Commit | Line | Data |
---|---|---|
1cd7adc6 | 1 | ;;; xml.el --- XML parser |
47db06aa | 2 | |
acaf905b | 3 | ;; Copyright (C) 2000-2012 Free Software Foundation, Inc. |
47db06aa GM |
4 | |
5 | ;; Author: Emmanuel Briot <briot@gnat.com> | |
720058f2 | 6 | ;; Maintainer: Mark A. Hershberger <mah@everybody.org> |
a98e819b | 7 | ;; Keywords: xml, data |
47db06aa GM |
8 | |
9 | ;; This file is part of GNU Emacs. | |
10 | ||
eb3fa2cf | 11 | ;; GNU Emacs is free software: you can redistribute it and/or modify |
47db06aa | 12 | ;; it under the terms of the GNU General Public License as published by |
eb3fa2cf GM |
13 | ;; the Free Software Foundation, either version 3 of the License, or |
14 | ;; (at your option) any later version. | |
47db06aa GM |
15 | |
16 | ;; GNU Emacs is distributed in the hope that it will be useful, | |
17 | ;; but WITHOUT ANY WARRANTY; without even the implied warranty of | |
18 | ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
19 | ;; GNU General Public License for more details. | |
20 | ||
21 | ;; You should have received a copy of the GNU General Public License | |
eb3fa2cf | 22 | ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. |
47db06aa GM |
23 | |
24 | ;;; Commentary: | |
25 | ||
a98e819b DL |
26 | ;; This file contains a somewhat incomplete non-validating XML parser. It |
27 | ;; parses a file, and returns a list that can be used internally by | |
a1dfa9a3 | 28 | ;; any other Lisp libraries. |
47db06aa GM |
29 | |
30 | ;;; FILE FORMAT | |
31 | ||
a98e819b DL |
32 | ;; The document type declaration may either be ignored or (optionally) |
33 | ;; parsed, but currently the parsing will only accept element | |
a1dfa9a3 | 34 | ;; declarations. The XML file is assumed to be well-formed. In case |
a98e819b DL |
35 | ;; of error, the parsing stops and the XML file is shown where the |
36 | ;; parsing stopped. | |
47db06aa | 37 | ;; |
a98e819b | 38 | ;; It also knows how to ignore comments and processing instructions. |
47db06aa GM |
39 | ;; |
40 | ;; The XML file should have the following format: | |
653558a1 GM |
41 | ;; <node1 attr1="name1" attr2="name2" ...>value |
42 | ;; <node2 attr3="name3" attr4="name4">value2</node2> | |
43 | ;; <node3 attr5="name5" attr6="name6">value3</node3> | |
47db06aa | 44 | ;; </node1> |
a1dfa9a3 | 45 | ;; Of course, the name of the nodes and attributes can be anything. There can |
47db06aa GM |
46 | ;; be any number of attributes (or none), as well as any number of children |
47 | ;; below the nodes. | |
48 | ;; | |
49 | ;; There can be only top level node, but with any number of children below. | |
50 | ||
51 | ;;; LIST FORMAT | |
52 | ||
c7f8d055 SM |
53 | ;; The functions `xml-parse-file', `xml-parse-region' and |
54 | ;; `xml-parse-tag' return a list with the following format: | |
47db06aa GM |
55 | ;; |
56 | ;; xml-list ::= (node node ...) | |
c7f8d055 | 57 | ;; node ::= (qname attribute-list . child_node_list) |
47db06aa GM |
58 | ;; child_node_list ::= child_node child_node ... |
59 | ;; child_node ::= node | string | |
c7f8d055 SM |
60 | ;; qname ::= (:namespace-uri . "name") | "name" |
61 | ;; attribute_list ::= ((qname . "value") (qname . "value") ...) | |
47db06aa GM |
62 | ;; | nil |
63 | ;; string ::= "..." | |
64 | ;; | |
a98e819b DL |
65 | ;; Some macros are provided to ease the parsing of this list. |
66 | ;; Whitespace is preserved. Fixme: There should be a tree-walker that | |
67 | ;; can remove it. | |
47db06aa | 68 | |
c7f8d055 SM |
69 | ;; TODO: |
70 | ;; * xml:base, xml:space support | |
71 | ;; * more complete DOCTYPE parsing | |
72 | ;; * pi support | |
73 | ||
47db06aa GM |
74 | ;;; Code: |
75 | ||
f6fcdfff CY |
76 | ;; Note that buffer-substring and match-string were formerly used in |
77 | ;; several places, because the -no-properties variants remove | |
78 | ;; composition info. However, after some discussion on emacs-devel, | |
79 | ;; the consensus was that the speed of the -no-properties variants was | |
80 | ;; a worthwhile tradeoff especially since we're usually parsing files | |
81 | ;; instead of hand-crafted XML. | |
a98e819b | 82 | |
a7aef6f5 | 83 | ;;; Macros to parse the list |
47db06aa | 84 | |
f8ab034e MH |
85 | (defconst xml-undefined-entity "?" |
86 | "What to substitute for undefined entities") | |
87 | ||
a7aef6f5 CY |
88 | (defconst xml-default-ns '(("" . "") |
89 | ("xml" . "http://www.w3.org/XML/1998/namespace") | |
90 | ("xmlns" . "http://www.w3.org/2000/xmlns/")) | |
91 | "Alist mapping default XML namespaces to their URIs.") | |
92 | ||
6d12a4df | 93 | (defvar xml-entity-alist |
a7aef6f5 | 94 | '(("lt" . "<") |
6d12a4df MH |
95 | ("gt" . ">") |
96 | ("apos" . "'") | |
97 | ("quot" . "\"") | |
a7aef6f5 CY |
98 | ("amp" . "&")) |
99 | "Alist mapping XML entities to their replacement text.") | |
7f3fbd5d | 100 | |
a76e6535 CY |
101 | (defvar xml-entity-expansion-limit 20000 |
102 | "The maximum size of entity reference expansions. | |
103 | If the size of the buffer increases by this many characters while | |
104 | expanding entity references in a segment of character data, the | |
105 | XML parser signals an error. Setting this to nil removes the | |
106 | limit (making the parser vulnerable to XML bombs).") | |
107 | ||
7f3fbd5d CY |
108 | (defvar xml-parameter-entity-alist nil |
109 | "Alist of defined XML parametric entities.") | |
6d12a4df MH |
110 | |
111 | (defvar xml-sub-parser nil | |
7f3fbd5d | 112 | "Non-nil when the XML parser is parsing an XML fragment.") |
6d12a4df MH |
113 | |
114 | (defvar xml-validating-parser nil | |
115 | "Set to non-nil to get validity checking.") | |
116 | ||
971489ea | 117 | (defsubst xml-node-name (node) |
47db06aa | 118 | "Return the tag associated with NODE. |
a1dfa9a3 SM |
119 | Without namespace-aware parsing, the tag is a symbol. |
120 | ||
121 | With namespace-aware parsing, the tag is a cons of a string | |
122 | representing the uri of the namespace with the local name of the | |
123 | tag. For example, | |
124 | ||
125 | <foo> | |
126 | ||
127 | would be represented by | |
128 | ||
049a0936 DE |
129 | '(\"\" . \"foo\"). |
130 | ||
131 | If you'd just like a plain symbol instead, use 'symbol-qnames in | |
132 | the PARSE-NS argument." | |
a1dfa9a3 | 133 | |
971489ea | 134 | (car node)) |
