xml.el 32.9 KB
Newer Older
1
;;; xml.el --- XML parser
Gerd Moellmann's avatar
Gerd Moellmann committed
2

3
;; Copyright (C) 2000, 2001, 2002, 2003, 2004,
Glenn Morris's avatar
Glenn Morris committed
4
;;   2005, 2006, 2007, 2008 Free Software Foundation, Inc.
Gerd Moellmann's avatar
Gerd Moellmann committed
5 6

;; Author: Emmanuel Briot  <briot@gnat.com>
Richard M. Stallman's avatar
Richard M. Stallman committed
7
;; Maintainer: Mark A. Hershberger <mah@everybody.org>
Dave Love's avatar
Dave Love committed
8
;; Keywords: xml, data
Gerd Moellmann's avatar
Gerd Moellmann committed
9 10 11

;; This file is part of GNU Emacs.

12
;; GNU Emacs is free software: you can redistribute it and/or modify
Gerd Moellmann's avatar
Gerd Moellmann committed
13
;; it under the terms of the GNU General Public License as published by
14 15
;; the Free Software Foundation, either version 3 of the License, or
;; (at your option) any later version.
Gerd Moellmann's avatar
Gerd Moellmann committed
16 17 18 19 20 21 22

;; GNU Emacs is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
;; GNU General Public License for more details.

;; You should have received a copy of the GNU General Public License
23
;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
Gerd Moellmann's avatar
Gerd Moellmann committed
24 25 26

;;; Commentary:

Dave Love's avatar
Dave Love committed
27 28
;; This file contains a somewhat incomplete non-validating XML parser.  It
;; parses a file, and returns a list that can be used internally by
Stefan Monnier's avatar
Stefan Monnier committed
29
;; any other Lisp libraries.
Gerd Moellmann's avatar
Gerd Moellmann committed
30 31 32

;;; FILE FORMAT

Dave Love's avatar
Dave Love committed
33 34
;; The document type declaration may either be ignored or (optionally)
;; parsed, but currently the parsing will only accept element
Stefan Monnier's avatar
Stefan Monnier committed
35
;; declarations.  The XML file is assumed to be well-formed.  In case
Dave Love's avatar
Dave Love committed
36 37
;; of error, the parsing stops and the XML file is shown where the
;; parsing stopped.
Gerd Moellmann's avatar
Gerd Moellmann committed
38
;;
Dave Love's avatar
Dave Love committed
39
;; It also knows how to ignore comments and processing instructions.
Gerd Moellmann's avatar
Gerd Moellmann committed
40 41
;;
;; The XML file should have the following format:
42 43 44
;;    <node1 attr1="name1" attr2="name2" ...>value
;;       <node2 attr3="name3" attr4="name4">value2</node2>
;;       <node3 attr5="name5" attr6="name6">value3</node3>
Gerd Moellmann's avatar
Gerd Moellmann committed
45
;;    </node1>
Stefan Monnier's avatar
Stefan Monnier committed
46
;; Of course, the name of the nodes and attributes can be anything.  There can
Gerd Moellmann's avatar
Gerd Moellmann committed
47 48 49 50 51 52 53
;; be any number of attributes (or none), as well as any number of children
;; below the nodes.
;;
;; There can be only top level node, but with any number of children below.

;;; LIST FORMAT

54 55
;; The functions `xml-parse-file', `xml-parse-region' and
;; `xml-parse-tag' return a list with the following format:
Gerd Moellmann's avatar
Gerd Moellmann committed
56 57
;;
;;    xml-list   ::= (node node ...)
58
;;    node       ::= (qname attribute-list . child_node_list)
Gerd Moellmann's avatar
Gerd Moellmann committed
59 60
;;    child_node_list ::= child_node child_node ...
;;    child_node ::= node | string
61 62
;;    qname      ::= (:namespace-uri . "name") | "name"
;;    attribute_list ::= ((qname . "value") (qname . "value") ...)
Gerd Moellmann's avatar
Gerd Moellmann committed
63 64 65
;;                       | nil
;;    string     ::= "..."
;;
Dave Love's avatar
Dave Love committed
66 67 68
;; Some macros are provided to ease the parsing of this list.
;; Whitespace is preserved.  Fixme: There should be a tree-walker that
;; can remove it.
Gerd Moellmann's avatar
Gerd Moellmann committed
69

70 71 72 73 74
;; TODO:
;;  * xml:base, xml:space support
;;  * more complete DOCTYPE parsing
;;  * pi support

Gerd Moellmann's avatar
Gerd Moellmann committed
75 76
;;; Code:

