.1.0
+  resolution: "buffer-alloc-unsafe@npm:1.1.0"
+  checksum: 10c0/06b9298c9369621a830227c3797ceb3ff5535e323946d7b39a7398fed8b3243798259b3c85e287608c5aad35ccc551cec1a0a5190cc8f39652e8eee25697fc9c
+  languageName: node
+  linkType: hard
+
+"buffer-alloc@npm:^1.2.0":
+  version: 1.2.0
+  resolution: "buffer-alloc@npm:1.2.0"
+  dependencies:
+    buffer-alloc-unsafe: "npm:^1.1.0"
+    buffer-fill: "npm:^1.0.0"
+  checksum: 10c0/09d87dd53996342ccfbeb2871257d8cdb25ce9ee2259adc95c6490200cd6e528c5fbae8f30bcc323fe8d8efb0fe541e4ac3bbe9ee3f81c6b7c4b27434cc02ab4
+  languageName: node
+  linkType: hard
+
 "buffer-crc32@npm:^1.0.0":
   version: 1.0.0
   resolution: "buffer-crc32@npm:1.0.0"
@@ -3050,6 +3078,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"buffer-fill@npm:^1.0.0":
+  version: 1.0.0
+  resolution: "buffer-fill@npm:1.0.0"
+  checksum: 10c0/55b5654fbbf2d7ceb4991bb537f5e5b5b5b9debca583fee416a74fcec47c16d9e7a90c15acd27577da7bd750b7fa6396e77e7c221e7af138b6d26242381c6e4d
+  languageName: node
+  linkType: hard
+
 "buffer-from@npm:^1.0.0":
   version: 1.1.2
   resolution: "buffer-from@npm:1.1.2"
@@ -3057,7 +3092,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"buffer@npm:^5.1.0, buffer@npm:^5.2.0":
+"buffer@npm:^5.1.0, buffer@npm:^5.2.0, buffer@npm:^5.2.1":
   version: 5.7.1
   resolution: "buffer@npm:5.7.1"
   dependencies:
@@ -3489,6 +3524,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"commander@npm:^2.8.1":
+  version: 2.20.3
+  resolution: "commander@npm:2.20.3"
+  checksum: 10c0/74c781a5248c2402a0a3e966a0a2bba3c054aad144f5c023364be83265e796b20565aa9feff624132ff629aa64e16999fa40a743c10c12f7c61e96a794b99288
+  languageName: node
+  linkType: hard
+
 "commander@npm:^5.0.0":
   version: 5.1.0
   resolution: "commander@npm:5.1.0"
@@ -3843,6 +3885,69 @@ __metadata:
   languageName: node
   linkType: hard
 
