Replacing Mandaic characters with Hebrew characters in specific HTML tags break links

Why does the given JavaScript code that replaces Mandaic characters with Hebrew characters in a webpage break all the links in the page? The code is as follows:

function replaceMandaicCharacters() {
  const mandaicLetters = "u{0840}u{0841}u{0842}u{0843}u{0844}u{0845}u{0846}u{0847}u{0848}u{0849}u{084A}u{084B}u{084C}u{084D}u{084E}u{084F}u{0850}u{0851}u{0852}u{0853}u{0854}u{0855}";
  const hebrewLetters = "אבגדחוזהטיכלמנסעפצקרשת";
  const targetTags = ["b", "td"];

  // Replace Mandaic characters in specific HTML tags
  for (const tag of targetTags) {
    const elements = document.getElementsByTagName(tag);
    for (const element of elements) {
      const originalText = element.textContent;
      const regex = new RegExp(`[${mandaicLetters}]`, "g");
      const newText = originalText.replace(regex, (match) => {
        const index = mandaicLetters.indexOf(match);
        return hebrewLetters.charAt(index);
      });
      if (newText !== originalText) {
        element.textContent = newText;
      }
    }
  }

  // Replace Mandaic characters in all text nodes on the webpage
  const textNodes = document.evaluate(
    "//text()",
    document,
    null,
    XPathResult.UNORDERED_NODE_SNAPSHOT_TYPE,
    null
  );
  for (let i = 0; i < textNodes.snapshotLength; i++) {
    const textNode = textNodes.snapshotItem(i);
    const originalText = textNode.textContent;
    const regex = new RegExp(`[${mandaicLetters}]`, "g");
    const newText = originalText.replace(regex, (match) => {
      const index = mandaicLetters.indexOf(match);
      return hebrewLetters.charAt(index);
    });
    if (newText !== originalText) {
      textNode.textContent = newText;
    }
  }
}

// Call replaceMandaicCharacters() whenever the DOM is updated
const observer = new MutationObserver(replaceMandaicCharacters);
observer.observe(document, { subtree: true, childList: true });

When I apply this code to a webpage, it replaces the Mandaic characters with Hebrew characters, but it also cancels links in the page, at least on some pages like this. How can I fix this issue and prevent links from being canceled?

Some suggestions, such as replacing only texts and not hrefs, did not work as expected.