first commit

2024-08-12 17:08:47 +02:00
commit acd1f8cc7e
6 changed files with 141 additions and 0 deletions
--- a/.gitignore
+++ b/.gitignore
@@ -0,0 +1 @@
+out/
--- a/app.ts
+++ b/app.ts
@@ -0,0 +1,62 @@
+export default class App {
+    input: HTMLInputElement;
+    outputContainer: HTMLTextAreaElement;
+
+    constructor() {
+        this.input = document.getElementById('in') as HTMLInputElement;
+        this.outputContainer = document.getElementById('out') as HTMLTextAreaElement;
+        (document.getElementById("run") as HTMLButtonElement).onclick = this.run.bind(this);
+    }
+
+    async run() {
+        const promises: Promise<string>[] = [];
+        for (const file of this.input.files!) {
+            promises.push(this.parseFile(file));
+        }
+
+        const parts = await Promise.all(promises);
+
+        const output = new Blob([parts.join(' ')], { type: 'text/plain' });
+        const a = document.createElement('a');
+        a.href = URL.createObjectURL(output);
+        a.download = "result.txt";
+        a.click();
+    }
+
+    async parseFile(file: File) {
+        const xml = await file.text();
+        const parser = new DOMParser();
+        const doc = parser.parseFromString(xml, "text/xml");
+
+        const results: string[] = [];
+
+        for (const node of doc.getElementsByTagNameNS('*', 'tartalom')) {
+            let resolve: (value: string) => void, reject: (reason?: any) => void;
+            const resultPromise = new Promise<string>((res, rej) => {
+                resolve = res, reject = rej;
+            });
+
+            const worker = new Worker('worker.js');
+
+            worker.onmessage = (e) => {
+                resolve(e.data);
+            };
+
+            worker.onerror = (e) => {
+                reject(e.error);
+            }
+
+            worker.postMessage(node.textContent);
+
+            try {
+                results.push(await resultPromise);
+            } catch (e) {
+                console.error(e);
+            }
+        }
+
+        return results.join(' ');
+    }
+}
+
+new App();
--- a/index.html
+++ b/index.html
@@ -0,0 +1,37 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <meta name="description" content="XML Cleaner">
+    <title>XML Cleaner</title>
+    <style>
+        body,
+        html {
+            margin: auto;
+            height: 100%;
+        }
+
+        body {
+            color: white;
+            background-color: black;
+            display: flex;
+            max-width: 800px;
+            justify-content: center;
+            align-items: center;
+        }
+
+        textarea {
+            width: 100%;
+            height: 100%;
+        }
+    </style>
+</head>
+
+<body>
+    <input type="file" id="in" multiple accept="text/xml"><button id="run">RUN</button>
+    <script type="module" src="out/app.js"></script>
+</body>
+
+</html>
--- a/readme.md
+++ b/readme.md
@@ -0,0 +1,16 @@
+# XML Cleaner
+
+Simple webapp to clean up XML datasets.
+
+Currently hardcoded to find the `tartalom` tags in multiple files,  
+and dump their text content into a single txt file.
+
+Removes all xml tags contained in CDATA strings, along with unnecessary whitespaces.
+
+## Building
+1. Run the Typescript compiler.
+```
+tsc
+```
+
+2. Deploy `index.html` and the `out` folder to a webserver.
--- a/tsconfig.json
+++ b/tsconfig.json
@@ -0,0 +1,11 @@
+{
+  "compilerOptions": {
+    "target": "ESNext",
+    "module": "ESNext",
+    "outDir": "./out",
+    "esModuleInterop": true,
+    "forceConsistentCasingInFileNames": true,
+    "strict": true,
+    "skipLibCheck": true
+  }
+}
--- a/worker.ts
+++ b/worker.ts
@@ -0,0 +1,14 @@
+const singleTagRegEx = /<[^>]+?>/g;
+const whitespaceRegEx = /(\s)\s+/g;
+
+function clean(text: string): string {
+    text = text.replaceAll(whitespaceRegEx, "$1");
+    text = text.replaceAll(singleTagRegEx, "");
+
+    return text;
+}
+
+onmessage = (e) => {
+    const result = clean(e.data);
+    postMessage(result);
+}