Fix(extractor): Correctly format LaTeX (KaTeX) equations in export (#192

)

* fix: format KaTeX in ChatGPT export (formulas between $ or $$)

* fix(rules): remove KaTeX duplication in ChatGPT export

* fix(rules): trim() KaTeX equations

* fix: format KaTeX in Perplexity export

* fix: format KaTeX in Phind export

* docs: update notes

* chore: version 3.4.0

* docs: update notes

package.json

@@ -1,6 +1,6 @@
 {
   "name": "save-my-chatbot",
-  "version": "3.2.6",
+  "version": "3.4.0",
   "license": "RMD-C v1.0 License",
   "author": "Hugo COLLIN",
   "homepage": "https://save.hugocollin.com",

public/files/updateNotes.md

@@ -1,8 +1,12 @@
 # Update notes
-# 3.3.0
-🐞 Facilitating bug reports
-When an error occurs during the export process, a modal will appear inviting you report the bug. 
-Plus, an option is available in the icon context menu.
+# 3.4.0
+🔢 Export math equations
+LaTeX equations from Perplexity, ChatGPT and Phind are now correctly formatted! 
+Math lovers, enjoy!
+
+🐞 Facilitating bug reports (or feature requests 🚀)
+When an error occurs during the export process, a modal will invite you to report the bug. 
+Plus, a new option is available in the icon context menu.
 
 # 3.2.6
 🧩 Some bug fixes

src/scripts/content/export/extractor/domains/ChatGPT.json

@@ -21,6 +21,10 @@
       "formatTables": {
         "filter": "filter_formatTables",
         "replacement": "replacement_formatTables"
+      },
+      "formatKatex": {
+        "filter": "filter_formatKatex",
+        "replacement": "replacement_formatKatex"
       }
     }
   },

src/scripts/content/export/extractor/domains/Perplexity.json

@@ -20,6 +20,10 @@
       "formatTables": {
         "filter": "filter_formatTables",
         "replacement": "replacement_formatTables"
+      },
+      "formatKatex": {
+        "filter": "filter_formatKatex",
+        "replacement": "replacement_formatKatex"
       }
     }
   },

src/scripts/content/export/extractor/domains/PerplexityPages.json

@@ -20,6 +20,10 @@
       "formatTables": {
         "filter": "filter_formatTables",
         "replacement": "replacement_formatTables"
+      },
+      "formatKatex": {
+        "filter": "filter_formatKatex",
+        "replacement": "replacement_formatKatex"
       }
     }
   },

src/scripts/content/export/extractor/domains/PhindChat.json

@@ -25,6 +25,10 @@
       "formatTables": {
         "filter": "filter_formatTables",
         "replacement": "replacement_formatTables"
+      },
+      "formatKatex": {
+        "filter": "filter_formatKatex",
+        "replacement": "replacement_formatKatex"
       }
     }
   },

src/scripts/content/export/extractor/domains/PhindSearch.json

@@ -21,6 +21,10 @@
       "formatTables": {
         "filter": "filter_formatTables",
         "replacement": "replacement_formatTables"
+      },
+      "formatKatex": {
+        "filter": "filter_formatKatex",
+        "replacement": "replacement_formatKatex"
       }
     }
   },

src/scripts/content/export/extractor/rules/rules.js

@@ -153,6 +153,32 @@ export function replacement_formatBeforeLinks(content, node) {
   return content.trim();
 }
 
+/*
+  --- Format KateX ---
+ */
+
+export function filter_formatKatex(node) {
+  return node.nodeName === 'SPAN' && node.classList.contains('katex');
+}
+
+export function replacement_formatKatex(content, node) {
+  // Cannot just select "math annotation" as it's missing in the node children (but rendered in the node.textContent !?!?)
+  const mrow = node.querySelector("math mrow")?.textContent ?? "";
+  const htmlMath = node.querySelector(".katex-html")?.textContent ?? "";
+
+  // Remove mrow at start and htmlMath at end of node.textContent
+  const mathml = node.textContent
+    ?.replace(mrow, "")
+    .replace(htmlMath, "")
+    .trim()
+    ?? node.textContent;
+
+  if (node.parentNode?.classList.contains('katex-display')) {
+    return '\n$$' + mathml + '$$\n';
+  }
+  return '$' + mathml + '$';
+}
+
 /*
   --- Phind rules ---
  */