From 4e1e3836c20f1423c1df56b17be38ea42961867d Mon Sep 17 00:00:00 2001
From: Xevion <xevion@xevion.dev>
Date: Fri, 11 Aug 2023 17:39:34 -0500
Subject: [PATCH] Add html entity decoding, class name pattern parsing

---
 index.ts     | 26 ++++++++++++++++++++++++--
 package.json |  1 +
 yarn.lock    |  5 +++++
 3 files changed, 30 insertions(+), 2 deletions(-)

diff --git a/index.ts b/index.ts
index caecb63..8ef3738 100644
--- a/index.ts
+++ b/index.ts
@@ -2,17 +2,22 @@ import {Cheerio, Element, load} from "cheerio";
 import {readFileSync} from "fs";
 import {parse} from "date-fns";
 import {inspect} from "util";
+import {decode} from "html-entities";
 import fetch from "node-fetch";
 import {z} from "zod";
 
 const $ = load(readFileSync('list.html'));
 const classes = $('#scheduleListView').children('.listViewWrapper');
 
+const name_pattern = /(.+) (\d{4}) Section (\d+)/;
+
 const subject_schema = z.object({
     code: z.string(),
-    description: z.string(),
+    description: z.string().transform((v) => decode(v, {level: 'all'})),
 });
 
+type Subject = z.infer<typeof subject_schema>;
+
 
 const subjects = await z.array(subject_schema).parseAsync(
     await fetch(
@@ -27,6 +32,8 @@ const subjects = await z.array(subject_schema).parseAsync(
     })
 );
 
+const subject_by_name = new Map(subjects.map((subject) => [subject.description, subject]));
+
 
 function getOffset(period: string) {
     switch (period) {
@@ -61,9 +68,24 @@ function extractDetails(source: Cheerio<Element>) {
     const [start_date, end_date] = raw_date.split("--").map((date) => parse(date.trim(), "MM/dd/yyyy", new Date()));
 
     const identifier = source.find('span.list-view-subj-course-section').text();
+    const name_match = name_pattern.exec(identifier);
+    if (name_match === null) {
+        throw new Error(`Could not parse identifier: ${identifier}`);
+    }
+
+    const [subject_name, code, section] = name_match.slice(1);
+    const subject = subject_by_name.get(subject_name);
+    if (subject === undefined) {
+        throw new Error(`Unknown subject: ${subject_name}`);
+    }
 
     return {
-        identifier,
+        identifier: {
+            subject,
+            code: code,
+            section: section,
+            crn: null,
+        },
         date: {
             start: start_date,
             end: end_date,
diff --git a/package.json b/package.json
index 1b4d0c9..94a947b 100644
--- a/package.json
+++ b/package.json
@@ -17,6 +17,7 @@
     "@types/node": "^20.4.9",
     "cheerio": "^1.0.0-rc.12",
     "date-fns": "^2.30.0",
+    "html-entities": "^2.4.0",
     "ics": "^3.2.0",
     "node-fetch": "^3.3.2",
     "ts-node": "^10.9.1",
diff --git a/yarn.lock b/yarn.lock
index 52dc023..66ba9ac 100644
--- a/yarn.lock
+++ b/yarn.lock
@@ -192,6 +192,11 @@ formdata-polyfill@^4.0.10:
   dependencies:
     fetch-blob "^3.1.2"
 
+html-entities@^2.4.0:
+  version "2.4.0"
+  resolved "https://registry.yarnpkg.com/html-entities/-/html-entities-2.4.0.tgz#edd0cee70402584c8c76cc2c0556db09d1f45061"
+  integrity sha512-igBTJcNNNhvZFRtm8uA6xMY6xYleeDwn3PeBCkDz7tHttv4F2hsDI2aPgNERWzvRcNYHNT3ymRaQzllmXj4YsQ==
+
 htmlparser2@^8.0.1:
   version "8.0.2"
   resolved "https://registry.yarnpkg.com/htmlparser2/-/htmlparser2-8.0.2.tgz#f002151705b383e62433b5cf466f5b716edaec21"