A script to run in a web browser's Javascript console to normalize the HTML in old HTML documentation from help.adobe.com. The resulting simplified HTML content can be converted more easily into other formats, like Markdown, using tools like pandoc. The help.adobe.com content is Creative Commons licensed.

normalize-help-adobe-com-html.js

document.body.querySelector("#ahpod")?.remove();
document.body.querySelector("#mboxScriptContainer")?.remove();
let h1 = document.querySelector("#content_wrapper h1:first-of-type");
let article = document.querySelector(
  "table#inner_content_table td:first-of-type"
);
if (!article) {
  // some pages seem to have a different table id, for some reason
  article = document.querySelector("table#page_content_table td:first-of-type");
}
article.querySelector("div:first-of-type")?.remove();
article.querySelector("#chcPromo")?.remove();
article.querySelectorAll("script")?.forEach((element) => element.remove());
article.querySelector("#userprefs")?.remove();
// article.querySelector("#related")?.remove();
article.querySelector("#footer")?.remove();
article.querySelector("#minitoc")?.remove();
article
  .querySelectorAll('a[name^="WS"')
  ?.forEach((element) => element.remove());
article
  .querySelectorAll('a[href^="#top"]')
  ?.forEach((element) => element.remove());
article
  .querySelectorAll('*[width^="NaN%"]')
  ?.forEach((element) => element.removeAttribute("width"));
article
  .querySelectorAll("a")
  ?.forEach((element) => element.removeAttribute("target"));
document.body.querySelectorAll("*")?.forEach((element) => {
  element.removeAttribute("class");
  element.removeAttribute("id");
  element.removeAttribute("style");
  element.removeAttribute("valign");
  element.removeAttribute("headers");
  element.removeAttribute("border");
  element.removeAttribute("cellpadding");
  element.removeAttribute("cellspacing");
  element.removeAttribute("align");
  element.removeAttribute("valign");
  element.removeAttribute("xmlns:adobe");
  element.removeAttribute("xmlns:fn");
  element.removeAttribute("xmlns:fo");
  element.removeAttribute("xmlns:xs");
});
article
  .querySelectorAll("a")
  ?.forEach((element) => element.removeAttribute("onclick"));
document.body.innerHTML = "";
document.body.appendChild(h1);
document.body.appendChild(article);
document.body.innerHTML = document.body.innerHTML
  .replaceAll(/( |\n)<samp>\s+/g, "$1<samp>") //no whitespace after <samp>
  .replaceAll(/\s+<\/samp>/g, "</samp>") // no whitespace before </samp>
  .replaceAll(/(<\/\w+>)\s+(\.|,|:|\))/g, "$1$2") // no whitespace between end tag and punctuation
  .replaceAll(/\s+(<\/a>)/g, "$1"); // no whitespace before </a>

Here are some additional manual tasks that aren't easy to automate from the web browser console, or they may apply only after Markdown conversion with Pandoc.

• An extra space may appear between Markdown formatting and punctuation.

  (\*\*|_)( | \n)[.,!:;\)]

• An empty comment may appear between lines: <!-- -->

• Hyperlinks may have a strange formatting where they duplicate the URL in quotes.

  \]\([\w\-\/\.:]+ "[\w\-\/\.:]+"\)

• Original relative URLs start with WS. They need to be replaced.

  \]\(WS\w+
<a href="WS

• The original formatting may include curly quotes that should be replaced with straight quotes: " and '

  • “
  • ”
  • ‘
  • ’

• Check that code examples are formatted correctly.

  • ^\s+(\{|\}) (curly brace after whitespace at beginning of line)
  • ^\s+(<\w+) (open HTML/XML tag after whitespace at beginning of line)

• Replace goo.gl links with canonical URLs.

• Replace adobe.com/go/ links with canonical URLs:

  • http://www.adobe.com/go/learn_flcs5_as3lr_en => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • https://help.adobe.com/en_US/AS3LCR/Flash_10.0/ => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • http://www.adobe.com/go/learn_flex45_platformref_en => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • http://www.adobe.com/go/learn_flex4_apiref_en => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • https://help.adobe.com/en_US/AS3LCR/Flex_4.0/ => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • http://help.adobe.com:80/en_US/Flex/4.0/langref/ => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • http://livedocs.adobe.com/flex/201/langref/ => https://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/index.html
  • http://www.adobe.com/go/learn_cs5_as2lr_en -> https://web.archive.org/web/20120114132936/http://help.adobe.com/en_US/FlashPlatform/reference/actionscript/2/help.html?content=Part2_AS2_LangRef_1.html
  • http://www.adobe.com/go/flash_devcenter: https://web.archive.org/web/20120504010203/http://www.adobe.com/devnet/flash.html
  • http://www.adobe.com/go/learn_fl_samples: https://web.archive.org/web/20120303062950/http://www.adobe.com/devnet/flash/samples.html

• Replace dead external links with WayBack Machine archive links.

• Replace http: URLs with https:, if possible.

• Search for helpexamples.com URLs in example code and download the content (possibly from WayBack Machine).

• Nice to have: Replace fake URLs like [yourDomain].com with example.com.

• Nice to have: Replace HTML tables with GFM table syntax. (Sometimes, pandoc doesn't convert them from HTML to Markdown)

• Nice to have: convert code samples from simple indented code blocks that are not syntax highlighted to as3 and mxml code fences.

• Replace <span class="dfn"> and <span class="kbd"> with italic and code Markdown.