Re: [basex-talk] Different interpretation of regex in eXist, Saxon and BaseX

Hi Andreas -

wow, that is a pretty nice regex :). I'm not nearly caffeinated enough right now to pick it apart, so I'm only able to ask a question - not provide any answers or help. Unless I'm reading the spec and Walmsley's coverage wrong, isn't the '?' a reluctant quantifier - given two choices it will always match the shorter choice? Or does the hash/octothorp give extra significance to the '?' quantifier?

In any event, thank you for the neat brain teaser! Best, Bridger

On Tue, Aug 7, 2018 at 3:38 PM Andreas Mixich mixich.andreas@gmail.com wrote:

Hi

[rfc3986](https://tools.ietf.org/html/rfc3986#appendix-B) defines a nice regular expression, which groups any URI, including URN, by URI component.

Interesting about this regex is the use of the '?' quantifier which makes every preceding group/component optional, thus matching either an URI or any other(!) string, since anything, that does not match one of the special groups, goes into a catch-all group (no.5), which keeps either the path or the full, arbitrary string. This is neglectable, since the input to this regex is guaranteed to be of the right type (a/@href/string()).

Here is the relevant part from the RFC.

Appendix B

^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(?([^#]*))?(#(.*))? 12 3 4 5 6 7 8 9

 The numbers in the second line above are only to assist
 readability; they indicate the reference points for each
 subexpression (i.e., each paired parenthesis).  We refer to the
 value matched for subexpression <n> as $<n>.  For example, matching
 the above expression to

    http://www.ics.uci.edu/pub/ietf/uri/#Related

 results in the following subexpression matches:

    $1 = http:
    $2 = http
    $3 = //www.ics.uci.edu
    $4 = www.ics.uci.edu
    $5 = /pub/ietf/uri/
    $6 = <undefined>
    $7 = <undefined>
    $8 = #Related
    $9 = Related

 where <undefined> indicates that the component is not present,
 as is the case for the query component in the above example.
 Therefore, we can determine the value of the five components as

    scheme    = $2
    authority = $4
    path      = $5
    query     = $7
    fragment  = $9

 Going in the opposite direction, we can recreate a URI reference
 from its components by using the algorithm of Section 5.3.

I tested this regex with Saxon, eXist and BaseX. eXist successfully parsed all the test-cases, I threw at it, into the right groups, Saxon and BaseX did not. The failure is:

[FORX0003] Pattern matches empty string..

And that got me baffled, since all three processors use Java underneath and since the definition of the '?' quantifier, when used like this, seems to be:

Makes the preceding item optional. Greedy, so the optional item
is included in the match if possible.

Which means, that *if* any of the group's contents match, they should be included, rather than producing an empty string.

Why is it like that? And what can I do about it? I found no other URI parsing regex, that componentizes this way and would be compatible with XQuery.

See, attached, a test-case.

-- Goody Bye, Minden jót, Mit freundlichen Grüßen, Andreas Mixich