bundle mammoth in core (#281)

* bundle mammoth * don't copy mammoth in node
microsoft · Mar 21, 2024 · 11701e3 · 11701e3
1 parent bd2d11d
commit 11701e3
Show file tree

Hide file tree

Showing 6 changed files with 5 additions and 29 deletions.
diff --git a/packages/cli/package.json b/packages/cli/package.json
@@ -25,13 +25,13 @@
   "optionalDependencies": {
     "llamaindex": "0.1.21",
     "llm-code-highlighter": "0.0.13",
-    "mammoth": "^1.7.0",
     "pdfjs-dist": "4.0.379"
   },
   "engines": {
     "node": ">=20.0.0"
   },
   "devDependencies": {
+    "mammoth": "^1.7.0",
     "commander": "^12.0.0",
     "dotenv": "^16.4.5",
     "execa": "^8.0.1",
@@ -58,7 +58,7 @@
     "zx": "^7.2.3"
   },
   "scripts": {
-    "compile": "esbuild src/main.ts --bundle --platform=node --target=node20 --outfile=built/genaiscript.cjs --external:pdfjs-dist --external:mammoth --external:llamaindex --external:llm-code-highlighter && echo '#!/usr/bin/env node' | cat - built/genaiscript.cjs > temp && mv temp built/genaiscript.cjs",
+    "compile": "esbuild src/main.ts --bundle --platform=node --target=node20 --outfile=built/genaiscript.cjs --external:pdfjs-dist --external:llamaindex --external:llm-code-highlighter && echo '#!/usr/bin/env node' | cat - built/genaiscript.cjs > temp && mv temp built/genaiscript.cjs",
     "postcompile": "cp built/genaiscript.cjs ../vscode/genaiscript.cjs && node built/genaiscript.cjs help-all > ../../docs/src/content/docs/reference/cli/commands.md",
     "go": "yarn compile && node built/genaiscript.cjs",
     "test": "node --import tsx --test src/**.test.ts",

diff --git a/packages/core/src/docx.ts b/packages/core/src/docx.ts
@@ -1,25 +1,7 @@
 import { host } from "./host"
 import { MarkdownTrace } from "./trace"
-import { installImport } from "./import"
 import { logError } from "./util"
-import { MAMMOTH_VERSION } from "./version"
-
-async function tryImportMammoth(trace?: MarkdownTrace) {
-    try {
-        const mod = await import("mammoth")
-        return mod
-    } catch (e) {
-        trace?.error(`mammoth not found, installing ${MAMMOTH_VERSION}...`)
-        try {
-            await installImport("mammoth", MAMMOTH_VERSION, trace)
-            const mod = await import("mammoth")
-            return mod
-        } catch (e) {
-            trace?.error("mammoth failed to load")
-            return undefined
-        }
-    }
-}
+import { extractRawText } from "mammoth"
 
 /**
  * parses docx, require mammoth to be installed
@@ -33,11 +15,10 @@ export async function DOCXTryParse(
 ): Promise<string> {
     const { trace } = options || {}
     try {
-        const mammoth = await tryImportMammoth(trace)
         const path = !/^\//.test(file)
             ? host.path.join(host.projectFolder(), file)
             : file
-        const results = await mammoth.extractRawText({ path })
+        const results = await extractRawText({ path })
         return results.value
     } catch (error) {
         logError(error.message)

diff --git a/packages/core/src/version.ts b/packages/core/src/version.ts
@@ -11,4 +11,3 @@ export const CORE_VERSION = packageJson.version
 export const GITHUB_REPO = packageJson.repository.url
 
 export const PDFJS_DIST_VERSION = packageJson.optionalDependencies["pdfjs-dist"]
-export const MAMMOTH_VERSION = packageJson.optionalDependencies["mammoth"]
diff --git a/packages/vscode/package.json b/packages/vscode/package.json
@@ -313,7 +313,7 @@
     "scripts": {
         "vscode:update-dts": "npx @vscode/dts dev",
         "vscode:prepublish": "yarn run compile",
-        "compile:extension": "esbuild src/extension.ts --sourcemap --bundle --format=cjs --platform=node --target=node20 --outfile=built/extension.js --external:vscode --external:pdfjs-dist --external:mammoth --external:llamaindex",
+        "compile:extension": "esbuild src/extension.ts --sourcemap --bundle --format=cjs --platform=node --target=node20 --outfile=built/extension.js --external:vscode --external:pdfjs-dist --external:llamaindex",
         "compile": "yarn compile:extension",
         "lint": "eslint src --ext ts",
         "run-in-browser": "vscode-test-web --browserType=chromium --extensionDevelopmentPath=. ../sample",

diff --git a/packages/vscode/postpackage.mjs b/packages/vscode/postpackage.mjs
@@ -1,7 +1,6 @@
 import 'zx/globals'
 
 await $`rm -Rf ./node_modules/pdfjs-dist`
-await $`rm -Rf ./node_modules/mammoth`
 
 const pkg = await fs.readJSON('./package.json')
 pkg.enabledApiProposals = pkg._enabledApiProposals

diff --git a/packages/vscode/prepackage.mjs b/packages/vscode/prepackage.mjs
@@ -4,9 +4,6 @@ import 'zx/globals'
 await fs.copy('../../node_modules/pdfjs-dist/package.json', './node_modules/pdfjs-dist/package.json', { overwrite: true })
 await $`cp -R ../../node_modules/pdfjs-dist/build ./node_modules/pdfjs-dist`
 
-await fs.copy('../../node_modules/mammoth/package.json', './node_modules/mammoth/package.json', { overwrite: true })
-await $`cp -R ../../node_modules/mammoth/lib ./node_modules/mammoth`
-
 const pkg = await fs.readJSON('./package.json')
 pkg._enabledApiProposals = pkg.enabledApiProposals
 pkg.displayName = "GenAIScript"
Original file line number	Diff line number	Diff line change
Expand Up		@@ -11,4 +11,3 @@ export const CORE_VERSION = packageJson.version
		export const GITHUB_REPO = packageJson.repository.url

		export const PDFJS_DIST_VERSION = packageJson.optionalDependencies["pdfjs-dist"]
		export const MAMMOTH_VERSION = packageJson.optionalDependencies["mammoth"]