47db06aa | 135 | |
971489ea | 136 | (defsubst xml-node-attributes (node) |
47db06aa GM |
137 | "Return the list of attributes of NODE. |
138 | The list can be nil." | |
971489ea | 139 | (nth 1 node)) |
47db06aa | 140 | |
971489ea | 141 | (defsubst xml-node-children (node) |
47db06aa GM |
142 | "Return the list of children of NODE. |
143 | This is a list of nodes, and it can be nil." | |
971489ea | 144 | (cddr node)) |
47db06aa GM |
145 | |
146 | (defun xml-get-children (node child-name) | |
147 | "Return the children of NODE whose tag is CHILD-NAME. | |
a1dfa9a3 | 148 | CHILD-NAME should match the value returned by `xml-node-name'." |
971489ea SM |
149 | (let ((match ())) |
150 | (dolist (child (xml-node-children node)) | |
a1dfa9a3 SM |
151 | (if (and (listp child) |
152 | (equal (xml-node-name child) child-name)) | |
153 | (push child match))) | |
971489ea | 154 | (nreverse match))) |
47db06aa | 155 | |
9bcd6a7e | 156 | (defun xml-get-attribute-or-nil (node attribute) |
47db06aa | 157 | "Get from NODE the value of ATTRIBUTE. |
a1dfa9a3 | 158 | Return nil if the attribute was not found. |
9bcd6a7e EZ |
159 | |
160 | See also `xml-get-attribute'." | |
2e9bdf15 | 161 | (cdr (assoc attribute (xml-node-attributes node)))) |
9bcd6a7e EZ |
162 | |
163 | (defsubst xml-get-attribute (node attribute) | |
164 | "Get from NODE the value of ATTRIBUTE. | |
165 | An empty string is returned if the attribute was not found. | |
166 | ||
167 | See also `xml-get-attribute-or-nil'." | |
168 | (or (xml-get-attribute-or-nil node attribute) "")) | |
47db06aa | 169 | |
566df3fc | 170 | ;;; Regular expressions for XML components |
47db06aa | 171 | |
566df3fc CY |
172 | ;; The following regexps are used as subexpressions in regexps that |
173 | ;; are `eval-when-compile'd for efficiency, so they must be defined at | |
174 | ;; compile time. | |
b3218de1 | 175 | (eval-and-compile |
566df3fc CY |
176 | |
177 | ;; [4] NameStartChar | |
178 | ;; See the definition of word syntax in `xml-syntax-table'. | |
179 | (defconst xml-name-start-char-re (concat "[[:word:]:_]")) | |
180 | ||
181 | ;; [4a] NameChar ::= NameStartChar | "-" | "." | [0-9] | #xB7 | |
182 | ;; | [#x0300-#x036F] | [#x203F-#x2040] | |
183 | (defconst xml-name-char-re (concat "[-0-9.[:word:]:_·̀-ͯ‿-⁀]")) | |
184 | ||
185 | ;; [5] Name ::= NameStartChar (NameChar)* | |
186 | (defconst xml-name-re (concat xml-name-start-char-re xml-name-char-re "*")) | |
187 | ||
188 | ;; [6] Names ::= Name (#x20 Name)* | |
189 | (defconst xml-names-re (concat xml-name-re "\\(?: " xml-name-re "\\)*")) | |
190 | ||
191 | ;; [7] Nmtoken ::= (NameChar)+ | |
192 | (defconst xml-nmtoken-re (concat xml-name-char-re "+")) | |
193 | ||
194 | ;; [8] Nmtokens ::= Nmtoken (#x20 Nmtoken)* | |
195 | (defconst xml-nmtokens-re (concat xml-nmtoken-re "\\(?: " xml-name-re "\\)*")) | |
196 | ||
197 | ;; [66] CharRef ::= '&#' [0-9]+ ';' | '&#x' [0-9a-fA-F]+ ';' | |
198 | (defconst xml-char-ref-re "\\(?:&#[0-9]+;\\|&#x[0-9a-fA-F]+;\\)") | |
199 | ||
200 | ;; [68] EntityRef ::= '&' Name ';' | |
201 | (defconst xml-entity-ref (concat "&" xml-name-re ";")) | |
202 | ||
203 | (defconst xml-entity-or-char-ref-re (concat "&\\(?:#\\(x\\)?\\([0-9]+\\)\\|\\(" | |
204 | xml-name-re "\\)\\);")) | |
205 | ||
206 | ;; [69] PEReference ::= '%' Name ';' | |
207 | (defconst xml-pe-reference-re (concat "%\\(" xml-name-re "\\);")) | |
208 | ||
209 | ;; [67] Reference ::= EntityRef | CharRef | |
210 | (defconst xml-reference-re (concat "\\(?:" xml-entity-ref "\\|" xml-char-ref-re "\\)")) | |
211 | ||
212 | ;; [10] AttValue ::= '"' ([^<&"] | Reference)* '"' | |
213 | ;; | "'" ([^<&'] | Reference)* "'" | |
214 | (defconst xml-att-value-re (concat "\\(?:\"\\(?:[^&\"]\\|" | |
215 | xml-reference-re "\\)*\"\\|" | |
216 | "'\\(?:[^&']\\|" xml-reference-re | |
217 | "\\)*'\\)")) | |
218 | ||
219 | ;; [56] TokenizedType ::= 'ID' | |
220 | ;; [VC: ID] [VC: One ID / Element Type] [VC: ID Attribute Default] | |
221 | ;; | 'IDREF' [VC: IDREF] | |
222 | ;; | 'IDREFS' [VC: IDREF] | |
223 | ;; | 'ENTITY' [VC: Entity Name] | |
224 | ;; | 'ENTITIES' [VC: Entity Name] | |
225 | ;; | 'NMTOKEN' [VC: Name Token] | |
226 | ;; | 'NMTOKENS' [VC: Name Token] | |
227 | (defconst xml-tokenized-type-re (concat "\\(?:ID\\|IDREF\\|IDREFS\\|ENTITY\\|" | |
228 | "ENTITIES\\|NMTOKEN\\|NMTOKENS\\)")) | |
229 | ||
230 | ;; [58] NotationType ::= 'NOTATION' S '(' S? Name (S? '|' S? Name)* S? ')' | |
231 | (defconst xml-notation-type-re | |
232 | (concat "\\(?:NOTATION\\s-+(\\s-*" xml-name-re | |
233 | "\\(?:\\s-*|\\s-*" xml-name-re "\\)*\\s-*)\\)")) | |
234 | ||
235 | ;; [59] Enumeration ::= '(' S? Nmtoken (S? '|' S? Nmtoken)* S? ')' | |
236 | ;; [VC: Enumeration] [VC: No Duplicate Tokens] | |
237 | (defconst xml-enumeration-re (concat "\\(?:(\\s-*" xml-nmtoken-re | |
238 | "\\(?:\\s-*|\\s-*" xml-nmtoken-re | |
239 | "\\)*\\s-+)\\)")) | |
240 | ||
241 | ;; [57] EnumeratedType ::= NotationType | Enumeration | |
242 | (defconst xml-enumerated-type-re (concat "\\(?:" xml-notation-type-re | |
243 | "\\|" xml-enumeration-re "\\)")) | |
244 | ||
245 | ;; [54] AttType ::= StringType | TokenizedType | EnumeratedType | |
246 | ;; [55] StringType ::= 'CDATA' | |
247 | (defconst xml-att-type-re (concat "\\(?:CDATA\\|" xml-tokenized-type-re | |
248 | "\\|" xml-notation-type-re | |
249 | "\\|" xml-enumerated-type-re "\\)")) | |
250 | ||
251 | ;; [60] DefaultDecl ::= '#REQUIRED' | '#IMPLIED' | (('#FIXED' S)? AttValue) | |
252 | (defconst xml-default-decl-re (concat "\\(?:#REQUIRED\\|#IMPLIED\\|" | |
253 | "\\(?:#FIXED\\s-+\\)*" | |
254 | xml-att-value-re "\\)")) | |
255 | ||
256 | ;; [53] AttDef ::= S Name S AttType S DefaultDecl | |
257 | (defconst xml-att-def-re (concat "\\(?:\\s-*" xml-name-re | |
258 | "\\s-*" xml-att-type-re | |
259 | "\\s-*" xml-default-decl-re "\\)")) | |
260 | ||
261 | ;; [9] EntityValue ::= '"' ([^%&"] | PEReference | Reference)* '"' | |
262 | ;; | "'" ([^%&'] | PEReference | Reference)* "'" | |
263 | (defconst xml-entity-value-re (concat "\\(?:\"\\(?:[^%&\"]\\|" | |
264 | xml-pe-reference-re | |
265 | "\\|" xml-reference-re | |
266 | "\\)*\"\\|'\\(?:[^%&']\\|" | |
267 | xml-pe-reference-re "\\|" | |
268 | xml-reference-re "\\)*'\\)")) | |
269 | ) ; End of `eval-when-compile' | |