+"decompress-tar@npm:^4.0.0, decompress-tar@npm:^4.1.0, decompress-tar@npm:^4.1.1":
+  version: 4.1.1
+  resolution: "decompress-tar@npm:4.1.1"
+  dependencies:
+    file-type: "npm:^5.2.0"
+    is-stream: "npm:^1.1.0"
+    tar-stream: "npm:^1.5.2"
+  checksum: 10c0/92d86c5dfe2a89f9b5db584668f8ed2a3107339083872c7f78b5f7d55222d954545e018c10346a50991542ad6d1406128bf1c97a24f023810993a1dcfb3c3f21
+  languageName: node
+  linkType: hard
+
+"decompress-tarbz2@npm:^4.0.0":
+  version: 4.1.1
+  resolution: "decompress-tarbz2@npm:4.1.1"
+  dependencies:
+    decompress-tar: "npm:^4.1.0"
+    file-type: "npm:^6.1.0"
+    is-stream: "npm:^1.1.0"
+    seek-bzip: "npm:^1.0.5"
+    unbzip2-stream: "npm:^1.0.9"
+  checksum: 10c0/d5ab2c2435a53f45da8348ffdb5ae0a3ff8fec55948b7890a1c55413de4d1e539a22978e7dcd8bd3561985878c9778253fe146cbdea429f04fa4529abb57c54e
+  languageName: node
+  linkType: hard
+
+"decompress-targz@npm:^4.0.0":
+  version: 4.1.1
+  resolution: "decompress-targz@npm:4.1.1"
+  dependencies:
+    decompress-tar: "npm:^4.1.1"
+    file-type: "npm:^5.2.0"
+    is-stream: "npm:^1.1.0"
+  checksum: 10c0/42514fb2df6248c56b2b115494b7d1d046bc582e960354ba4faad5792f261782a61d17d9ef53845abe78c0f0ecafc195cb0754c00227fa0bd0642a1bfd8eafad
+  languageName: node
+  linkType: hard
+
+"decompress-unzip@npm:^4.0.1":
+  version: 4.0.1
+  resolution: "decompress-unzip@npm:4.0.1"
+  dependencies:
+    file-type: "npm:^3.8.0"
+    get-stream: "npm:^2.2.0"
+    pify: "npm:^2.3.0"
+    yauzl: "npm:^2.4.2"
+  checksum: 10c0/896f88e1c23b59cdce022227a8910c06158bd4b296c21d61af7167bd50d00e9e4355b605bdbfd7ba75d46ad277d4f881cdd037aec7165a40ccd0ee4ef59443a8
+  languageName: node
+  linkType: hard
+
+"decompress@npm:^4.2.0":
+  version: 4.2.1
+  resolution: "decompress@npm:4.2.1"
+  dependencies:
+    decompress-tar: "npm:^4.0.0"
+    decompress-tarbz2: "npm:^4.0.0"
+    decompress-targz: "npm:^4.0.0"
+    decompress-unzip: "npm:^4.0.1"
+    graceful-fs: "npm:^4.1.10"
+    make-dir: "npm:^1.0.0"
+    pify: "npm:^2.3.0"
+    strip-dirs: "npm:^2.0.0"
+  checksum: 10c0/6730279fa206aad04a8338a88ab49c596034c502b2d5f23a28d0a28290b82d9217f9e60c8b5739805474ca842fc856e08e2d64ed759f2118c2bcabe42fa9eece
+  languageName: node
+  linkType: hard
+
 "deep-is@npm:^0.1.3":
   version: 0.1.4
   resolution: "deep-is@npm:0.1.4"
@@ -4295,7 +4400,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"end-of-stream@npm:^1.1.0":
+"end-of-stream@npm:^1.0.0, end-of-stream@npm:^1.1.0":
   version: 1.4.4
   resolution: "end-of-stream@npm:1.4.4"
   dependencies:
@@ -4993,6 +5098,27 @@ __metadata:
   languageName: node
   linkType: hard
 
+"file-type@npm:^3.8.0":
+  version: 3.9.0
+  resolution: "file-type@npm:3.9.0"
+  checksum: 10c0/7ae074b350c2300807a99d428600a8ee6b2ace901400898706a20ddc2c43c9abb7e05177ff55ed67a2fd26dfa9b91857b21ec9c0ab3202b9cabebc7e65900240
+  languageName: node
+  linkType: hard
+
+"file-type@npm:^5.2.0":
+  version: 5.2.0
+  resolution: "file-type@npm:5.2.0"
+  checksum: 10c0/c16c2f4e484a838c12b63e08637277905f08aebb1afbc291086029210aea17ded5ed701c9a4588313446ae0c1da71566b58df9a9c758a1ec300c4f80b9713cbf
+  languageName: node
+  linkType: hard
+
+"file-type@npm:^6.1.0":
+  version: 6.2.0
+  resolution: "file-type@npm:6.2.0"
+  checksum: 10c0/3d7fe85a10bd97ca0c35fd9a20d21f5b20849bbb70985d37c34475051433f3c6109c76a3e5893bff6773037b769be9730a2db762789ecf25def9b62a4c2ee953
+  languageName: node
+  linkType: hard
+
 "file-url@npm:^2.0.0":
   version: 2.0.2
   resolution: "file-url@npm:2.0.2"
@@ -5147,6 +5273,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"fs-constants@npm:^1.0.0":
+  version: 1.0.0
+  resolution: "fs-constants@npm:1.0.0"
+  checksum: 10c0/a0cde99085f0872f4d244e83e03a46aa387b74f5a5af750896c6b05e9077fac00e9932fdf5aef84f2f16634cd473c63037d7a512576da7d5c2b9163d1909f3a8
+  languageName: node
+  linkType: hard
+
 "fs-extra@npm:^1.0.0":
   version: 1.0.0
   resolution: "fs-extra@npm:1.0.0"
@@ -5307,6 +5440,16 @@ __metadata:
   languageName: node
   linkType: hard
 
