1 ! Copyright (C) 2005, 2006 Daniel Ehrenberg
2 ! See http://factorcode.org/license.txt for BSD license.
3 USING: accessors arrays ascii assocs combinators
4 combinators.short-circuit fry io.encodings io.encodings.iana
5 io.encodings.string io.encodings.utf16 io.encodings.utf8 kernel make
6 math math.parser namespaces sequences sets splitting state-parser
7 strings xml.char-classes xml.data xml.entities xml.errors ;
10 ! XML namespace processing: ns = namespace
12 ! A stack of hashtables
15 : attrs>ns ( attrs-alist -- hash )
16 ! this should check to make sure URIs are valid
19 swap dup space>> "xmlns" =
22 T{ name f "" "xmlns" f } names-match?
23 [ "" set ] [ drop ] if
26 ] { } make-assoc f like ;
29 dup space>> dup ns-stack get assoc-stack
30 [ nip ] [ nonexist-ns ] if* >>url drop ;
38 : init-ns-stack ( -- )
40 { "xml" "http://www.w3.org/XML/1998/namespace" }
41 { "xmlns" "http://www.w3.org/2000/xmlns" }
46 : tag-ns ( name attrs-alist -- name attrs )
48 [ dup add-ns ] dip dup [ drop add-ns ] assoc-each <attrs> ;
52 : version=1.0? ( -- ? )
53 prolog-data get version>> "1.0" = ;
55 ! version=1.0? is calculated once and passed around for efficiency
57 : assure-name ( str version=1.0? -- str )
60 [ rest-slice [ name-char? ] with all? ]
61 } 2&& [ "Malformed name" xml-string-error ] unless ;
63 : (parse-name) ( start -- str )
65 [ [ get-char name-char? not ] curry take-until append ]
68 : parse-name-starting ( start -- name )
69 (parse-name) get-char CHAR: : =
70 [ next "" (parse-name) ] [ "" swap ] if f <name> ;
72 : parse-name ( -- name )
73 "" parse-name-starting ;
77 : (parse-entity) ( string -- )
78 dup entities at [ , ] [
79 prolog-data get standalone>>
81 dup extra-entities get at
82 [ , ] [ no-entity ] ?if
87 next CHAR: ; take-char next
89 "x" ?head 16 10 ? base> ,
90 ] [ (parse-entity) ] if ;
92 : (parse-char) ( ch -- )
94 { [ dup not ] [ 2drop ] }
95 { [ 2dup = ] [ 2drop next ] }
96 { [ dup CHAR: & = ] [ drop parse-entity (parse-char) ] }
97 [ , next (parse-char) ]
100 : parse-char ( ch -- string )
101 [ (parse-char) ] "" make ;
103 : parse-quot ( ch -- string )
105 [ "XML file ends in a quote" xml-string-error ] unless ;
107 : parse-text ( -- string )
112 : start-tag ( -- name ? )
113 #! Outputs the name and whether this is a closing tag
114 get-char CHAR: / = dup [ next ] when
117 : parse-attr-value ( -- seq )
118 get-char dup "'\"" member? [
121 "Attribute lacks quote" xml-string-error
125 [ parse-name ] with-scope
126 pass-blank CHAR: = expect pass-blank
127 [ parse-attr-value ] with-scope
130 : (middle-tag) ( -- )
131 pass-blank version=1.0? get-char name-start?
132 [ parse-attr (middle-tag) ] when ;
134 : middle-tag ( -- attrs-alist )
135 ! f make will make a vector if it has any elements
136 [ (middle-tag) ] f make pass-blank ;
138 : end-tag ( name attrs-alist -- tag )
139 tag-ns pass-blank get-char CHAR: / =
140 [ pop-ns <contained> next ] [ <opener> ] if ;
142 : take-comment ( -- comment )
148 : take-cdata ( -- string )
149 "[CDATA[" expect-string "]]>" take-string ;
151 : take-element-decl ( -- element-decl )
152 pass-blank " " take-string pass-blank ">" take-string <element-decl> ;
154 : take-attlist-decl ( -- doctype-decl )
155 pass-blank " " take-string pass-blank ">" take-string <attlist-decl> ;
157 : take-until-one-of ( seps -- str sep )
158 '[ get-char _ member? ] take-until get-char ;
160 : only-blanks ( str -- )
161 [ blank? ] all? [ bad-doctype-decl ] unless ;
163 : take-system-literal ( -- str )
164 pass-blank get-char next {
165 { CHAR: ' [ "'" take-string ] }
166 { CHAR: " [ "\"" take-string ] }
169 : take-system-id ( -- system-id )
170 take-system-literal <system-id>
171 ">" take-string only-blanks ;
173 : take-public-id ( -- public-id )
175 take-system-literal <public-id>
176 ">" take-string only-blanks ;
180 : (take-internal-subset) ( -- )
181 pass-blank get-char {
183 [ drop "<!" expect-string direct , (take-internal-subset) ]
