rewriterule [B]后向标记

Question

I am trying to have a snippet in php to understand how B flag works.

My htaccess file location is Applications/AMPPS/www/h or localhost/h having following code

RewriteRule search/(.*)$ /search.php?$1 [R,QSA,B]

when I hit url localhost/h/search/x & y/z with B Flag I got this: localhost/search.php?x+%2526+y%252fz and without B flag I got this: http://localhost/search.php?x%20&%20y/z

Now if I use php's urlencode for string x & y/z that gives me x+%26+y%2Fz which is replacing space with + as B Flag does but whereas for & B flag gives output like %2526 , urlencode gives %26 , when I re-encode the encoded output again, becomes something like: x%2B%2526%2By%252Fz which yes, encodes %26 into %2526 but encode + so in both cases , it does not do the way, B does. I need to match it the way B does.

in addition, what possible characters (may be there can be a list ) where I can find out which non-alphanumeric characters can be encoded by B and how B encode them if I try to understand from php's prospective.

Answer 1

The [B] flag will encode all non-alphanumeric characters. So every character except the 36 (AZ 0-9) or 62 (AZ az 0-9) alphanumeric characters will be escaped.

So first, why do you need the [B] flag ? As you said, the string x & y/z in a URL would result into x%20&%20y/z which is invalid since & is one of the reserved characters in URL's even after percent encoding .

So x%20&%20y/z would only give you x in your PHP.

So now you'll need the flag. Let's analyse what it does with your example string:

x & y/z
x%20&%20y/z       // encode all spaces (%20)
x%20%26%20y%2Fz   // encode the & (%26) and / (%2F)

And this will result in x & y/z in your PHP.

What you're probably confused about is, that you don't see the correct encoding in your browser.

Let's analyse why you get x+%2526+y%252fz using the [R] flag :

The & gets encoded into %26 If you would encode this again, the % would be encoded into %25 the result would be %2526 . But why do the both special (reserved) characters get encoded twice?

This is because the rewrite module does encode the special characters first and then encode everything by default.

If you want to suppress this, you can use the [NE] flag.

The behavior of getting + instead of %20 seems to be a side effect of using [B] and [R] since this won't happen if you only use [R]

To answer your question: the [B] flag does the same as PHP's rawurlencode .