! Copyright (C) 2010 Doug Coleman.
! See http://factorcode.org/license.txt for BSD license.
-USING: html.parser.analyzer math tools.test ;
+USING: html.parser html.parser.analyzer math tools.test ;
IN: html.parser.analyzer.tests
[ 0 3 ]
[ 0 { 3 5 7 9 11 } [ odd? ] find-last-nth ]
[ undefined-find-nth? ] must-fail-with
+
+[ V{
+ T{ tag f text f "foo" f }
+}
+] [
+ "<html><head><title>foo</title></head></html>" parse-html
+ "title" find-between-first
+] unit-test
+
+[ V{
+ T{ tag f "p" H{ } f f }
+ T{ tag f text f "para" f }
+ T{ tag f "p" H{ } f t }
+}
+] [
+ "<body><div><p>para</p></div></body>" parse-html "div" find-between-first
+] unit-test
+
+[ V{
+ T{ tag f "div" H{ { "class" "foo" } } f f }
+ T{ tag f "p" H{ } f f }
+ T{ tag f text f "para" f }
+ T{ tag f "p" H{ } f t }
+ T{ tag f "div" H{ } f t }
+}
+] [
+ "<body><div class=\"foo\"><p>para</p></div></body>" parse-html
+ "foo" find-by-class-between
+] unit-test
+
+[ V{
+ T{ tag f "div" H{ { "class" "foo" } } f f }
+ T{ tag f "div" H{ } f f }
+ T{ tag f "p" H{ } f f }
+ T{ tag f text f "para" f }
+ T{ tag f "p" H{ } f t }
+ T{ tag f "div" H{ } f t }
+ T{ tag f "div" H{ } f t }
+}
+] [
+ "<body><div class=\"foo\"><div><p>para</p></div></div></body>" parse-html
+ "foo" find-by-class-between
+] unit-test
! Copyright (C) 2008 Doug Coleman.
! See http://factorcode.org/license.txt for BSD license.
USING: accessors assocs combinators combinators.short-circuit
-fry html.parser http.client io kernel locals math sequences
-sets splitting unicode.case unicode.categories urls
+fry html.parser http.client io kernel locals math math.statistics
+sequences sets splitting unicode.case unicode.categories urls
urls.encoding shuffle ;
IN: html.parser.analyzer
: find-first-name ( vector string -- i/f tag/f )
>lower '[ name>> _ = ] find ; inline
-: find-matching-close ( vector string -- i/f tag/f )
+! Takes a sequence and a quotation expected to return -1 if the
+! element decrements the stack, 0 if it doesnt affect it and 1 if it
+! increments it. Then finds the matching element where the stack is
+! empty.
+: stack-find ( seq quot -- i/f )
+ map cum-sum [ 0 = ] find drop ; inline
+
+! Produces a function which returns 1 if the input item is an opening
+! tag element with the specified name, -1 if it is a closing tag of
+! the same name and 0 otherwise.
+: tag-classifier ( string -- quot )
>lower
- '[ [ name>> _ = ] [ closing?>> ] bi and ] find ; inline
+ '[ dup name>> _ = [ closing?>> [ -1 ] [ 1 ] if ] [ drop 0 ] if ] ; inline
: find-between* ( vector i/f tag/f -- vector )
over integer? [
[ tail-slice ] [ name>> ] bi*
- dupd find-matching-close drop [ 1 + ] [ 1 ] if*
+ dupd tag-classifier stack-find [ 1 + ] [ 1 ] if*
head
] [
3drop V{ } clone