Andy <>
Sat, 21 Dec 2019 01:52:26 -0800 (PST)

          From comp.compilers

Related articles
Fragments (Andy) (2019-12-21)
Re: Fragments (Kaz Kylheku) (2019-12-21)
Re: Fragments (Ben Hanson) (2019-12-22)
| List of all articles for this month |

From: Andy <>
Newsgroups: comp.compilers
Date: Sat, 21 Dec 2019 01:52:26 -0800 (PST)
Organization: Compilers Central
Injection-Info:; posting-host=""; logging-data="81191"; mail-complaints-to=""
Keywords: lex
Posted-Date: 21 Dec 2019 13:10:24 EST

In examples is usually used very small alphabet: 3 to 5 letters but in
lexical analysing is not only Ascii but many thousands of Unicode.
Many chars are grouped by the same action: for example digits->a
letter->b whitepsaces->c
We can use "fragments" [A-Za-z], [0-9] instead of alone letters.
Problem that fragments not always are disjoint: digits and all chars, letters and letter 'a', etc.

How to handle with not disjoint fragments? on input we get regular
expression in Posix standard and we want make DFA with a few

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.