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 hashtables ;
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&& [ bad-name ] 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-named-entity ( string -- )
78 dup entities at [ , ] [
79 dup extra-entities get at
80 [ dup number? [ , ] [ % ] if ] [ no-entity ] ?if ! Make less hackish
84 next CHAR: ; take-char next
86 "x" ?head 16 10 ? base> ,
87 ] [ parse-named-entity ] if ;
89 : (parse-char) ( ch -- )
91 { [ dup not ] [ 2drop ] }
92 { [ 2dup = ] [ 2drop next ] }
93 { [ dup CHAR: & = ] [ drop parse-entity (parse-char) ] }
94 [ , next (parse-char) ]
97 : parse-char ( ch -- string )
98 [ (parse-char) ] "" make ;
100 : parse-text ( -- string )
105 : start-tag ( -- name ? )
106 #! Outputs the name and whether this is a closing tag
107 get-char CHAR: / = dup [ next ] when
110 : (parse-quote) ( ch -- string )
112 [ unclosed-quote ] unless ;
114 : parse-quote ( -- seq )
115 pass-blank get-char dup "'\"" member?
116 [ next (parse-quote) ] [ quoteless-attr ] if ;
120 pass-blank CHAR: = expect
124 : (middle-tag) ( -- )
125 pass-blank version=1.0? get-char name-start?
126 [ parse-attr (middle-tag) ] when ;
128 : middle-tag ( -- attrs-alist )
129 ! f make will make a vector if it has any elements
130 [ (middle-tag) ] f make pass-blank ;
132 : end-tag ( name attrs-alist -- tag )
133 tag-ns pass-blank get-char CHAR: / =
134 [ pop-ns <contained> next ] [ <opener> ] if ;
136 : take-comment ( -- comment )
142 : take-cdata ( -- string )
143 "[CDATA[" expect-string "]]>" take-string ;
145 : take-element-decl ( -- element-decl )
146 pass-blank " " take-string pass-blank ">" take-string <element-decl> ;
148 : take-attlist-decl ( -- doctype-decl )
149 pass-blank " " take-string pass-blank ">" take-string <attlist-decl> ;
151 : take-until-one-of ( seps -- str sep )
152 '[ get-char _ member? ] take-until get-char ;
154 : only-blanks ( str -- )
155 [ blank? ] all? [ bad-doctype-decl ] unless ;
157 : take-system-literal ( -- str ) ! replace with parse-quote?
158 pass-blank get-char next {
159 { CHAR: ' [ "'" take-string ] }
160 { CHAR: " [ "\"" take-string ] }
163 : take-system-id ( -- system-id )
164 take-system-literal <system-id>
165 ">" take-string only-blanks ;
167 : take-public-id ( -- public-id )
169 take-system-literal <public-id>
170 ">" take-string only-blanks ;
174 : (take-internal-subset) ( -- )
175 pass-blank get-char {
177 [ drop "<!" expect-string direct , (take-internal-subset) ]
180 : take-internal-subset ( -- seq )
181 [ (take-internal-subset) ] { } make ;
183 : (take-external-id) ( token -- external-id )
185 { "SYSTEM" [ take-system-id ] }
186 { "PUBLIC" [ take-public-id ] }
190 : take-external-id ( -- external-id )
191 " " take-string (take-external-id) ;
193 : take-doctype-decl ( -- doctype-decl )
194 pass-blank " >" take-until-one-of {
196 pass-blank get-char CHAR: [ = [
197 next take-internal-subset f swap
198 ">" take-string only-blanks
200 " >" take-until-one-of {
201 { CHAR: \s [ (take-external-id) ] }
202 { CHAR: > [ only-blanks f ] }
207 } case <doctype-decl> ;
209 : take-entity-def ( -- entity-name entity-def )
210 " " take-string pass-blank get-char {
211 { CHAR: ' [ parse-quote ] }
212 { CHAR: " [ parse-quote ] }
213 [ drop take-external-id ]
216 : associate-entity ( entity-name entity-def -- )