270 | ||
b3218de1 CY |
271 | |
272 | ;; [75] ExternalID ::= 'SYSTEM' S SystemLiteral | |
273 | ;; | 'PUBLIC' S PubidLiteral S SystemLiteral | |
274 | ;; [76] NDataDecl ::= S 'NDATA' S | |
275 | ;; [73] EntityDef ::= EntityValue| (ExternalID NDataDecl?) | |
276 | ;; [71] GEDecl ::= '<!ENTITY' S Name S EntityDef S? '>' | |
277 | ;; [74] PEDef ::= EntityValue | ExternalID | |
278 | ;; [72] PEDecl ::= '<!ENTITY' S '%' S Name S PEDef S? '>' | |
279 | ;; [70] EntityDecl ::= GEDecl | PEDecl | |
6d12a4df | 280 | |
a98e819b DL |
281 | ;; Note that this is setup so that we can do whitespace-skipping with |
282 | ;; `(skip-syntax-forward " ")', inter alia. Previously this was slow | |
566df3fc | 283 | ;; compared with `re-search-forward', but that has been fixed. |
a98e819b DL |
284 | |
285 | (defvar xml-syntax-table | |
566df3fc CY |
286 | ;; By default, characters have symbol syntax. |
287 | (let ((table (make-char-table 'syntax-table '(3)))) | |
288 | ;; The XML space chars [3], and nothing else, have space syntax. | |
289 | (dolist (c '(?\s ?\t ?\r ?\n)) | |
a98e819b | 290 | (modify-syntax-entry c " " table)) |
566df3fc CY |
291 | ;; The characters in NameStartChar [4], aside from ':' and '_', |
292 | ;; have word syntax. This is used by `xml-name-start-char-re'. | |
293 | (modify-syntax-entry '(?A . ?Z) "w" table) | |
294 | (modify-syntax-entry '(?a . ?z) "w" table) | |
295 | (modify-syntax-entry '(#xC0 . #xD6) "w" table) | |
296 | (modify-syntax-entry '(#xD8 . #XF6) "w" table) | |
297 | (modify-syntax-entry '(#xF8 . #X2FF) "w" table) | |
298 | (modify-syntax-entry '(#x370 . #X37D) "w" table) | |
299 | (modify-syntax-entry '(#x37F . #x1FFF) "w" table) | |
300 | (modify-syntax-entry '(#x200C . #x200D) "w" table) | |
301 | (modify-syntax-entry '(#x2070 . #x218F) "w" table) | |
302 | (modify-syntax-entry '(#x2C00 . #x2FEF) "w" table) | |
303 | (modify-syntax-entry '(#x3001 . #xD7FF) "w" table) | |
304 | (modify-syntax-entry '(#xF900 . #xFDCF) "w" table) | |
305 | (modify-syntax-entry '(#xFDF0 . #xFFFD) "w" table) | |
306 | (modify-syntax-entry '(#x10000 . #xEFFFF) "w" table) | |
a98e819b | 307 | table) |
566df3fc CY |
308 | "Syntax table used by the XML parser. |
309 | In this syntax table, the XML space characters [ \\t\\r\\n], and | |
310 | only those characters, have whitespace syntax.") | |
a98e819b | 311 | |
566df3fc | 312 | ;;; Entry points: |
a98e819b | 313 | |
566df3fc CY |
314 | ;;;###autoload |
315 | (defun xml-parse-file (file &optional parse-dtd parse-ns) | |
316 | "Parse the well-formed XML file FILE. | |
317 | Return the top node with all its children. | |
318 | If PARSE-DTD is non-nil, the DTD is parsed rather than skipped. | |
049a0936 DE |
319 | |
320 | If PARSE-NS is non-nil, then QNAMES are expanded. By default, | |
321 | the variable `xml-default-ns' is the mapping from namespaces to | |
322 | URIs, and expanded names will be returned as a cons | |
323 | ||
324 | (\"namespace:\" . \"foo\"). | |
325 | ||
326 | If PARSE-NS is an alist, it will be used as the mapping from | |
327 | namespace to URIs instead. | |
328 | ||
329 | If it is the symbol 'symbol-qnames, expanded names will be | |
330 | returned as a plain symbol 'namespace:foo instead of a cons. | |
331 | ||
332 | Both features can be combined by providing a cons cell | |
333 | ||
334 | (symbol-qnames . ALIST)." | |
566df3fc CY |
335 | (with-temp-buffer |
336 | (insert-file-contents file) | |
337 | (xml--parse-buffer parse-dtd parse-ns))) | |
a98e819b DL |
338 | |
339 | ;;;###autoload | |
a7aef6f5 | 340 | (defun xml-parse-region (&optional beg end buffer parse-dtd parse-ns) |
47db06aa | 341 | "Parse the region from BEG to END in BUFFER. |
566df3fc CY |
342 | Return the XML parse tree, or raise an error if the region does |
343 | not contain well-formed XML. | |
344 | ||
a7aef6f5 CY |
345 | If BEG is nil, it defaults to `point-min'. |
346 | If END is nil, it defaults to `point-max'. | |
47db06aa | 347 | If BUFFER is nil, it defaults to the current buffer. |
566df3fc CY |
348 | If PARSE-DTD is non-nil, parse the DTD and return it as the first |
349 | element of the list. | |
049a0936 DE |
350 | If PARSE-NS is non-nil, then QNAMES are expanded. By default, |
351 | the variable `xml-default-ns' is the mapping from namespaces to | |
352 | URIs, and expanded names will be returned as a cons | |
353 | ||
354 | (\"namespace:\" . \"foo\"). | |
355 | ||
356 | If PARSE-NS is an alist, it will be used as the mapping from | |
357 | namespace to URIs instead. | |
358 | ||
359 | If it is the symbol 'symbol-qnames, expanded names will be | |
360 | returned as a plain symbol 'namespace:foo instead of a cons. | |
361 | ||
362 | Both features can be combined by providing a cons cell | |
363 | ||
364 | (symbol-qnames . ALIST)." | |
39d58fc0 MH |
365 | ;; Use fixed syntax table to ensure regexp char classes and syntax |
366 | ;; specs DTRT. | |
fbf2e7ad CY |
367 | (unless buffer |
368 | (setq buffer (current-buffer))) | |
369 | (with-temp-buffer | |
a7aef6f5 | 370 | (insert-buffer-substring-no-properties buffer beg end) |
fbf2e7ad CY |
371 | (xml--parse-buffer parse-dtd parse-ns))) |
372 | ||
566df3fc CY |
373 | ;; XML [5] |
374 | ||
375 | ;; Fixme: This needs re-writing to deal with the XML grammar properly, i.e. | |
376 | ;; document ::= prolog element Misc* | |
377 | ;; prolog ::= XMLDecl? Misc* (doctypedecl Misc*)? | |
378 | ||
fbf2e7ad | 379 | (defun xml--parse-buffer (parse-dtd parse-ns) |
566df3fc | 380 | (with-syntax-table xml-syntax-table |
39d58fc0 | 381 | (let ((case-fold-search nil) ; XML is case-sensitive. |
7f3fbd5d CY |
382 | ;; Prevent entity definitions from changing the defaults |
383 | (xml-entity-alist xml-entity-alist) | |
384 | (xml-parameter-entity-alist xml-parameter-entity-alist) | |
39d58fc0 | 385 | xml result dtd) |
fbf2e7ad CY |
386 | (goto-char (point-min)) |
387 | (while (not (eobp)) | |
388 | (if (search-forward "<" nil t) | |
389 | (progn | |
390 | (forward-char -1) | |
a7aef6f5 | 391 | (setq result (xml-parse-tag-1 parse-dtd parse-ns)) |
fbf2e7ad CY |
392 | (cond |
393 | ((null result) | |
394 | ;; Not looking at an xml start tag. | |
395 | (unless (eobp) | |
396 | (forward-char 1))) | |
397 | ((and xml (not xml-sub-parser)) | |
398 | ;; Translation of rule [1] of XML specifications | |
