RegEx Issue in parse()

[gabrielodom]

Consider the following example:

sentence <- 'I had "a bunch" of coke.'

The regular expression fails to remove these quotation marks before matching against stop words. Thus, "a" is still included in the output of parse().

[gabrielodom]

Also, the function discards all output if "Bup/Nx" is included in the string.

[gabrielodom]

Concerning smart quotes: https://www.cl.cam.ac.uk/~mgk25/ucs/quotes.html#:~:text=If%20you%20can%20use%20Unicode,'%20or%20%E2%80%9Cquote%E2%80%9D%20)

U+201C, U+201D

[gabrielodom]

Another issue:
If I call lookup(m-CAT) (for mephedrone), I get

  original_word     class            category synonym
1         m-cat stimulant synthetic cathinone   m-cat

However, when we put this word in a sentence:

> lookup(parse("try some m-CAT my friend"))
  original_word             class category synonym
1        friend narcotic (opioid) fentanyl  friend

The parse() function removes m-CAT because of the hyphen. Additionally, even if parse() just removed the hyphen and left mCAT or mcat, lookup(mcat) returns NAs. In fact, there are 129 slang terms which contain - in their strings. Our current framework for parse() would fail to match these drugs. Code to find the 129 drugs mentioned:

DOPE::lookup_df %>% filter(str_detect(synonym, pattern = "-"))