Remove special characters from string in Red language

Question

I want to remove all characters in a string except:

• - or _ or .
• A thru Z
• a thru z
• 0 to 9
• space

On linux command line, using sed I would do this:

$ echo "testing-#$% yes.no" | sed 's/[^-_.a-zA-Z0-9 ]//g'

Output:

testing- yes.no

How can I achieve the same effect in Red language with PARSE? I looked at:

• http://www.rebol.com/docs/core23/rebolcore-15.html#section-1
• http://rebol-land.blogspot.in/2013/03/rebols-answer-to-regex-parse-and-rebol.html
• http://ross-gill.com/page/Beyond_Regular_Expressions

However, I could not codify it. I tried:

>> parse "mystring%^&" [#a - #z #A - #Z #0 - #9]
== false
>> parse "mystring%^&" [#a-#z#A-#Z#0-#9]        
== false

Answer 1

First note the difference between ISSUE! and CHAR!

#a #b #c  ; issues
#"a" #"b" #"c"   ; chars

You can then establish a character set (BITSET! type) either for the characters you want to keep or those you wish to discard. We'll do the former here:

good-chars: charset [#"a" - #"z" #"A" - #"Z" #"0" - #"9"]

Now that we have that, we can approach this in some different ways:

Parse

A fairly basic parse loop—skips any good-chars and removes anything else.

parse "mystring%^&" [any [some good-chars | remove skip]]

Remove-Each

Hopefully self-explanatory:

remove-each char "mystring%^&" [not find good-chars char]

Answer 2

First, characters must be in quotes, #a is issue!, char! is #"a". You've got the specification right, but you must pass it to charset function, to make a bitset! form it.

Then you can parse your string, keeping valid characters and skiping invalid:

>> chars: charset [#"a" - #"z" #"A" - #"Z" #"0" - #"9"]
== make bitset! #{000000000000FFC07FFFFFE07FFFFFE0}
>> rejoin parse "mystring%^&asdf" [collect some [keep chars | skip]]
== "mystringasdf"