77 78 79 80 81 82
;; Note that buffer-substring and match-string were formerly used in
;; several places, because the -no-properties variants remove
;; composition info.  However, after some discussion on emacs-devel,
;; the consensus was that the speed of the -no-properties variants was
;; a worthwhile tradeoff especially since we're usually parsing files
;; instead of hand-crafted XML.
Dave Love's avatar
Dave Love committed
83

Gerd Moellmann's avatar
Gerd Moellmann committed
84 85 86 87 88 89
;;*******************************************************************
;;**
;;**  Macros to parse the list
;;**
;;*******************************************************************

90 91 92
(defconst xml-undefined-entity "?"
  "What to substitute for undefined entities")

93 94 95 96 97 98 99 100 101 102 103 104 105 106
(defvar xml-entity-alist
  '(("lt"   . "<")
    ("gt"   . ">")
    ("apos" . "'")
    ("quot" . "\"")
    ("amp"  . "&"))
  "The defined entities.  Entities are added to this when the DTD is parsed.")

(defvar xml-sub-parser nil
  "Dynamically set this to a non-nil value if you want to parse an XML fragment.")

(defvar xml-validating-parser nil
  "Set to non-nil to get validity checking.")

107
(defsubst xml-node-name (node)
Gerd Moellmann's avatar
Gerd Moellmann committed
108
  "Return the tag associated with NODE.
Stefan Monnier's avatar
Stefan Monnier committed
109 110 111 112 113 114 115 116 117 118 119 120
Without namespace-aware parsing, the tag is a symbol.

With namespace-aware parsing, the tag is a cons of a string
representing the uri of the namespace with the local name of the
tag.  For example,

    <foo>

would be represented by

    '(\"\" . \"foo\")."

121
  (car node))
Gerd Moellmann's avatar
Gerd Moellmann committed
122

123
(defsubst xml-node-attributes (node)
Gerd Moellmann's avatar
Gerd Moellmann committed
124 125
  "Return the list of attributes of NODE.
The list can be nil."
126
  (nth 1 node))
Gerd Moellmann's avatar
Gerd Moellmann committed
127

128
(defsubst xml-node-children (node)
Gerd Moellmann's avatar
Gerd Moellmann committed
129 130
  "Return the list of children of NODE.
This is a list of nodes, and it can be nil."
131
  (cddr node))
Gerd Moellmann's avatar
Gerd Moellmann committed
132 133 134

(defun xml-get-children (node child-name)
  "Return the children of NODE whose tag is CHILD-NAME.
Stefan Monnier's avatar
Stefan Monnier committed
135
CHILD-NAME should match the value returned by `xml-node-name'."
136 137
  (let ((match ()))
    (dolist (child (xml-node-children node))
Stefan Monnier's avatar
Stefan Monnier committed
138 139 140
      (if (and (listp child)
               (equal (xml-node-name child) child-name))
          (push child match)))
141
    (nreverse match)))
Gerd Moellmann's avatar
Gerd Moellmann committed
142

143
(defun xml-get-attribute-or-nil (node attribute)
Gerd Moellmann's avatar
Gerd Moellmann committed
144
  "Get from NODE the value of ATTRIBUTE.
Stefan Monnier's avatar
Stefan Monnier committed
145
Return nil if the attribute was not found.
146 147

See also `xml-get-attribute'."
148
  (cdr (assoc attribute (xml-node-attributes node))))
149 150 151 152 153 154 155

(defsubst xml-get-attribute (node attribute)
  "Get from NODE the value of ATTRIBUTE.
An empty string is returned if the attribute was not found.

See also `xml-get-attribute-or-nil'."
  (or (xml-get-attribute-or-nil node attribute) ""))
Gerd Moellmann's avatar
Gerd Moellmann committed
156 157 158 159 160 161 162

;;*******************************************************************
;;**
;;**  Creating the list
;;**
;;*******************************************************************

Dave Love's avatar
Dave Love committed
163
;;;###autoload
164
(defun xml-parse-file (file &optional parse-dtd parse-ns)
Dave Love's avatar
Dave Love committed
165 166
  "Parse the well-formed XML file FILE.
If FILE is already visited, use its buffer and don't kill it.
Gerd Moellmann's avatar
Gerd Moellmann committed
167
Returns the top node with all its children.
168 169
If PARSE-DTD is non-nil, the DTD is parsed rather than skipped.
If PARSE-NS is non-nil, then QNAMES are expanded."
170 171 172 173 174 175 176 177 178 179 180 181 182
  (if (get-file-buffer file)
      (with-current-buffer (get-file-buffer file)
	(save-excursion
	  (xml-parse-region (point-min)
			    (point-max)
			    (current-buffer)
			    parse-dtd parse-ns)))
    (with-temp-buffer
      (insert-file-contents file)
      (xml-parse-region (point-min)
			(point-max)
			(current-buffer)
			parse-dtd parse-ns))))
Gerd Moellmann's avatar
Gerd Moellmann committed
183

