dynamic import of string-strip-html

packages/core/package.json

@@ -3,33 +3,33 @@
   "version": "0.0.31",
   "license": "MIT",
   "dependencies": {
-    "@anthropic-ai/sdk": "^0.8.0",
+    "@anthropic-ai/sdk": "^0.8.1",
     "@notionhq/client": "^2.2.13",
     "lodash": "^4.17.21",
     "mammoth": "^1.6.0",
     "md-utils-ts": "^2.0.0",
     "mongodb": "^6.2.0",
     "notion-md-crawler": "^0.0.2",
-    "openai": "^4.13.0",
+    "openai": "^4.14.0",
     "papaparse": "^5.4.1",
     "pdf-parse": "^1.1.1",
     "portkey-ai": "^0.1.13",
     "rake-modified": "^1.0.8",
     "replicate": "^0.20.1",
-    "string-strip-html": "^8.5.0",
+    "string-strip-html": "^13.4.3",
     "tiktoken": "^1.0.10",
     "uuid": "^9.0.1",
     "wink-nlp": "^1.14.3"
   },
   "devDependencies": {
     "@types/lodash": "^4.14.200",
-    "@types/node": "^18.18.6",
+    "@types/node": "^18.18.7",
     "@types/papaparse": "^5.3.10",
     "@types/pdf-parse": "^1.1.3",
     "@types/uuid": "^9.0.6",
     "node-stdlib-browser": "^1.2.0",
     "tsup": "^7.2.0",
-    "typescript": "^4.9.5"
+    "typescript": "^5.2.2"
   },
   "engines": {
     "node": ">=18.0.0"

packages/core/src/readers/HTMLReader.ts

-import { stripHtml } from "string-strip-html";
 import { Document } from "../Node";
 import { DEFAULT_FS } from "../storage/constants";
 import { GenericFileSystem } from "../storage/FileSystem";
@@ -25,7 +24,7 @@ export class HTMLReader implements BaseReader {
   ): Promise<Document[]> {
     const dataBuffer = await fs.readFile(file, "utf-8");
     const htmlOptions = this.getOptions();
-    const content = this.parseContent(dataBuffer, htmlOptions);
+    const content = await this.parseContent(dataBuffer, htmlOptions);
     return [new Document({ text: content, id_: file })];
   }
 
@@ -36,7 +35,8 @@ export class HTMLReader implements BaseReader {
    * @see getOptions
    * @returns The HTML content, stripped of unwanted tags and attributes
    */
-  parseContent(html: string, options: any = {}): string {
+  async parseContent(html: string, options: any = {}): Promise<string> {
+    const { stripHtml } = await import("string-strip-html"); // ESM only
     return stripHtml(html).result;
   }
 

packages/core/tsconfig.json

@@ -3,6 +3,7 @@
     "esModuleInterop": true,
     "forceConsistentCasingInFileNames": true,
     "isolatedModules": true,
+    "module": "esnext",
     "moduleResolution": "node",
     "preserveWatchOutput": true,
     "skipLibCheck": true,