399 | (error "XML: (Not Well-Formed) Only one root tag allowed")) | |
400 | ((and (listp (car result)) | |
401 | parse-dtd) | |
402 | (setq dtd (car result)) | |
403 | (if (cdr result) ; possible leading comment | |
404 | (add-to-list 'xml (cdr result)))) | |
405 | (t | |
406 | (add-to-list 'xml result)))) | |
407 | (goto-char (point-max)))) | |
408 | (if parse-dtd | |
409 | (cons dtd (nreverse xml)) | |
410 | (nreverse xml))))) | |
47db06aa | 411 | |
c7f8d055 | 412 | (defun xml-maybe-do-ns (name default xml-ns) |
a1dfa9a3 SM |
413 | "Perform any namespace expansion. |
414 | NAME is the name to perform the expansion on. | |
c7f8d055 SM |
415 | DEFAULT is the default namespace. XML-NS is a cons of namespace |
416 | names to uris. When namespace-aware parsing is off, then XML-NS | |
417 | is nil. | |
418 | ||
419 | During namespace-aware parsing, any name without a namespace is | |
420 | put into the namespace identified by DEFAULT. nil is used to | |
049a0936 DE |
421 | specify that the name shouldn't be given a namespace. |
422 | Expanded names will by default be returned as a cons. If you | |
423 | would like to get plain symbols instead, provide a cons cell | |
424 | ||
425 | (symbol-qnames . ALIST) | |
426 | ||
427 | in the XML-NS argument." | |
c7f8d055 | 428 | (if (consp xml-ns) |
049a0936 DE |
429 | (let* ((symbol-qnames (eq (car-safe xml-ns) 'symbol-qnames)) |
430 | (nsp (string-match ":" name)) | |
c7f8d055 SM |
431 | (lname (if nsp (substring name (match-end 0)) name)) |
432 | (prefix (if nsp (substring name 0 (match-beginning 0)) default)) | |
433 | (special (and (string-equal lname "xmlns") (not prefix))) | |
434 | ;; Setting default to nil will insure that there is not | |
435 | ;; matching cons in xml-ns. In which case we | |
436 | (ns (or (cdr (assoc (if special "xmlns" prefix) | |
049a0936 | 437 | (if symbol-qnames (cdr xml-ns) xml-ns))) |
6d12a4df | 438 | ""))) |
049a0936 DE |
439 | (if (and symbol-qnames |
440 | (not (string= prefix "xmlns"))) | |
441 | (intern (concat ns lname)) | |
442 | (cons ns (if special "" lname)))) | |
c7f8d055 | 443 | (intern name))) |
47db06aa | 444 | |
2d42509a | 445 | (defun xml-parse-tag (&optional parse-dtd parse-ns) |
a98e819b | 446 | "Parse the tag at point. |
47db06aa GM |
447 | If PARSE-DTD is non-nil, the DTD of the document, if any, is parsed and |
448 | returned as the first element in the list. | |
049a0936 DE |
449 | If PARSE-NS is non-nil, expand QNAMES; for further details, see |
450 | `xml-parse-region'. | |
a7aef6f5 CY |
451 | |
452 | Return one of: | |
a98e819b DL |
453 | - a list : the matching node |
454 | - nil : the point is not looking at a tag. | |
455 | - a pair : the first element is the DTD, the second is the node." | |
566df3fc CY |
456 | (let* ((case-fold-search nil) |
457 | ;; Prevent entity definitions from changing the defaults | |
458 | (xml-entity-alist xml-entity-alist) | |
459 | (xml-parameter-entity-alist xml-parameter-entity-alist) | |
460 | (buf (current-buffer)) | |
461 | (pos (point))) | |
a7aef6f5 | 462 | (with-temp-buffer |
566df3fc CY |
463 | (with-syntax-table xml-syntax-table |
464 | (insert-buffer-substring-no-properties buf pos) | |
465 | (goto-char (point-min)) | |
466 | (xml-parse-tag-1 parse-dtd parse-ns))))) | |
a7aef6f5 CY |
467 | |
468 | (defun xml-parse-tag-1 (&optional parse-dtd parse-ns) | |
469 | "Like `xml-parse-tag', but possibly modify the buffer while working." | |
049a0936 DE |
470 | (let* ((xml-validating-parser (or parse-dtd xml-validating-parser)) |
471 | (xml-ns | |
472 | (cond ((eq parse-ns 'symbol-qnames) | |
473 | (cons 'symbol-qnames xml-default-ns)) | |
474 | ((or (consp (car-safe parse-ns)) | |
475 | (and (eq (car-safe parse-ns) 'symbol-qnames) | |
476 | (listp (cdr parse-ns)))) | |
477 | parse-ns) | |
478 | (parse-ns | |
479 | xml-default-ns)))) | |
2d42509a | 480 | (cond |
a7aef6f5 | 481 | ;; Processing instructions, like <?xml version="1.0"?>. |
2d42509a JB |
482 | ((looking-at "<\\?") |
483 | (search-forward "?>") | |
484 | (skip-syntax-forward " ") | |
a7aef6f5 CY |
485 | (xml-parse-tag-1 parse-dtd xml-ns)) |
486 | ;; Character data (CDATA) sections, in which no tag should be interpreted | |
2d42509a JB |
487 | ((looking-at "<!\\[CDATA\\[") |
488 | (let ((pos (match-end 0))) | |
489 | (unless (search-forward "]]>" nil t) | |
6d12a4df | 490 | (error "XML: (Not Well Formed) CDATA section does not end anywhere in the document")) |
ae026110 | 491 | (concat |
f6fcdfff | 492 | (buffer-substring-no-properties pos (match-beginning 0)) |
ae026110 | 493 | (xml-parse-string)))) |
a7aef6f5 | 494 | ;; DTD for the document |
7f3fbd5d | 495 | ((looking-at "<!DOCTYPE[ \t\n\r]") |
6d12a4df MH |
496 | (let ((dtd (xml-parse-dtd parse-ns))) |
497 | (skip-syntax-forward " ") | |
498 | (if xml-validating-parser | |
a7aef6f5 CY |
499 | (cons dtd (xml-parse-tag-1 nil xml-ns)) |
500 | (xml-parse-tag-1 nil xml-ns)))) | |
501 | ;; skip comments | |
2d42509a JB |
502 | ((looking-at "<!--") |
503 | (search-forward "-->") | |
a7aef6f5 | 504 | ;; FIXME: This loses the skipped-over spaces. |
a268160b | 505 | (skip-syntax-forward " ") |
18edb22d | 506 | (unless (eobp) |
772b2e2c | 507 | (let ((xml-sub-parser t)) |
a7aef6f5 CY |
508 | (xml-parse-tag-1 parse-dtd xml-ns)))) |
509 | ;; end tag | |
2d42509a JB |
510 | ((looking-at "</") |
511 | '()) | |
a7aef6f5 CY |
512 | ;; opening tag |
513 | ((looking-at (eval-when-compile (concat "<\\(" xml-name-re "\\)"))) | |
2d42509a | 514 | (goto-char (match-end 1)) |
34638996 | 515 | ;; Parse this node |
f6fcdfff | 516 | (let* ((node-name (match-string-no-properties 1)) |
5178753d MH |
517 | ;; Parse the attribute list. |
518 | (attrs (xml-parse-attlist xml-ns)) | |
06b60517 | 519 | children) |
5178753d MH |
520 | ;; add the xmlns:* attrs to our cache |
521 | (when (consp xml-ns) | |
c7f8d055 SM |
522 | (dolist (attr attrs) |
523 | (when (and (consp (car attr)) | |
6d12a4df MH |
524 | (equal "http://www.w3.org/2000/xmlns/" |
525 | (caar attr))) | |
526 | (push (cons (cdar attr) (cdr attr)) | |
049a0936 DE |
527 | (if (symbolp (car xml-ns)) |
528 | (cdr xml-ns) | |
529 | xml-ns))))) | |
5178753d | 530 | (setq children (list attrs (xml-maybe-do-ns node-name "" xml-ns))) |
a7aef6f5 CY |