+"get-stream@npm:^2.2.0":
+  version: 2.3.1
+  resolution: "get-stream@npm:2.3.1"
+  dependencies:
+    object-assign: "npm:^4.0.1"
+    pinkie-promise: "npm:^2.0.0"
+  checksum: 10c0/46c12f496e7edec688a1cc570fe7556ce91e91201fa7efb146853fb9f0a8f0b0bb9a02cf9d9e4e9d4e2097f98c83b09621d9034c25ca0cf80ae6f4dace9c3465
+  languageName: node
+  linkType: hard
+
 "get-stream@npm:^5.1.0":
   version: 5.2.0
   resolution: "get-stream@npm:5.2.0"
@@ -5497,7 +5640,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"graceful-fs@npm:^4.1.2, graceful-fs@npm:^4.1.6, graceful-fs@npm:^4.1.9, graceful-fs@npm:^4.2.0, graceful-fs@npm:^4.2.2, graceful-fs@npm:^4.2.4, graceful-fs@npm:^4.2.6":
+"graceful-fs@npm:^4.1.10, graceful-fs@npm:^4.1.2, graceful-fs@npm:^4.1.6, graceful-fs@npm:^4.1.9, graceful-fs@npm:^4.2.0, graceful-fs@npm:^4.2.2, graceful-fs@npm:^4.2.4, graceful-fs@npm:^4.2.6":
   version: 4.2.11
   resolution: "graceful-fs@npm:4.2.11"
   checksum: 10c0/386d011a553e02bc594ac2ca0bd6d9e4c22d7fa8cfbfc448a6d148c59ea881b092db9dbe3547ae4b88e55f1b01f7c4a2ecc53b310c042793e63aa44cf6c257f2
@@ -6329,6 +6472,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"is-natural-number@npm:^4.0.1":
+  version: 4.0.1
+  resolution: "is-natural-number@npm:4.0.1"
+  checksum: 10c0/f05c544cb0ad39d4410e2ae2244282bf61918ebbb808b665436ffca4f6bbe908d3ae3a8d21fe143d302951f157d969986dd432098b63899561639fcd1ce1c280
+  languageName: node
+  linkType: hard
+
 "is-negative-zero@npm:^2.0.3":
   version: 2.0.3
   resolution: "is-negative-zero@npm:2.0.3"
@@ -6413,7 +6563,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"is-stream@npm:^1.0.1":
+"is-stream@npm:^1.0.1, is-stream@npm:^1.1.0":
   version: 1.1.0
   resolution: "is-stream@npm:1.1.0"
   checksum: 10c0/b8ae7971e78d2e8488d15f804229c6eed7ed36a28f8807a1815938771f4adff0e705218b7dab968270433f67103e4fef98062a0beea55d64835f705ee72c7002
@@ -7087,6 +7237,15 @@ __metadata:
   languageName: node
   linkType: hard
 
+"make-dir@npm:^1.0.0":
+  version: 1.3.0
+  resolution: "make-dir@npm:1.3.0"
+  dependencies:
+    pify: "npm:^3.0.0"
+  checksum: 10c0/5eb94f47d7ef41d89d1b8eef6539b8950d5bd99eeba093a942bfd327faa37d2d62227526b88b73633243a2ec7972d21eb0f4e5d62ae4e02a79e389f4a7bb3022
+  languageName: node
+  linkType: hard
+
 "make-fetch-happen@npm:^13.0.0":
   version: 13.0.1
   resolution: "make-fetch-happen@npm:13.0.1"
@@ -8053,6 +8212,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"node-ensure@npm:^0.0.0":
+  version: 0.0.0
+  resolution: "node-ensure@npm:0.0.0"
+  checksum: 10c0/7af391aee024a8b7df77c239ed8b90417e3f2539824fa06b60f243ce14c75ee455766464c7c3ba9407d5b1e4d1d74ed5cf5f8af10c67b0fc05aa6e29f5d2462b
+  languageName: node
+  linkType: hard
+
 "node-fetch@npm:^2.6.7":
   version: 2.7.0
   resolution: "node-fetch@npm:2.7.0"
@@ -8152,7 +8318,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"object-assign@npm:^4.1.1":
+"object-assign@npm:^4.0.1, object-assign@npm:^4.1.1":
   version: 4.1.1
   resolution: "object-assign@npm:4.1.1"
   checksum: 10c0/1f4df9945120325d041ccf7b86f31e8bcc14e73d29171e37a7903050e96b81323784ec59f93f102ec635bcf6fa8034ba3ea0a8c7e69fa202b87ae3b6cec5a414
@@ -8219,6 +8385,21 @@ __metadata:
   languageName: node
   linkType: hard
 
