janus-parse 1.0.3 → 1.1.1

package/README.md

@@ -1,4 +1,4 @@
-# janus-parse
+# janus-parse [![npm version](https://img.shields.io/npm/v/janus-parse?style=flat-round)](https://www.npmjs.com/package/janus-parse) ![language](https://img.shields.io/badge/Language-Typescript-blue?style=flat-round) [![release](https://github.com/Casneil/janus-parse/actions/workflows/publish.yml/badge.svg)](https://github.com/Casneil/janus-parse/actions/workflows/publish.yml) [![build](https://github.com/Casneil/janus-parse/actions/workflows/build.yml/badge.svg)](https://github.com/Casneil/janus-parse/actions/workflows/build.yml)
 
 A lightweight, high-performance, isomorphic HTML text extractor and sanitizer built entirely in TypeScript. It safely strips out targeted layout elements or malicious scripts and returns clean, whitespace-normalized text.
 

package/dist/index.d.ts

@@ -1,2 +1,2 @@
-export * from "./janus-parse.js";
+export * from "./janus-parse";
 //# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../lib/index.ts"],"names":[],"mappings":"AAAA,cAAc,kBAAkB,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../lib/index.ts"],"names":[],"mappings":"AAAA,cAAc,eAAe,CAAC"}

package/dist/index.js

@@ -1 +1 @@
-export * from "./janus-parse.js";
+export * from "./janus-parse";

package/dist/janus-parse.d.ts

@@ -1,4 +1,4 @@
-import { Config } from "./utils/index.js";
+import type { Config } from "./utils";
 export declare function janusServer(text: string, config?: Config): Promise<string>;
-export declare function janusClient(text: string, config?: Config): string | null;
+export declare function janusClient(text: string, config?: Config): string;
 //# sourceMappingURL=janus-parse.d.ts.map

package/dist/janus-parse.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"janus-parse.d.ts","sourceRoot":"","sources":["../lib/janus-parse.ts"],"names":[],"mappings":"AAAA,OAAO,EAA6B,MAAM,EAAW,MAAM,kBAAkB,CAAC;AAO9E,wBAAsB,WAAW,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,GAAE,MAAoB,mBAU3E;AAED,wBAAgB,WAAW,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,GAAE,MAAoB,iBAYrE"}
+{"version":3,"file":"janus-parse.d.ts","sourceRoot":"","sources":["../lib/janus-parse.ts"],"names":[],"mappings":"AAMA,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AAOtC,wBAAsB,WAAW,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,GAAE,MAAoB,mBAS3E;AAED,wBAAgB,WAAW,CAAC,IAAI,EAAE,MAAM,EAAE,MAAM,GAAE,MAAoB,UAUrE"}

package/dist/janus-parse.js

@@ -1,4 +1,4 @@
-import { validateText, removeNodes, getTags } from "./utils/index.js";
+import { validateText, removeNodes, getTags, normalizeWhitespace, } from "./utils";
 const janusConfig = {
     addBlacklistTags: [],
     removeBlacklistTags: [],
@@ -9,17 +9,15 @@ export async function janusServer(text, config = janusConfig) {
     const parser = await import("node-html-parser");
     const root = parser.parse(text);
     const nodesToRemove = root.querySelectorAll(tags);
-    // @ts-expect-error
     removeNodes(nodesToRemove);
-    return root.textContent.replace(new RegExp("\\s+", "g"), " ").trim();
+    return normalizeWhitespace(root.textContent);
 }
 export function janusClient(text, config = janusConfig) {
     validateText(text);
     const tags = getTags(config);
     const parser = new DOMParser();
-    const virtualDoc = parser.parseFromString(text, "text/html");
-    const nodesToRemove = virtualDoc.querySelectorAll(tags);
+    const virtualDocument = parser.parseFromString(text, "text/html");
+    const nodesToRemove = virtualDocument.querySelectorAll(tags);
     removeNodes(nodesToRemove);
-    const cleanText = virtualDoc.body.textContent;
-    return cleanText;
+    return normalizeWhitespace(virtualDocument.body.textContent);
 }

package/dist/utils/index.d.ts

@@ -1,8 +1,10 @@
+import type { HTMLElement as NodeHTMLElement } from "node-html-parser";
 export type Config = {
     addBlacklistTags?: string[];
     removeBlacklistTags?: string[];
 };
 export declare function validateText(text: string): void;
+export declare function normalizeWhitespace(text?: string): string;
 export declare function getTags(config: Config): string;
-export declare function removeNodes(nodesToRemove: HTMLElement[] | NodeListOf<Element>): NodeListOf<Element> | HTMLElement[];
+export declare function removeNodes(nodesToRemove: NodeHTMLElement[] | NodeListOf<Element>): NodeListOf<Element> | NodeHTMLElement[];
 //# sourceMappingURL=index.d.ts.map

