<zacts>mark_weaver: what do you think of the Oracle versus Google case? <zacts>feel free to /msg me if you like ***slucxx is now known as slucx
***slucx is now known as s-botx
***s-botx is now known as sbot
***sbot is now known as sgbotx
***sgbotx is now known as gsbot
***gsbot is now known as sbot
***sbot is now known as slucx
<davexunit>is it possible to use 'match' to match based on a predicate function *and* bind the result if it isn't falsy? <davexunit>in my case, I have am using a regexp to extract some text from a string. If the string doesn't match, #f is returned, otherwise the relevant substring is returned. <davexunit>I could do something like: (match "foo" ((and (? matches? _) (= extract-stuff s)) s)) <davexunit>but I don't like it because it requires repeating the call to regexp-exec <mark_weaver>davexunit: I think you can do something like (match "foo" ((= extract-stuff (? identity s)) ...)) <davexunit>using pattern matching for the URI routing part of 'guix publish' <mark_weaver>sometimes I think we should have an alias 'true?' for 'identity' to better readability in cases like this. <davexunit>I have looked for a procedure named 'true?' on more than one occasion. :) *jao would find such an alias very confusing