531 | (cond |
532 | ;; is this an empty element ? | |
533 | ((looking-at "/>") | |
534 | (forward-char 2) | |
535 | (nreverse children)) | |
536 | ;; is this a valid start tag ? | |
537 | ((eq (char-after) ?>) | |
538 | (forward-char 1) | |
539 | ;; Now check that we have the right end-tag. | |
540 | (let ((end (concat "</" node-name "\\s-*>"))) | |
541 | (while (not (looking-at end)) | |
542 | (cond | |
543 | ((eobp) | |
a76e6535 | 544 | (error "XML: (Not Well-Formed) End of document while reading element `%s'" |
a7aef6f5 CY |
545 | node-name)) |
546 | ((looking-at "</") | |
547 | (forward-char 2) | |
548 | (error "XML: (Not Well-Formed) Invalid end tag `%s' (expecting `%s')" | |
549 | (let ((pos (point))) | |
550 | (buffer-substring pos (if (re-search-forward "\\s-*>" nil t) | |
551 | (match-beginning 0) | |
552 | (point-max)))) | |
553 | node-name)) | |
554 | ;; Read a sub-element and push it onto CHILDREN. | |
555 | ((= (char-after) ?<) | |
556 | (let ((tag (xml-parse-tag-1 nil xml-ns))) | |
557 | (when tag | |
558 | (push tag children)))) | |
559 | ;; Read some character data. | |
560 | (t | |
561 | (let ((expansion (xml-parse-string))) | |
562 | (push (if (stringp (car children)) | |
563 | ;; If two strings were separated by a | |
564 | ;; comment, concat them. | |
565 | (concat (pop children) expansion) | |
566 | expansion) | |
567 | children))))) | |
568 | ;; Move point past the end-tag. | |
569 | (goto-char (match-end 0)) | |
570 | (nreverse children))) | |
571 | ;; Otherwise this was an invalid start tag (expected ">" not found.) | |
572 | (t | |
573 | (error "XML: (Well-Formed) Couldn't parse tag: %s" | |
574 | (buffer-substring-no-properties (- (point) 10) (+ (point) 1))))))) | |
575 | ||
576 | ;; (Not one of PI, CDATA, Comment, End tag, or Start tag) | |
577 | (t | |
578 | (unless xml-sub-parser ; Usually, we error out. | |
6d12a4df | 579 | (error "XML: (Well-Formed) Invalid character")) |
6d12a4df MH |
580 | ;; However, if we're parsing incrementally, then we need to deal |
581 | ;; with stray CDATA. | |
582 | (xml-parse-string))))) | |
583 | ||
584 | (defun xml-parse-string () | |
a7aef6f5 CY |
585 | "Parse character data at point, and return it as a string. |
586 | Leave point at the start of the next thing to parse. This | |
587 | function can modify the buffer by expanding entity and character | |
588 | references." | |
589 | (let ((start (point)) | |
a76e6535 CY |
590 | ;; Keep track of the size of the rest of the buffer: |
591 | (old-remaining-size (- (buffer-size) (point))) | |
a7aef6f5 CY |
592 | ref val) |
593 | (while (and (not (eobp)) | |
594 | (not (looking-at "<"))) | |
595 | ;; Find the next < or & character. | |
596 | (skip-chars-forward "^<&") | |
597 | (when (eq (char-after) ?&) | |
598 | ;; If we find an entity or character reference, expand it. | |
566df3fc | 599 | (unless (looking-at xml-entity-or-char-ref-re) |
a7aef6f5 CY |
600 | (error "XML: (Not Well-Formed) Invalid entity reference")) |
601 | ;; For a character reference, the next entity or character | |
602 | ;; reference must be after the replacement. [4.6] "Numerical | |
603 | ;; character references are expanded immediately when | |
604 | ;; recognized and MUST be treated as character data." | |
566df3fc CY |
605 | (if (setq ref (match-string 2)) |
606 | (progn ; Numeric char reference | |
607 | (setq val (save-match-data | |
608 | (decode-char 'ucs (string-to-number | |
609 | ref (if (match-string 1) 16))))) | |
610 | (and (null val) | |
611 | xml-validating-parser | |
612 | (error "XML: (Validity) Invalid character reference `%s'" | |
613 | (match-string 0))) | |
614 | (replace-match (or (string val) xml-undefined-entity) t t)) | |
615 | ;; For an entity reference, search again from the start of | |
616 | ;; the replaced text, since the replacement can contain | |
617 | ;; entity or character references, or markup. | |
618 | (setq ref (match-string 3) | |
619 | val (assoc ref xml-entity-alist)) | |
620 | (and (null val) | |
621 | xml-validating-parser | |
622 | (error "XML: (Validity) Undefined entity `%s'" ref)) | |
623 | (replace-match (cdr val) t t) | |
624 | (goto-char (match-beginning 0))) | |
a76e6535 CY |
625 | ;; Check for XML bombs. |
626 | (and xml-entity-expansion-limit | |
627 | (> (- (buffer-size) (point)) | |
628 | (+ old-remaining-size xml-entity-expansion-limit)) | |
629 | (error "XML: Entity reference expansion \ | |
630 | surpassed `xml-entity-expansion-limit'")))) | |
a7aef6f5 CY |
631 | ;; [2.11] Clean up line breaks. |
632 | (let ((end-marker (point-marker))) | |
633 | (goto-char start) | |
634 | (while (re-search-forward "\r\n?" end-marker t) | |
635 | (replace-match "\n" t t)) | |
636 | (goto-char end-marker) | |
637 | (buffer-substring start (point))))) | |
47db06aa | 638 | |
c7f8d055 | 639 | (defun xml-parse-attlist (&optional xml-ns) |
a1dfa9a3 SM |
640 | "Return the attribute-list after point. |
641 | Leave point at the first non-blank character after the tag." | |
971489ea | 642 | (let ((attlist ()) |
34638996 | 643 | end-pos name) |
a98e819b DL |
644 | (skip-syntax-forward " ") |
645 | (while (looking-at (eval-when-compile | |
30eabd7a | 646 | (concat "\\(" xml-name-re "\\)\\s-*=\\s-*"))) |
c7f8d055 | 647 | (setq end-pos (match-end 0)) |
f6fcdfff | 648 | (setq name (xml-maybe-do-ns (match-string-no-properties 1) nil xml-ns)) |
c7f8d055 | 649 | (goto-char end-pos) |
47db06aa | 650 | |
a158ff81 JB |
651 | ;; See also: http://www.w3.org/TR/2000/REC-xml-20001006#AVNormalize |
652 | ||
47db06aa GM |
653 | ;; Do we have a string between quotes (or double-quotes), |
654 | ;; or a simple word ? | |
a158ff81 | 655 | (if (looking-at "\"\\([^\"]*\\)\"") |
34638996 | 656 | (setq end-pos (match-end 0)) |
f0ec1711 | 657 | (if (looking-at "'\\([^']*\\)'") |
34638996 | 658 | (setq end-pos (match-end 0)) |
6d12a4df | 659 | (error "XML: (Not Well-Formed) Attribute values must be given between quotes"))) |
47db06aa GM |
660 | |
661 | ;; Each attribute must be unique within a given element | |
662 | (if (assoc name attlist) | |
6d12a4df | 663 | (error "XML: (Not Well-Formed) Each attribute must be unique within an element")) |
524425ae | 664 | |
a158ff81 JB |
665 | ;; Multiple whitespace characters should be replaced with a single one |
