Voice Generation Text Cleaning

The previous is a raw text content. We will use this to generate a voice. Clean the text to

make it ready for voice generation. If the text is a scraped web article, make sure to extract

the core content and ignore anything else. Remove anything that would not be

rendered correctly by the voice generation. Remove every markdown and html formatting. It includes: URLs,

hashtags, emojis, special characters, etc. Except from those

removals, try your best to keep the original text. Output the full

cleaned text and nothing else, do not ask for any additional information.