+"officeparser@npm:^4.1.1":
+  version: 4.1.1
+  resolution: "officeparser@npm:4.1.1"
+  dependencies:
+    "@xmldom/xmldom": "npm:^0.8.10"
+    decompress: "npm:^4.2.0"
+    file-type: "npm:^16.5.4"
+    node-ensure: "npm:^0.0.0"
+    rimraf: "npm:^2.6.3"
+  bin:
+    officeparser: officeParser.js
+  checksum: 10c0/7c5f01b3fe74fd31ca6ae6a6f46b2336d953eece3ca61d171e417b50d7e1962268af630bef0c8d40e39a1c471e7146281eecd880f0eecc576895eee9bc20b6a7
+  languageName: node
+  linkType: hard
+
 "omggif@npm:^1.0.10, omggif@npm:^1.0.9":
   version: 1.0.10
   resolution: "omggif@npm:1.0.10"
@@ -8610,13 +8791,20 @@ __metadata:
   languageName: node
   linkType: hard
 
-"pify@npm:^2.0.0":
+"pify@npm:^2.0.0, pify@npm:^2.3.0":
   version: 2.3.0
   resolution: "pify@npm:2.3.0"
   checksum: 10c0/551ff8ab830b1052633f59cb8adc9ae8407a436e06b4a9718bcb27dc5844b83d535c3a8512b388b6062af65a98c49bdc0dd523d8b2617b188f7c8fee457158dc
   languageName: node
   linkType: hard
 
+"pify@npm:^3.0.0":
+  version: 3.0.0
+  resolution: "pify@npm:3.0.0"
+  checksum: 10c0/fead19ed9d801f1b1fcd0638a1ac53eabbb0945bf615f2f8806a8b646565a04a1b0e7ef115c951d225f042cca388fdc1cd3add46d10d1ed6951c20bd2998af10
+  languageName: node
+  linkType: hard
+
 "pinkie-promise@npm:^2.0.0":
   version: 2.0.1
   resolution: "pinkie-promise@npm:2.0.1"
@@ -9664,7 +9852,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"readable-stream@npm:^2.0.2, readable-stream@npm:^2.0.5, readable-stream@npm:^2.2.2, readable-stream@npm:~2.3.6":
+"readable-stream@npm:^2.0.2, readable-stream@npm:^2.0.5, readable-stream@npm:^2.2.2, readable-stream@npm:^2.3.0, readable-stream@npm:^2.3.5, readable-stream@npm:~2.3.6":
   version: 2.3.8
   resolution: "readable-stream@npm:2.3.8"
   dependencies:
@@ -10073,6 +10261,17 @@ __metadata:
   languageName: node
   linkType: hard
 
+"rimraf@npm:^2.6.3":
+  version: 2.7.1
+  resolution: "rimraf@npm:2.7.1"
+  dependencies:
+    glob: "npm:^7.1.3"
+  bin:
+    rimraf: ./bin.js
+  checksum: 10c0/4eef73d406c6940927479a3a9dee551e14a54faf54b31ef861250ac815172bade86cc6f7d64a4dc5e98b65e4b18a2e1c9ff3b68d296be0c748413f092bb0dd40
+  languageName: node
+  linkType: hard
+
 "rimraf@npm:^3.0.2":
   version: 3.0.2
   resolution: "rimraf@npm:3.0.2"
@@ -10196,7 +10395,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"safe-buffer@npm:^5.0.1, safe-buffer@npm:^5.1.2, safe-buffer@npm:~5.2.0":
+"safe-buffer@npm:^5.0.1, safe-buffer@npm:^5.1.1, safe-buffer@npm:^5.1.2, safe-buffer@npm:~5.2.0":
   version: 5.2.1
   resolution: "safe-buffer@npm:5.2.1"
   checksum: 10c0/6501914237c0a86e9675d4e51d89ca3c21ffd6a31642efeba25ad65720bce6921c9e7e974e5be91a786b25aa058b5303285d3c15dbabf983a919f5f630d349f3
@@ -10284,6 +10483,18 @@ __metadata:
   languageName: node
   linkType: hard
 
