core
/
oga
6
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Updated parsing of doctypes. 

The resulting nodes now separate the type, public and system IDs in to separate
string values.
This commit is contained in:
Yorick Peterse 2014-03-11 22:08:21 +01:00
parent c07edc767b
commit c9592856f0
2 changed files with 33 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
lib/oga/parser.y
View File

@ -1,8 +1,8 @@
class Oga::Parser class Oga::Parser
token T_SPACE T_NEWLINE T_SMALLER T_GREATER T_SLASH token T_NEWLINE T_SPACE
token T_DQUOTE T_SQUOTE T_DASH T_RBRACKET T_LBRACKET token T_STRING
token T_COLON T_BANG T_EQUALS T_TEXT T_DOCTYPE token T_DOCTYPE_START T_DOCTYPE_END T_DOCTYPE_TYPE
options no_result_var options no_result_var
@ -18,38 +18,32 @@ rule
; ;
expression expression
: tag : doctype
| doctype
; ;
# Doctypes # Doctypes
doctype doctype
: T_DOCTYPE { s(:doctype, val[0]) } # <!DOCTYPE html>
; : T_DOCTYPE_START T_DOCTYPE_END { s(:doctype) }
# Generic HTML tags # <!DOCTYPE html PUBLIC>
| T_DOCTYPE_START T_DOCTYPE_TYPE T_DOCTYPE_END
tag_start
# <p>
: T_SMALLER T_TEXT T_GREATER { val[1] }
;
tag_end
# </p>
: T_SMALLER T_SLASH T_TEXT T_GREATER
;
tag
# <p>foo</p>
: tag_start tag_body tag_end
{ {
s(:element, val[0], val[1]) s(:doctype, val[1])
} }
;
tag_body # <!DOCTYPE html PUBLIC "foo">
: T_TEXT | T_DOCTYPE_START T_DOCTYPE_TYPE T_STRING T_DOCTYPE_END
{
s(:doctype, val[1], val[2])
}
# <!DOCTYPE html PUBLIC "foo" "bar">
| T_DOCTYPE_START T_DOCTYPE_TYPE T_STRING T_STRING T_DOCTYPE_END
{
s(:doctype, val[1], val[2], val[3])
}
; ;
whitespaces whitespaces

18
spec/oga/parser/doctype_spec.rb
View File

@ -6,17 +6,24 @@ describe Oga::Parser do
parse_html('<!DOCTYPE html>').should == s(:document, s(:doctype)) parse_html('<!DOCTYPE html>').should == s(:document, s(:doctype))
end end
example 'parse a doctype with a public ID' do example 'parse a doctype with the doctype type' do
parse_html('<!DOCTYPE html "foo">').should == s( parse_html('<!DOCTYPE html PUBLIC>').should == s(
:document, :document,
s(:doctype, 'foo') s(:doctype, 'PUBLIC')
)
end
example 'parse a doctype with a public ID' do
parse_html('<!DOCTYPE html PUBLIC "foo">').should == s(
:document,
s(:doctype, 'PUBLIC', 'foo')
) )
end end
example 'parse a doctype with a public and private ID' do example 'parse a doctype with a public and private ID' do
parse_html('<!DOCTYPE html "foo" "bar">').should == s( parse_html('<!DOCTYPE html PUBLIC "foo" "bar">').should == s(
:document, :document,
s(:doctype, 'foo', 'bar') s(:doctype, 'PUBLIC', 'foo', 'bar')
) )
end end
@ -28,6 +35,7 @@ describe Oga::Parser do
:document, :document,
s( s(
:doctype, :doctype,
'PUBLIC',
'-//W3C//DTD HTML 4.01//EN', '-//W3C//DTD HTML 4.01//EN',
'http://www.w3.org/TR/html4/strict.dtd' 'http://www.w3.org/TR/html4/strict.dtd'
) )