xmode: fix handling of HASH_CHAR and always rules

[factor.git] / basis / xmode / loader / syntax / syntax.factor

diff --git a/basis/xmode/loader/syntax/syntax.factor b/basis/xmode/loader/syntax/syntax.factor
index 0e7293da976f54d16fe4222a658580a736cbe570..a26fa6be716a838462187d4f802edec2f64bd01e 100644 (file)
--- a/basis/xmode/loader/syntax/syntax.factor
+++ b/basis/xmode/loader/syntax/syntax.factor
@@ -1,19 +1,20 @@
 ! Copyright (C) 2007, 2009 Slava Pestov.
 ! See http://factorcode.org/license.txt for BSD license.
-USING: accessors xmode.tokens xmode.rules xmode.keyword-map
-xml.data xml.traversal xml assocs kernel combinators sequences
-math.parser namespaces make parser lexer xmode.utilities
-parser-combinators.regexp io.files splitting arrays ;
+USING: accessors arrays assocs combinators kernel lexer make
+namespaces parser sequences splitting xml.data xml.syntax
+xml.syntax.private xml.traversal xml.traversal.private
+xmode.rules xmode.tokens xmode.utilities ;
 IN: xmode.loader.syntax
 
 ! Rule tag parsing utilities
 : (parse-rule-tag) ( rule-set tag specs class -- )
     new swap init-from-tag swap add-rule ; inline
 
-: RULE:
-    scan scan-word
-    parse-definition { } make
-    swap [ (parse-rule-tag) ] 2curry (TAG:) ; parsing
+SYNTAX: RULE:
+    scan-token scan-word scan-word [
+        [ parse-definition call( -- ) ] { } make
+        swap [ (parse-rule-tag) ] 2curry
+    ] dip swap define-tag ;
 
 ! Attribute utilities
 : string>boolean ( string -- ? ) "TRUE" = ;
@@ -27,13 +28,15 @@ IN: xmode.loader.syntax
 
 : string>rule-set-name ( string -- name ) "MAIN" or ;
 
+: cdata>string ( tag -- string )
+    children>> [ dup cdata? [ text>> ] when ] map (children>string) ;
+
 ! PROP, PROPS
 : parse-prop-tag ( tag -- key value )
     [ "NAME" attr ] [ "VALUE" attr ] bi ;
 
 : parse-props-tag ( tag -- assoc )
-    child-tags
-    [ parse-prop-tag ] H{ } map>assoc ;
+    children-tags [ parse-prop-tag ] H{ } map>assoc ;
 
 : position-attrs ( tag -- at-line-start? at-whitespace-end? at-word-start? )
     ! XXX Wrong logic!
@@ -41,33 +44,43 @@ IN: xmode.loader.syntax
     [ attr string>boolean ] with map first3 ;
 
 : parse-literal-matcher ( tag -- matcher )
-    dup children>string
+    dup cdata>string
     rule-set get ignore-case?>> <string-matcher>
     swap position-attrs <matcher> ;
 
 : parse-regexp-matcher ( tag -- matcher )
-    dup children>string rule-set get ignore-case?>> <regexp>
+    dup cdata>string
+    rule-set get ignore-case?>> <?insensitive-regexp>
     swap position-attrs <matcher> ;
 
 : shared-tag-attrs ( -- )
-    { "TYPE" string>token (>>body-token) } , ; inline
+    { "TYPE" string>token body-token<< } , ; inline
 
 : parse-delegate ( string -- pair )
     "::" split1 [ rule-set get swap ] unless* 2array ;
 
 : delegate-attr ( -- )
-    { "DELEGATE" f (>>delegate) } , ;
+    { "DELEGATE" f delegate<< } , ;
+
+! XXX: check HASH_CHAR for full prefix, not just first character
+
+: char<< ( value object -- )
+    '[ 1 head _ chars<< ] unless-empty ;
 
 : regexp-attr ( -- )
-    { "HASH_CHAR" f (>>chars) } , ;
+    { "HASH_CHARS" f chars<< } ,
+    { "HASH_CHAR" f char<< } , ;
 
 : match-type-attr ( -- )
-    { "MATCH_TYPE" string>match-type (>>match-token) } , ;
+    { "MATCH_TYPE" string>match-type match-token<< } , ;
+
+: string>escape ( str -- escape/f )
+    [ f ] [ <escape-rule> ] if-empty ;
 
 : span-attrs ( -- )
-    { "NO_LINE_BREAK" string>boolean (>>no-line-break?) } ,
-    { "NO_WORD_BREAK" string>boolean (>>no-word-break?) } ,
-    { "NO_ESCAPE" string>boolean (>>no-escape?) } , ;
+    { "NO_LINE_BREAK" string>boolean no-line-break?<< } ,
+    { "NO_WORD_BREAK" string>boolean no-word-break?<< } ,
+    { "ESCAPE" string>escape escape-rule<< } , ;
 
 : literal-start ( -- )
     [ parse-literal-matcher >>start drop ] , ;
@@ -78,28 +91,34 @@ IN: xmode.loader.syntax
 : literal-end ( -- )
     [ parse-literal-matcher >>end drop ] , ;
 
-! SPAN's children
-<TAGS: parse-begin/end-tag ( rule tag -- )
+TAGS: parse-begin/end-tag ( rule tag -- )
 
-TAG: BEGIN
-    ! XXX
+TAG: BEGIN parse-begin/end-tag
     parse-literal-matcher >>start drop ;
 
-TAG: END
-    ! XXX
+TAG: END parse-begin/end-tag
     parse-literal-matcher >>end drop ;
 
-TAGS>
-
 : parse-begin/end-tags ( -- )
-    [
-        ! XXX: handle position attrs on span tag itself
-        child-tags [ parse-begin/end-tag ] with each
-    ] , ;
+    [ children-tags [ parse-begin/end-tag ] with each ] , ;
+
+TAGS: parse-regexp-begin/end-tag ( rule tag -- )
+
+TAG: BEGIN parse-regexp-begin/end-tag
+    parse-regexp-matcher >>start drop ;
+
+! XXX: END AT_WHITESPACE_END="TRUE"?
+
+TAG: END parse-regexp-begin/end-tag
+    dup "REGEXP" attr string>boolean
+    [ parse-regexp-matcher ] [ parse-literal-matcher ] if >>end drop ;
+
+: parse-regexp-begin/end-tags ( -- )
+    [ children-tags [ parse-regexp-begin/end-tag ] with each ] , ;
 
 : init-span-tag ( -- ) [ drop init-span ] , ;
 
 : init-eol-span-tag ( -- ) [ drop init-eol-span ] , ;
 
 : parse-keyword-tag ( tag keyword-map -- )
-    [ dup main>> string>token swap children>string ] dip set-at ;
+    [ dup main>> string>token swap cdata>string ] dip set-at ;