186 : take-internal-subset ( -- seq )
187 [ (take-internal-subset) ] { } make ;
189 : (take-external-id) ( token -- external-id )
191 { "SYSTEM" [ take-system-id ] }
192 { "PUBLIC" [ take-public-id ] }
196 : take-external-id ( -- external-id )
197 " " take-string (take-external-id) ;
199 : take-doctype-decl ( -- doctype-decl )
200 pass-blank " >" take-until-one-of {
202 pass-blank get-char CHAR: [ = [
203 next take-internal-subset f swap
204 ">" take-string only-blanks
206 " >" take-until-one-of {
207 { CHAR: \s [ (take-external-id) ] }
208 { CHAR: > [ only-blanks f ] }
213 } case <doctype-decl> ;
215 : take-entity-def ( -- entity-name entity-def )
216 " " take-string pass-blank get-char {
217 { CHAR: ' [ take-system-literal ] }
218 { CHAR: " [ take-system-literal ] }
219 [ drop take-external-id ]
222 : take-entity-decl ( -- entity-decl )
223 pass-blank get-char {
224 { CHAR: % [ next pass-blank take-entity-def ] }
225 [ drop take-entity-def ]
227 ">" take-string only-blanks <entity-decl> ;
229 : take-directive ( -- directive )
231 { "ELEMENT" [ take-element-decl ] }
232 { "ATTLIST" [ take-attlist-decl ] }
233 { "DOCTYPE" [ take-doctype-decl ] }
234 { "ENTITY" [ take-entity-decl ] }
238 : direct ( -- object )
240 { CHAR: - [ take-comment ] }
241 { CHAR: [ [ take-cdata ] }
242 [ drop take-directive ]
245 : yes/no>bool ( string -- t/f )
252 : assure-no-extra ( seq -- )
254 T{ name f "" "version" f }
255 T{ name f "" "encoding" f }
256 T{ name f "" "standalone" f }
258 [ extra-attrs ] unless-empty ;
260 : good-version ( version -- version )
261 dup { "1.0" "1.1" } member? [ bad-version ] unless ;
263 : prolog-attrs ( alist -- prolog )
264 [ T{ name f "" "version" f } swap at
265 [ good-version ] [ versionless-prolog ] if* ] keep
266 [ T{ name f "" "encoding" f } swap at
268 T{ name f "" "standalone" f } swap at
269 [ yes/no>bool ] [ f ] if*
272 SYMBOL: string-input?
273 : decode-input-if ( encoding -- )
274 string-input? get [ drop ] [ decode-input ] if ;
276 : parse-prolog ( -- prolog )
277 pass-blank middle-tag "?>" expect-string
278 dup assure-no-extra prolog-attrs
279 dup encoding>> dup "UTF-16" =
280 [ drop ] [ name>encoding [ decode-input-if ] when* ] if
281 dup prolog-data set ;
283 : instruct ( -- instruction )
284 "" (parse-name) dup "xml" =
285 [ drop parse-prolog ] [
287 [ capitalized-prolog ]
288 [ "?>" take-string append <instruction> ] if
291 : make-tag ( -- tag )
293 { [ get-char dup CHAR: ! = ] [ drop next direct ] }
294 { [ CHAR: ? = ] [ next instruct ] }
296 start-tag [ dup add-ns pop-ns <closer> ]
297 [ middle-tag end-tag ] if
302 ! Autodetecting encodings
304 : continue-make-tag ( str -- tag )
305 parse-name-starting middle-tag end-tag CHAR: > expect ;
307 : start-utf16le ( -- tag )
308 utf16le decode-input-if
312 : 10xxxxxx? ( ch -- ? )
313 -6 shift 3 bitand 2 = ;
315 : start<name ( ch -- tag )
317 [ utf8 decode-input-if next make-tag ] [
319 [ get-next 10xxxxxx? not ] take-until
320 get-char suffix utf8 decode
321 utf8 decode-input-if next
327 { 0 [ next next start-utf16le ] }
328 { CHAR: ? [ next next instruct ] } ! XML prolog parsing sets the encoding
329 { CHAR: ! [ utf8 decode-input next next direct ] }
333 : skip-utf8-bom ( -- tag )
334 "\u0000bb\u0000bf" expect utf8 decode-input
335 CHAR: < expect make-tag ;
337 : start-utf16be ( -- tag )
338 utf16be decode-input-if
339 next CHAR: < expect make-tag ;
341 : skip-utf16le-bom ( -- tag )
342 utf16le decode-input-if
344 CHAR: < expect make-tag ;
346 : skip-utf16be-bom ( -- tag )
347 utf16be decode-input-if
349 CHAR: < expect make-tag ;
351 : start-document ( -- tag )
353 { CHAR: < [ start< ] }
354 { 0 [ start-utf16be ] }
355 { HEX: EF [ skip-utf8-bom ] }
356 { HEX: FF [ skip-utf16le-bom ] }
357 { HEX: FE [ skip-utf16be-bom ] }
360 [ drop pass-blank utf8 decode-input-if CHAR: < expect make-tag ]
361 [ 1string ] if ! Replace with proper error