Checking suitable characters for a URL

[Amnesiac]

In Accelerated C++ section 6.1 pg 107 (for those who have it to hand) there is a function that returns true when passed a character that cannot be part of a URL.

The specific part I'm interested in is the following:

bool not_url_char(char c)
{
     static const string url_ch = "~;/?:@=&$-_.+!*'(),";

     return !(isalnum(c) ||
              find(url_ch.begin(), url_ch.end(), c) != url_ch.end());
}

So, as I understand it, the function is passed a character, and on line 5 returns false if the character is an alphanumeric character, or any of the characters in the range (url_ch.begin(), url_ch.end()]

But after the call to find() on line 6, what is the use of != url_ch.end()? I don't understand why this is in the code and what its purpose is; I can't seem to relate it to anything in the function. It seems to be saying "do something if something != url_ch.end() but I don't know what either of the somethings are...I hope that makes sense.

[modoran]

The entire expression is this:
find(url_ch.begin(), url_ch.end(), c) != url_ch.end()

Look at std::strings's find method and you will understand for yourself. (I suppose you know wahat iallnum() and OR operator do )

[Veltas]

It seems find returns the first iterator it finds to the given character, c.

I think the idea was that if the character isn't any of the characters in url_ch then it would end up returning the last iterator. != is the 'if not equals' operator, so returns true upon find not returning the end iterator, AKA returns true if any of the url_ch characters are found to be equal to c.

EDIT: And it returns in the range [url_ch.begin(), url_ch.end()]. It's possible for the first character to be the 'match' character.

[Lowest0ne]

~local static const? How useful.

This whole function is odd, just do the whole string at once. Not sure what isalnum() does (especially because this is just one character).

bool not_url(const string& str)
{
     return (str.find_first_of("~;/?:@=&$-_.+!*'(),") != string::npos);
}

[Veltas]

I'm not sure I like the double-negatives. If isalnum is necessary, maybe this is more appropriate:

bool not_url_char(char c)
{
     static const string url_ch = "~;/?:@=&$-_.+!*'(),";

     return (!isalnum(c)) && (find(url_ch.begin(), url_ch.end(), c) == url_ch.end());
}