666 | ;; in the attributes | |
06b60517 | 667 | (let ((string (match-string-no-properties 1))) |
a98e819b | 668 | (replace-regexp-in-string "\\s-\\{2,\\}" " " string) |
6d12a4df MH |
669 | (let ((expansion (xml-substitute-special string))) |
670 | (unless (stringp expansion) | |
566df3fc CY |
671 | ;; We say this is the constraint. It is actually that |
672 | ;; neither external entities nor "<" can be in an | |
673 | ;; attribute value. | |
6d12a4df MH |
674 | (error "XML: (Not Well-Formed) Entities in attributes cannot expand into elements")) |
675 | (push (cons name expansion) attlist))) | |
a158ff81 | 676 | |
34638996 | 677 | (goto-char end-pos) |
a98e819b | 678 | (skip-syntax-forward " ")) |
971489ea | 679 | (nreverse attlist))) |
47db06aa | 680 | |
a7aef6f5 | 681 | ;;; DTD (document type declaration) |
47db06aa | 682 | |
a7aef6f5 CY |
683 | ;; The following functions know how to skip or parse the DTD of a |
684 | ;; document. FIXME: it fails at least if the DTD contains conditional | |
685 | ;; sections. | |
a98e819b DL |
686 | |
687 | (defun xml-skip-dtd () | |
688 | "Skip the DTD at point. | |
47db06aa | 689 | This follows the rule [28] in the XML specifications." |
6d12a4df MH |
690 | (let ((xml-validating-parser nil)) |
691 | (xml-parse-dtd))) | |
47db06aa | 692 | |
6d12a4df | 693 | (defun xml-parse-dtd (&optional parse-ns) |
a98e819b DL |
694 | "Parse the DTD at point." |
695 | (forward-char (eval-when-compile (length "<!DOCTYPE"))) | |
696 | (skip-syntax-forward " ") | |
6d12a4df MH |
697 | (if (and (looking-at ">") |
698 | xml-validating-parser) | |
699 | (error "XML: (Validity) Invalid DTD (expecting name of the document)")) | |
524425ae | 700 | |
971489ea | 701 | ;; Get the name of the document |
30eabd7a | 702 | (looking-at xml-name-re) |
f6fcdfff | 703 | (let ((dtd (list (match-string-no-properties 0) 'dtd)) |
fbf2e7ad | 704 | (xml-parameter-entity-alist xml-parameter-entity-alist) |
fbf2e7ad | 705 | next-parameter-entity) |
47db06aa | 706 | (goto-char (match-end 0)) |
a98e819b | 707 | (skip-syntax-forward " ") |
7f3fbd5d CY |
708 | |
709 | ;; External subset (XML [75]) | |
a98e819b DL |
710 | (cond ((looking-at "PUBLIC\\s-+") |
711 | (goto-char (match-end 0)) | |
712 | (unless (or (re-search-forward | |
713 | "\\=\"\\([[:space:][:alnum:]-'()+,./:=?;!*#@$_%]*\\)\"" | |
714 | nil t) | |
715 | (re-search-forward | |
716 | "\\='\\([[:space:][:alnum:]-()+,./:=?;!*#@$_%]*\\)'" | |
717 | nil t)) | |
6d12a4df | 718 | (error "XML: Missing Public ID")) |
f6fcdfff | 719 | (let ((pubid (match-string-no-properties 1))) |
6d12a4df | 720 | (skip-syntax-forward " ") |
a98e819b DL |
721 | (unless (or (re-search-forward "\\='\\([^']*\\)'" nil t) |
722 | (re-search-forward "\\=\"\\([^\"]*\\)\"" nil t)) | |
6d12a4df | 723 | (error "XML: Missing System ID")) |
f6fcdfff | 724 | (push (list pubid (match-string-no-properties 1) 'public) dtd))) |
a98e819b DL |
725 | ((looking-at "SYSTEM\\s-+") |
726 | (goto-char (match-end 0)) | |
727 | (unless (or (re-search-forward "\\='\\([^']*\\)'" nil t) | |
728 | (re-search-forward "\\=\"\\([^\"]*\\)\"" nil t)) | |
6d12a4df | 729 | (error "XML: Missing System ID")) |
f6fcdfff | 730 | (push (list (match-string-no-properties 1) 'system) dtd))) |
a98e819b | 731 | (skip-syntax-forward " ") |
7f3fbd5d CY |
732 | |
733 | (if (eq (char-after) ?>) | |
734 | ||
735 | ;; No internal subset | |
a98e819b | 736 | (forward-char) |
a98e819b | 737 | |
7f3fbd5d CY |
738 | ;; Internal subset (XML [28b]) |
739 | (unless (eq (char-after) ?\[) | |
740 | (error "XML: Bad DTD")) | |
741 | (forward-char) | |
742 | ||
fbf2e7ad CY |
743 | ;; [2.8]: "markup declarations may be made up in whole or in |
744 | ;; part of the replacement text of parameter entities." | |
745 | ||
746 | ;; Since parameter entities are valid only within the DTD, we | |
747 | ;; first search for the position of the next possible parameter | |
748 | ;; entity. Then, search for the next DTD element; if it ends | |
749 | ;; before the next parameter entity, expand the parameter entity | |
750 | ;; and try again. | |
751 | (setq next-parameter-entity | |
752 | (save-excursion | |
566df3fc | 753 | (if (re-search-forward xml-pe-reference-re nil t) |
fbf2e7ad CY |
754 | (match-beginning 0)))) |
755 | ||
7f3fbd5d CY |
756 | ;; Parse the rest of the DTD |
757 | ;; Fixme: Deal with NOTATION, PIs. | |
758 | (while (not (looking-at "\\s-*\\]")) | |
759 | (skip-syntax-forward " ") | |
760 | (cond | |
a76e6535 CY |
761 | ((eobp) |
762 | (error "XML: (Well-Formed) End of document while reading DTD")) | |
7f3fbd5d | 763 | ;; Element declaration [45]: |
6fe566a7 CY |
764 | ((and (looking-at (eval-when-compile |
765 | (concat "<!ELEMENT\\s-+\\(" xml-name-re | |
766 | "\\)\\s-+\\([^>]+\\)>"))) | |
fbf2e7ad CY |
767 | (or (null next-parameter-entity) |
768 | (<= (match-end 0) next-parameter-entity))) | |
7f3fbd5d CY |
769 | (let ((element (match-string-no-properties 1)) |
770 | (type (match-string-no-properties 2)) | |
771 | (end-pos (match-end 0))) | |
772 | ;; Translation of rule [46] of XML specifications | |
a98e819b | 773 | (cond |
6fe566a7 | 774 | ((string-match "\\`EMPTY\\s-*\\'" type) ; empty declaration |
a98e819b | 775 | (setq type 'empty)) |
6fe566a7 | 776 | ((string-match "\\`ANY\\s-*$" type) ; any type of contents |
a98e819b | 777 | (setq type 'any)) |
6fe566a7 CY |
778 | ((string-match "\\`(\\(.*\\))\\s-*\\'" type) ; children ([47]) |
779 | (setq type (xml-parse-elem-type | |
780 | (match-string-no-properties 1 type)))) | |
781 | ((string-match "^%[^;]+;[ \t\n\r]*\\'" type) ; substitution | |
a98e819b | 782 | nil) |
7f3fbd5d CY |
783 | (xml-validating-parser |
784 | (error "XML: (Validity) Invalid element type in the DTD"))) | |
27720433 | 785 | |
7f3fbd5d CY |
786 | ;; rule [45]: the element declaration must be unique |
787 | (and (assoc element dtd) | |
788 | xml-validating-parser | |
789 | (error "XML: (Validity) DTD element declarations must be unique (<%s>)" | |
790 | element)) | |
a98e819b DL |
791 | |
792 | ;; Store the element in the DTD | |
793 | (push (list element type) dtd) | |
7f3fbd5d CY |
794 | (goto-char end-pos))) |
795 | ||
796 | ;; Attribute-list declaration [52] (currently unsupported): | |
fbf2e7ad CY |
797 | ((and (looking-at (eval-when-compile |