package/dist/utils/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../lib/utils/index.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,MAAM,GAAG;IACnB,gBAAgB,CAAC,EAAE,MAAM,EAAE,CAAC;IAC5B,mBAAmB,CAAC,EAAE,MAAM,EAAE,CAAC;CAChC,CAAC;AAIF,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,QAUxC;AAED,wBAAgB,OAAO,CAAC,MAAM,EAAE,MAAM,UAWrC;AAED,wBAAgB,WAAW,CACzB,aAAa,EAAE,WAAW,EAAE,GAAG,UAAU,CAAC,OAAO,CAAC,uCAOnD"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../lib/utils/index.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,WAAW,IAAI,eAAe,EAAE,MAAM,kBAAkB,CAAC;AAEvE,MAAM,MAAM,MAAM,GAAG;IACnB,gBAAgB,CAAC,EAAE,MAAM,EAAE,CAAC;IAC5B,mBAAmB,CAAC,EAAE,MAAM,EAAE,CAAC;CAChC,CAAC;AAIF,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,QAUxC;AAED,wBAAgB,mBAAmB,CAAC,IAAI,SAAK,UAE5C;AAED,wBAAgB,OAAO,CAAC,MAAM,EAAE,MAAM,UAWrC;AAED,wBAAgB,WAAW,CACzB,aAAa,EAAE,eAAe,EAAE,GAAG,UAAU,CAAC,OAAO,CAAC,2CAOvD"}

package/dist/utils/index.js

@@ -5,6 +5,9 @@ export function validateText(text) {
         throw new Error("The text is not a valid string. Please provide a valid string to parse.", { cause: text });
     }
 }
+export function normalizeWhitespace(text = "") {
+    return text.replace(new RegExp(String.raw `\s+`, "g"), " ").trim();
+}
 export function getTags(config) {
     let tags = new Set();
     if (config.addBlacklistTags) {

package/package.json

@@ -1,9 +1,10 @@
 {
   "name": "janus-parse",
   "license": "MIT",
-  "version": "1.0.3",
+  "version": "1.1.1",
   "main": "./dist/index.js",
   "types": "./dist/index.d.ts",
+  "packageManager": "pnpm@11.15.1",
   "keywords": [
     "html-parser",
     "text-extractor",
@@ -19,6 +20,17 @@
   "files": [
     "dist"
   ],
+  "release": {
+    "branches": [
+      "main"
+    ],
+    "plugins": [
+      "@semantic-release/commit-analyzer",
+      "@semantic-release/release-notes-generator",
+      "@semantic-release/npm",
+      "@semantic-release/github"
+    ]
+  },
   "repository": {
     "type": "git",
     "url": "https://github.com/Casneil/janus-parse"
@@ -27,25 +39,42 @@
   "bugs": {
     "url": "https://github.com/Casneil/janus-parse/issues"
   },
+  "publishConfig": {
+    "registry": "https://registry.npmjs.org"
+  },
   "type": "module",
   "prepack": "pnpm build",
+  "scripts": {
+    "janus": "tsx ./lib",
+    "test": "vitest",
+    "code:lint": "pnpm eslint --flag unstable_native_nodejs_ts_config --cache .",
+    "fmt": "prettier --write .",
+    "type:check": "tsc --noEmit",
+    "build": "tsc",
+    "prepublishOnly": "pnpm build"
+  },
   "dependencies": {
     "node-html-parser": "9.0.0"
   },
   "devDependencies": {
+    "@eslint/js": "^10.0.1",
+    "@semantic-release/commit-analyzer": "^13.0.1",
+    "@semantic-release/github": "^12.0.9",
+    "@semantic-release/npm": "^13.1.5",
+    "@semantic-release/release-notes-generator": "^14.1.1",
     "@types/node": "^26.1.1",
     "eslint": "^10.7.0",
+    "eslint-import-resolver-typescript": "^4.4.5",
+    "eslint-plugin-import-x": "^4.17.1",
+    "eslint-plugin-unicorn": "^72.0.0",
     "happy-dom": "^20.11.0",
     "prettier": "^3.9.5",
+    "semantic-release": "^25.0.8",
     "tsx": "4.23.1",
-    "typescript": "5.7.2",
+    "typescript": "6.0.3",
+    "typescript-eslint": "^8.65.0",
     "vitest": "^4.1.10"
   },
   "prettier": {},
-  "scripts": {
-    "janus": "tsx ./lib/index.ts",
-    "test": "vitest",
-    "fmt": "prettier --write .",
-    "build": "tsc"
-  }
-}
+  "--Typescript V 7--": "cannot use TS v7 until typescript-eslint adds support"
+}