From 27b8e7cdf1a9c36a76cece5325a625d228b6e51d Mon Sep 17 00:00:00 2001
From: alberto <albedef@protonmail.com>
Date: Sun, 9 Mar 2025 09:22:18 +0100
Subject: [PATCH] Add parseunimap script.

---
 README.md      | 22 +++++++++++++++++++
 parseunimap.js | 58 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 80 insertions(+)
 create mode 100644 parseunimap.js

diff --git a/README.md b/README.md
index 74590ee..5b696c1 100644
--- a/README.md
+++ b/README.md
@@ -105,3 +105,25 @@ Questa repository contiene i seguenti script:
     ```
 
     Qui un esempio di utilizzo: [Argomenti di G2](https://git.phc.dm.unipi.it/lukefleed/domande-orali/src/branch/master/geoemtria-2)
+
+-   `parseunimap` &mdash; [@aziis98](https://git.phc.dm.unipi.it/aziis98) + [@alberto](https://git.phc.dm.unipi.it/alberto)
+
+    Script che converte il registro delle lezioni dell'Unipi in markdown
+
+    Installare:
+    ```bash
+    bun init
+    bun install jsdom
+    bun install node-fetch
+    ```
+
+    Per eseguire:
+    ```bash
+    bun run script.js <registro_link>
+    ```
+    Una volta ottenuto in output il file markdown, è possibile convertirlo in pdf utilizzando ad
+    esempio Pandoc tramite il comando
+
+    ```bash
+    pandoc fileregistro.md -o fileregistro.pdf
+    ```
diff --git a/parseunimap.js b/parseunimap.js
new file mode 100644
index 0000000..f5a5648
--- /dev/null
+++ b/parseunimap.js
@@ -0,0 +1,58 @@
+import {createWriteStream} from 'fs';
+import {JSDOM} from "jsdom";
+import fetch from "node-fetch";
+
+(async () => {
+  try {
+    const url = process.argv[2];
+    if (!url) {
+      console.error("Usage: node script.js <registro_link>");
+      process.exit(1);
+    }
+
+    const response =
+        await fetch(url, {headers : {"User-Agent" : "Mozilla/5.0"}});
+    const buffer = await response.arrayBuffer();
+    const isoHtml = Buffer.from(buffer, "binary").toString("latin1");
+
+    const dom = new JSDOM(isoHtml);
+    const document = dom.window.document;
+
+    let ol = document.querySelector("ol");
+    if (!ol) {
+      console.error("No <ol> found on the page");
+      return;
+    }
+
+    ol.querySelectorAll("a, i").forEach(el => el.remove());
+
+    const results = [...ol.children ]
+                        .map(li => li.textContent.match(/\w+:(.+)/sm))
+                        .filter(match => match)
+                        .map(match => match[1].trim())
+                        .map((x, index) => (index + 1) + ". " + x + " \n");
+
+    const writeStream = createWriteStream("registro.md")
+    const pathName = writeStream.path;
+    writeStream.on('error', function(err) { /* error handling */ });
+    const head = "# Registro delle lezioni \n\n";
+    writeStream.write(head)
+    results.forEach(value => writeStream.write(`${value}`));
+    // the finish event is emitted when all data has been flushed from the
+    // stream
+    writeStream.on(
+        'finish',
+        () => { console.log(`Registro salvato nel file ${pathName}`); });
+
+    // handle the errors on the write process
+    writeStream.on(
+        'error',
+        (err) => {console.error(
+            `There is an error writing the file ${pathName} => ${err}`)});
+
+    // close the stream
+    writeStream.end();
+  } catch (error) {
+    console.error("Error fetching or processing page:", error);
+  }
+})();