798 | (concat "<!ATTLIST[ \t\n\r]*\\(" xml-name-re | |
799 | "\\)[ \t\n\r]*\\(" xml-att-def-re | |
800 | "\\)*[ \t\n\r]*>"))) | |
801 | (or (null next-parameter-entity) | |
802 | (<= (match-end 0) next-parameter-entity))) | |
7f3fbd5d CY |
803 | (goto-char (match-end 0))) |
804 | ||
fbf2e7ad | 805 | ;; Comments (skip to end, ignoring parameter entity): |
7f3fbd5d | 806 | ((looking-at "<!--") |
fbf2e7ad CY |
807 | (search-forward "-->") |
808 | (and next-parameter-entity | |
809 | (> (point) next-parameter-entity) | |
810 | (setq next-parameter-entity | |
811 | (save-excursion | |
566df3fc | 812 | (if (re-search-forward xml-pe-reference-re nil t) |
fbf2e7ad | 813 | (match-beginning 0)))))) |
7f3fbd5d CY |
814 | |
815 | ;; Internal entity declarations: | |
fbf2e7ad CY |
816 | ((and (looking-at (eval-when-compile |
817 | (concat "<!ENTITY[ \t\n\r]+\\(%[ \t\n\r]+\\)?\\(" | |
818 | xml-name-re "\\)[ \t\n\r]*\\(" | |
819 | xml-entity-value-re "\\)[ \t\n\r]*>"))) | |
820 | (or (null next-parameter-entity) | |
821 | (<= (match-end 0) next-parameter-entity))) | |
7f3fbd5d CY |
822 | (let* ((name (prog1 (match-string-no-properties 2) |
823 | (goto-char (match-end 0)))) | |
824 | (alist (if (match-string 1) | |
825 | 'xml-parameter-entity-alist | |
826 | 'xml-entity-alist)) | |
827 | ;; Retrieve the deplacement text: | |
828 | (value (xml--entity-replacement-text | |
829 | ;; Entity value, sans quotation marks: | |
830 | (substring (match-string-no-properties 3) 1 -1)))) | |
831 | ;; If the same entity is declared more than once, the | |
832 | ;; first declaration is binding. | |
833 | (unless (assoc name (symbol-value alist)) | |
834 | (set alist (cons (cons name value) (symbol-value alist)))))) | |
835 | ||
836 | ;; External entity declarations (currently unsupported): | |
fbf2e7ad CY |
837 | ((and (or (looking-at (eval-when-compile |
838 | (concat "<!ENTITY[ \t\n\r]+\\(%[ \t\n\r]+\\)?\\(" | |
839 | xml-name-re "\\)[ \t\n\r]+SYSTEM[ \t\n\r]+" | |
840 | "\\(\"[^\"]*\"\\|'[^']*'\\)[ \t\n\r]*>"))) | |
841 | (looking-at (eval-when-compile | |
842 | (concat "<!ENTITY[ \t\n\r]+\\(%[ \t\n\r]+\\)?\\(" | |
843 | xml-name-re "\\)[ \t\n\r]+PUBLIC[ \t\n\r]+" | |
844 | "\"[- \r\na-zA-Z0-9'()+,./:=?;!*#@$_%]*\"" | |
845 | "\\|'[- \r\na-zA-Z0-9()+,./:=?;!*#@$_%]*'" | |
846 | "[ \t\n\r]+\\(\"[^\"]*\"\\|'[^']*'\\)" | |
847 | "[ \t\n\r]*>")))) | |
848 | (or (null next-parameter-entity) | |
849 | (<= (match-end 0) next-parameter-entity))) | |
7f3fbd5d CY |
850 | (goto-char (match-end 0))) |
851 | ||
fbf2e7ad CY |
852 | ;; If a parameter entity is in the way, expand it. |
853 | (next-parameter-entity | |
854 | (save-excursion | |
855 | (goto-char next-parameter-entity) | |
566df3fc | 856 | (unless (looking-at xml-pe-reference-re) |
fbf2e7ad CY |
857 | (error "XML: Internal error")) |
858 | (let* ((entity (match-string 1)) | |
859 | (beg (point-marker)) | |
860 | (elt (assoc entity xml-parameter-entity-alist))) | |
861 | (if elt | |
862 | (progn | |
863 | (replace-match (cdr elt) t t) | |
864 | ;; The replacement can itself be a parameter entity. | |
865 | (goto-char next-parameter-entity)) | |
866 | (goto-char (match-end 0)))) | |
867 | (setq next-parameter-entity | |
566df3fc | 868 | (if (re-search-forward xml-pe-reference-re nil t) |
fbf2e7ad | 869 | (match-beginning 0))))) |
7f3fbd5d | 870 | |
a76e6535 | 871 | ;; Anything else is garbage (ignored if not validating). |
7f3fbd5d | 872 | (xml-validating-parser |
a76e6535 CY |
873 | (error "XML: (Validity) Invalid DTD item")) |
874 | (t | |
875 | (skip-chars-forward "^]")))) | |
7f3fbd5d | 876 | |
6d12a4df | 877 | (if (looking-at "\\s-*]>") |
23d519e4 | 878 | (goto-char (match-end 0)))) |
461f3ad0 | 879 | (nreverse dtd))) |
47db06aa | 880 | |
7f3fbd5d CY |
881 | (defun xml--entity-replacement-text (string) |
882 | "Return the replacement text for the entity value STRING. | |
883 | The replacement text is obtained by replacing character | |
884 | references and parameter-entity references." | |
b3218de1 CY |
885 | (let ((ref-re (eval-when-compile |
886 | (concat "\\(?:&#\\([0-9]+\\)\\|&#x\\([0-9a-fA-F]+\\)\\|%\\(" | |
887 | xml-name-re "\\)\\);"))) | |
7f3fbd5d CY |
888 | children) |
889 | (while (string-match ref-re string) | |
890 | (push (substring string 0 (match-beginning 0)) children) | |
891 | (let ((remainder (substring string (match-end 0))) | |
892 | ref val) | |
893 | (cond ((setq ref (match-string 1 string)) | |
894 | ;; Decimal character reference | |
895 | (setq val (decode-char 'ucs (string-to-number ref))) | |
896 | (if val (push (string val) children))) | |
897 | ;; Hexadecimal character reference | |
898 | ((setq ref (match-string 2 string)) | |
899 | (setq val (decode-char 'ucs (string-to-number ref 16))) | |
900 | (if val (push (string val) children))) | |
901 | ;; Parameter entity reference | |
902 | ((setq ref (match-string 3 string)) | |
903 | (setq val (assoc ref xml-parameter-entity-alist)) | |
a7aef6f5 CY |
904 | (and (null val) |
905 | xml-validating-parser | |
906 | (error "XML: (Validity) Undefined parameter entity `%s'" ref)) | |
907 | (push (or (cdr val) xml-undefined-entity) children))) | |
7f3fbd5d CY |
908 | (setq string remainder))) |
909 | (mapconcat 'identity (nreverse (cons string children)) ""))) | |
910 | ||
47db06aa | 911 | (defun xml-parse-elem-type (string) |
a98e819b | 912 | "Convert element type STRING into a Lisp structure." |
47db06aa GM |
913 | |
914 | (let (elem modifier) | |
915 | (if (string-match "(\\([^)]+\\))\\([+*?]?\\)" string) | |
916 | (progn | |
f6fcdfff CY |
917 | (setq elem (match-string-no-properties 1 string) |
918 | modifier (match-string-no-properties 2 string)) | |
47db06aa | 919 | (if (string-match "|" elem) |
971489ea | 920 | (setq elem (cons 'choice |
47db06aa GM |
921 | (mapcar 'xml-parse-elem-type |
922 | (split-string elem "|")))) | |
923 | (if (string-match "," elem) | |
971489ea | 924 | (setq elem (cons 'seq |
47db06aa | 925 | (mapcar 'xml-parse-elem-type |
a98e819b | 926 | (split-string elem ","))))))) |
a158ff81 | 927 | (if (string-match "[ \t\n\r]*\\([^+*?]+\\)\\([+*?]?\\)" string) |
f6fcdfff CY |
928 | (setq elem (match-string-no-properties 1 string) |
929 | modifier (match-string-no-properties 2 string)))) | |
47db06aa | 930 | |
971489ea SM |
