Atleast in most common use, markdown parsers aren't that smart. They don't do much lookahead, they parse text in a stream with little buffering.

The most complex behaviour is tables, and even there you can easily spot where the parsers simply go character by character along the line with no difference other than being in "table mode" (or for code fences, in "code mode")