+"seek-bzip@npm:^1.0.5":
+  version: 1.0.6
+  resolution: "seek-bzip@npm:1.0.6"
+  dependencies:
+    commander: "npm:^2.8.1"
+  bin:
+    seek-bunzip: bin/seek-bunzip
+    seek-table: bin/seek-bzip-table
+  checksum: 10c0/e4019e4498bb725ff855603595c4116ca003674b13d29cb9ed9891b2ceec884ccf7c1cb5dba0d6b50fe6ce760a011078f5744efb79823f4ddb9decb1571e9912
+  languageName: node
+  linkType: hard
+
 "semver-compare@npm:^1.0.0":
   version: 1.0.0
   resolution: "semver-compare@npm:1.0.0"
@@ -10776,6 +10987,15 @@ __metadata:
   languageName: node
   linkType: hard
 
+"strip-dirs@npm:^2.0.0":
+  version: 2.1.0
+  resolution: "strip-dirs@npm:2.1.0"
+  dependencies:
+    is-natural-number: "npm:^4.0.1"
+  checksum: 10c0/073d6d08331ec2e87afc2c2535d7336fee1d63797384045e4ecb9908a5ac6615022ee000cc278d6bbc94147bed7350f7cf4657b6d18c377813f37e7ae329fb52
+  languageName: node
+  linkType: hard
+
 "strip-json-comments@npm:^3.1.1":
   version: 3.1.1
   resolution: "strip-json-comments@npm:3.1.1"
@@ -10908,6 +11128,21 @@ __metadata:
   languageName: node
   linkType: hard
 
+"tar-stream@npm:^1.5.2":
+  version: 1.6.2
+  resolution: "tar-stream@npm:1.6.2"
+  dependencies:
+    bl: "npm:^1.0.0"
+    buffer-alloc: "npm:^1.2.0"
+    end-of-stream: "npm:^1.0.0"
+    fs-constants: "npm:^1.0.0"
+    readable-stream: "npm:^2.3.0"
+    to-buffer: "npm:^1.1.1"
+    xtend: "npm:^4.0.0"
+  checksum: 10c0/ab8528d2cc9ccd0906d1ce6d8089030b2c92a578c57645ff4971452c8c5388b34c7152c04ed64b8510d22a66ffaf0fee58bada7d6ab41ad1e816e31993d59cf3
+  languageName: node
+  linkType: hard
+
 "tar-stream@npm:^3.0.0":
   version: 3.1.7
   resolution: "tar-stream@npm:3.1.7"
@@ -10982,6 +11217,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"through@npm:^2.3.8":
+  version: 2.3.8
+  resolution: "through@npm:2.3.8"
+  checksum: 10c0/4b09f3774099de0d4df26d95c5821a62faee32c7e96fb1f4ebd54a2d7c11c57fe88b0a0d49cf375de5fee5ae6bf4eb56dbbf29d07366864e2ee805349970d3cc
+  languageName: node
+  linkType: hard
+
 "timm@npm:^1.6.1":
   version: 1.7.1
   resolution: "timm@npm:1.7.1"
@@ -11026,6 +11268,13 @@ __metadata:
   languageName: node
   linkType: hard
 
+"to-buffer@npm:^1.1.1":
+  version: 1.1.1
+  resolution: "to-buffer@npm:1.1.1"
+  checksum: 10c0/fb9fc6a0103f2b06e2e01c3d291586d0148759d5584f35d0973376434d1b58bd6ee5df9273f0bb1190eb2a5747c894bf49fed571325a7ac10208a48f31736439
+  languageName: node
+  linkType: hard
+
 "to-fast-properties@npm:^2.0.0":
   version: 2.0.0
   resolution: "to-fast-properties@npm:2.0.0"
@@ -11280,6 +11529,16 @@ __metadata:
   languageName: node
   linkType: hard
 
+"unbzip2-stream@npm:^1.0.9":
+  version: 1.4.3
+  resolution: "unbzip2-stream@npm:1.4.3"
+  dependencies:
+    buffer: "npm:^5.2.1"
+    through: "npm:^2.3.8"
+  checksum: 10c0/2ea2048f3c9db3499316ccc1d95ff757017ccb6f46c812d7c42466247e3b863fb178864267482f7f178254214247779daf68e85f50bd7736c3c97ba2d58b910a
+  languageName: node
+  linkType: hard
+
 "undici-types@npm:~5.26.4":
   version: 5.26.5
   resolution: "undici-types@npm:5.26.5"
@@ -12101,7 +12360,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"yauzl@npm:^2.10.0":
+"yauzl@npm:^2.10.0, yauzl@npm:^2.4.2":
   version: 2.10.0
   resolution: "yauzl@npm:2.10.0"
   dependencies: