I want to remove all http: and https: in the HTML files but exclude placeholder="http: and placeholder="https:. I have tried the following example but every http: and https: will be removed:
/(?!placeholder=")(http:|https:)/
You need to replace the lookahead with a lookbehind. Besides, you may reduce the alternation to a mere https?: pattern, where s? means 1 or 0 s:
'/(?<!placeholder=")https?:/'
^ ^^
If you want to make sure the placeholder is matched as a whole word, add a word boundary:
'/(?<!\bplaceholder=")https?:/'
^^
If there must be a whitespace before placeholder, replace \b with \s.
Details
(?<!\bplaceholder=") - a location inside a string that is immediately preceded with a whole word placeholder and then ="http - a http substrings? - an optional s: - a colon.