Ignoring quoted strings in regular expressions

[nick]

I couldn't find a usenet group on regular expressions, so I'm posing
here. Sorry if this is the wrong place to post.

Anyway, I'm wondering if there's some way to specify in a regular
expression a way to ignore quoted strings. For example:

1. blah whatever blah
2. blah "whatever" blah
3. blah " whatever " blah

I want to match 1, but not 2 or 3. Essentially, I want to match if
whatever is anywhere in the string and unquoted.

At http://www.cs.sfu.ca/~cameron/REX.html I found the following:

([^]"'><]+|"[^"]*"|'[^']*')*>

"This expression scans through arbitrary content searching for the
closing ">" delimiter. Quoted strings are skipped and the scan may
terminate with failure if an erroneous "]" or "<" delimiter is
encountered."

I don't see where in the above RE the actual quoted string skipping is
being done.

Any help would be appreciated.

Thanks,

Nick

 

[Ben Morrow]

I couldn't find a usenet group on regular expressions, so I'm posing
here. Sorry if this is the wrong place to post.

Which language/regex package are you using? You would be best off in a
group specific to that. Unless, of course, this is a homework problem.

Anyway, I'm wondering if there's some way to specify in a regular
expression a way to ignore quoted strings. For example:

At http://www.cs.sfu.ca/~cameron/REX.html I found the following:

([^]"'><]+|"[^"]*"|'[^']*')*>

Bleech. Talk about line-noise .

Let's break it up. Stuff from # to end-of-line is my commentary.

( # start group
[^]"'><]+ # match as much stuff as possible that doesn't include
# ] " ' > or <.
| # or
" [^"]* " # a double-quoted string (" followed by as much not-" as
# you can followed by another ")
| # or
' [^']* ' # a single-quoted string (as above)
)* # keep doing all of the above as much as you can

# end by matching a >.

Ben