931 | (if (and (stringp elem) (string= elem "#PCDATA")) |
932 | (setq elem 'pcdata)) | |
524425ae | 933 | |
971489ea SM |
934 | (cond |
935 | ((string= modifier "+") | |
936 | (list '+ elem)) | |
937 | ((string= modifier "*") | |
938 | (list '* elem)) | |
939 | ((string= modifier "?") | |
0fa6f70c | 940 | (list '\? elem)) |
971489ea SM |
941 | (t |
942 | elem)))) | |
47db06aa | 943 | |
a7aef6f5 | 944 | ;;; Substituting special XML sequences |
47db06aa GM |
945 | |
946 | (defun xml-substitute-special (string) | |
a7aef6f5 CY |
947 | "Return STRING, after substituting entity and character references. |
948 | STRING is assumed to occur in an XML attribute value." | |
566df3fc | 949 | (let ((strlen (length string)) |
a7aef6f5 | 950 | children) |
566df3fc | 951 | (while (string-match xml-entity-or-char-ref-re string) |
a7aef6f5 CY |
952 | (push (substring string 0 (match-beginning 0)) children) |
953 | (let* ((remainder (substring string (match-end 0))) | |
566df3fc CY |
954 | (is-hex (match-string 1 string)) ; Is it a hex numeric reference? |
955 | (ref (match-string 2 string))) ; Numeric part of reference | |
a7aef6f5 CY |
956 | (if ref |
957 | ;; [4.6] Character references are included as | |
958 | ;; character data. | |
566df3fc | 959 | (let ((val (decode-char 'ucs (string-to-number ref (if is-hex 16))))) |
a7aef6f5 CY |
960 | (push (cond (val (string val)) |
961 | (xml-validating-parser | |
962 | (error "XML: (Validity) Undefined character `x%s'" ref)) | |
963 | (t xml-undefined-entity)) | |
964 | children) | |
a76e6535 CY |
965 | (setq string remainder |
966 | strlen (length string))) | |
a7aef6f5 CY |
967 | ;; [4.4.5] Entity references are "included in literal". |
968 | ;; Note that we don't need do anything special to treat | |
969 | ;; quotes as normal data characters. | |
566df3fc | 970 | (setq ref (match-string 3 string)) ; entity name |
a7aef6f5 CY |
971 | (let ((val (or (cdr (assoc ref xml-entity-alist)) |
972 | (if xml-validating-parser | |
973 | (error "XML: (Validity) Undefined entity `%s'" ref) | |
974 | xml-undefined-entity)))) | |
a76e6535 CY |
975 | (setq string (concat val remainder))) |
976 | (and xml-entity-expansion-limit | |
977 | (> (length string) (+ strlen xml-entity-expansion-limit)) | |
978 | (error "XML: Passed `xml-entity-expansion-limit' while expanding `&%s;'" | |
979 | ref))))) | |
a7aef6f5 | 980 | (mapconcat 'identity (nreverse (cons string children)) ""))) |
a3110b5d | 981 | |
571855b6 UJ |
982 | (defun xml-substitute-numeric-entities (string) |
983 | "Substitute SGML numeric entities by their respective utf characters. | |
984 | This function replaces numeric entities in the input STRING and | |
985 | returns the modified string. For example \"*\" gets replaced | |
986 | by \"*\"." | |
987 | (if (and string (stringp string)) | |
988 | (let ((start 0)) | |
989 | (while (string-match "&#\\([0-9]+\\);" string start) | |
990 | (condition-case nil | |
991 | (setq string (replace-match | |
992 | (string (read (substring string | |
993 | (match-beginning 1) | |
994 | (match-end 1)))) | |
995 | nil nil string)) | |
996 | (error nil)) | |
997 | (setq start (1+ (match-beginning 0)))) | |
998 | string) | |
999 | nil)) | |
1000 | ||
a7aef6f5 | 1001 | ;;; Printing a parse tree (mainly for debugging). |
47db06aa | 1002 | |
27240aa4 AS |
1003 | (defun xml-debug-print (xml &optional indent-string) |
1004 | "Outputs the XML in the current buffer. | |
1005 | XML can be a tree or a list of nodes. | |
1006 | The first line is indented with the optional INDENT-STRING." | |
1007 | (setq indent-string (or indent-string "")) | |
971489ea | 1008 | (dolist (node xml) |
27240aa4 AS |
1009 | (xml-debug-print-internal node indent-string))) |
1010 | ||
1011 | (defalias 'xml-print 'xml-debug-print) | |
47db06aa | 1012 | |
7731c9f4 | 1013 | (defun xml-escape-string (string) |
17975d7f CY |
1014 | "Convert STRING into a string containing valid XML character data. |
1015 | Replace occurrences of &<>'\" in STRING with their default XML | |
1016 | entity references (e.g. replace each & with &). | |
1017 | ||
1018 | XML character data must not contain & or < characters, nor the > | |
1019 | character under some circumstances. The XML spec does not impose | |
1020 | restriction on \" or ', but we just substitute for these too | |
1021 | \(as is permitted by the spec)." | |
1022 | (with-temp-buffer | |
1023 | (insert string) | |
1024 | (dolist (substitution '(("&" . "&") | |
1025 | ("<" . "<") | |
1026 | (">" . ">") | |
1027 | ("'" . "'") | |
1028 | ("\"" . """))) | |
1029 | (goto-char (point-min)) | |
1030 | (while (search-forward (car substitution) nil t) | |
1031 | (replace-match (cdr substitution) t t nil))) | |
1032 | (buffer-string))) | |
7731c9f4 | 1033 | |
971489ea | 1034 | (defun xml-debug-print-internal (xml indent-string) |
47db06aa | 1035 | "Outputs the XML tree in the current buffer. |
a98e819b | 1036 | The first line is indented with INDENT-STRING." |
47db06aa GM |
1037 | (let ((tree xml) |
1038 | attlist) | |
a98e819b | 1039 | (insert indent-string ?< (symbol-name (xml-node-name tree))) |
524425ae | 1040 | |
47db06aa | 1041 | ;; output the attribute list |
971489ea | 1042 | (setq attlist (xml-node-attributes tree)) |
47db06aa | 1043 | (while attlist |
7731c9f4 MH |
1044 | (insert ?\ (symbol-name (caar attlist)) "=\"" |
1045 | (xml-escape-string (cdar attlist)) ?\") | |
971489ea | 1046 | (setq attlist (cdr attlist))) |
524425ae | 1047 | |
971489ea | 1048 | (setq tree (xml-node-children tree)) |
47db06aa | 1049 | |
27240aa4 AS |
1050 | (if (null tree) |
1051 | (insert ?/ ?>) | |
1052 | (insert ?>) | |
1053 | ||
1054 | ;; output the children | |
1055 | (dolist (node tree) | |
1056 | (cond | |
1057 | ((listp node) | |
1058 | (insert ?\n) | |
1059 | (xml-debug-print-internal node (concat indent-string " "))) | |
7731c9f4 MH |
1060 | ((stringp node) |
1061 | (insert (xml-escape-string node))) | |
27240aa4 AS |
1062 | (t |
1063 | (error "Invalid XML tree")))) | |
1064 | ||
1065 | (when (not (and (null (cdr tree)) | |
1066 | (stringp (car tree)))) | |
1067 | (insert ?\n indent-string)) | |
1068 | (insert ?< ?/ (symbol-name (xml-node-name xml)) ?>)))) | |
47db06aa GM |
1069 | |
1070 | (provide 'xml) | |
1071 | ||
1072 | ;;; xml.el ends here |