217 swap extra-entities [ ?set-at ] change ;
219 : take-entity-decl ( -- entity-decl )
220 pass-blank get-char {
221 { CHAR: % [ next pass-blank take-entity-def ] }
222 [ drop take-entity-def 2dup associate-entity ]
224 ">" take-string only-blanks <entity-decl> ;
226 : take-directive ( -- directive )
228 { "ELEMENT" [ take-element-decl ] }
229 { "ATTLIST" [ take-attlist-decl ] }
230 { "DOCTYPE" [ take-doctype-decl ] }
231 { "ENTITY" [ take-entity-decl ] }
235 : direct ( -- object )
237 { CHAR: - [ take-comment ] }
238 { CHAR: [ [ take-cdata ] }
239 [ drop take-directive ]
242 : yes/no>bool ( string -- t/f )
249 : assure-no-extra ( seq -- )
251 T{ name f "" "version" f }
252 T{ name f "" "encoding" f }
253 T{ name f "" "standalone" f }
255 [ extra-attrs ] unless-empty ;
257 : good-version ( version -- version )
258 dup { "1.0" "1.1" } member? [ bad-version ] unless ;
260 : prolog-version ( alist -- version )
261 T{ name f "" "version" f } swap at
262 [ good-version ] [ versionless-prolog ] if* ;
264 : prolog-encoding ( alist -- encoding )
265 T{ name f "" "encoding" f } swap at "UTF-8" or ;
267 : prolog-standalone ( alist -- version )
268 T{ name f "" "standalone" f } swap at
269 [ yes/no>bool ] [ f ] if* ;
271 : prolog-attrs ( alist -- prolog )
274 [ prolog-standalone ]
277 SYMBOL: string-input?
278 : decode-input-if ( encoding -- )
279 string-input? get [ drop ] [ decode-input ] if ;
281 : parse-prolog ( -- prolog )
282 pass-blank middle-tag "?>" expect-string
283 dup assure-no-extra prolog-attrs
284 dup encoding>> dup "UTF-16" =
285 [ drop ] [ name>encoding [ decode-input-if ] when* ] if
286 dup prolog-data set ;
288 : instruct ( -- instruction )
289 "" (parse-name) dup "xml" =
290 [ drop parse-prolog ] [
292 [ capitalized-prolog ]
293 [ "?>" take-string append <instruction> ] if
296 : make-tag ( -- tag )
298 { [ get-char dup CHAR: ! = ] [ drop next direct ] }
299 { [ CHAR: ? = ] [ next instruct ] }
301 start-tag [ dup add-ns pop-ns <closer> ]
302 [ middle-tag end-tag ] if
307 ! Autodetecting encodings
309 : continue-make-tag ( str -- tag )
310 parse-name-starting middle-tag end-tag CHAR: > expect ;
312 : start-utf16le ( -- tag )
313 utf16le decode-input-if
317 : 10xxxxxx? ( ch -- ? )
318 -6 shift 3 bitand 2 = ;
320 : start<name ( ch -- tag )
322 [ utf8 decode-input-if next make-tag ] [
324 [ get-next 10xxxxxx? not ] take-until
325 get-char suffix utf8 decode
326 utf8 decode-input-if next
332 { 0 [ next next start-utf16le ] }
333 { CHAR: ? [ next next instruct ] } ! XML prolog parsing sets the encoding
334 { CHAR: ! [ utf8 decode-input next next direct ] }
338 : skip-utf8-bom ( -- tag )
339 "\u0000bb\u0000bf" expect utf8 decode-input
340 CHAR: < expect make-tag ;
342 : decode-expecting ( encoding string -- tag )
343 [ decode-input-if next ] [ expect-string ] bi* make-tag ;
345 : start-utf16be ( -- tag )
346 utf16be "<" decode-expecting ;
348 : skip-utf16le-bom ( -- tag )
349 utf16le "\u0000fe<" decode-expecting ;
351 : skip-utf16be-bom ( -- tag )
352 utf16be "\u0000ff<" decode-expecting ;
354 : start-document ( -- tag )
356 { CHAR: < [ start< ] }
357 { 0 [ start-utf16be ] }
358 { HEX: EF [ skip-utf8-bom ] }
359 { HEX: FF [ skip-utf16le-bom ] }
360 { HEX: FE [ skip-utf16be-bom ] }
362 [ drop utf8 decode-input-if f ]
363 ! Same problem as with <e`>, in the case of XML chunks?