184

185 186
(defvar xml-name-re)
(defvar xml-entity-value-re)
187
(defvar xml-att-def-re)
188
(let* ((start-chars (concat "[:alpha:]:_"))
189
       (name-chars  (concat "-[:digit:]." start-chars))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
190
       ;;[3]   	S	   ::=   	(#x20 | #x9 | #xD | #xA)+
191
       (whitespace  "[ \t\n\r]"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
192 193 194 195
  ;;[4] NameStartChar ::= ":" | [A-Z] | "_" | [a-z] | [#xC0-#xD6] 
  ;;                      | [#xD8-#xF6] | [#xF8-#x2FF] | [#x370-#x37D] | [#x37F-#x1FFF]
  ;;                      | [#x200C-#x200D] | [#x2070-#x218F] | [#x2C00-#x2FEF] | [#x3001-#xD7FF]
  ;;                      | [#xF900-#xFDCF] | [#xFDF0-#xFFFD] | [#x10000-#xEFFFF]
196
  (defvar xml-name-start-char-re (concat "[" start-chars "]"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
197
  ;;[4a] NameChar	::= NameStartChar | "-" | "." | [0-9] | #xB7 | [#x0300-#x036F] | [#x203F-#x2040]
198
  (defvar xml-name-char-re       (concat "[" name-chars  "]"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
199
  ;;[5] Name     ::= NameStartChar (NameChar)*
200
  (defvar xml-name-re            (concat xml-name-start-char-re xml-name-char-re "*"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
201
  ;;[6] Names    ::= Name (#x20 Name)*
202
  (defvar xml-names-re           (concat xml-name-re "\\(?: " xml-name-re "\\)*"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
203
  ;;[7] Nmtoken ::= (NameChar)+
204
  (defvar xml-nmtoken-re         (concat xml-name-char-re "+"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
205
  ;;[8] Nmtokens ::= Nmtoken (#x20 Nmtoken)*
206
  (defvar xml-nmtokens-re        (concat xml-nmtoken-re "\\(?: " xml-name-re "\\)*"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
207
  ;;[66] CharRef ::= '&#' [0-9]+ ';' | '&#x' [0-9a-fA-F]+ ';'
208
  (defvar xml-char-ref-re        "\\(?:&#[0-9]+;\\|&#x[0-9a-fA-F]+;\\)")
Mark A. Hershberger's avatar
Mark A. Hershberger committed
209
  ;;[68] EntityRef   ::= '&' Name ';'
210
  (defvar xml-entity-ref         (concat "&" xml-name-re ";"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
211
  ;;[69] PEReference ::= '%' Name ';'
212
  (defvar xml-pe-reference-re    (concat "%" xml-name-re ";"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
213
  ;;[67] Reference   ::= EntityRef | CharRef
214
  (defvar xml-reference-re       (concat "\\(?:" xml-entity-ref "\\|" xml-char-ref-re "\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
215
  ;;[10]   	AttValue	   ::=   	'"' ([^<&"] | Reference)* '"' |  "'" ([^<&'] | Reference)* "'"
216 217
  (defvar xml-att-value-re    (concat "\\(?:\"\\(?:[^&\"]\\|" xml-reference-re "\\)*\"\\|"
				      "'\\(?:[^&']\\|" xml-reference-re "\\)*'\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
218 219 220 221 222 223 224
  ;;[56]   	TokenizedType	   ::=   	'ID'	   [VC: ID] [VC: One ID per Element Type] [VC: ID Attribute Default]
  ;;                                            | 'IDREF'    [VC: IDREF]
  ;;                             	              | 'IDREFS'   [VC: IDREF]
  ;;                                            | 'ENTITY'   [VC: Entity Name]
  ;;                                            | 'ENTITIES' [VC: Entity Name]
  ;;                                            | 'NMTOKEN'  [VC: Name Token]
  ;;                                            | 'NMTOKENS' [VC: Name Token]
225
  (defvar xml-tokenized-type-re "\\(?:ID\\|IDREF\\|IDREFS\\|ENTITY\\|ENTITIES\\|NMTOKEN\\|NMTOKENS\\)")
Mark A. Hershberger's avatar
Mark A. Hershberger committed
226
  ;;[58]   	NotationType	   ::=   	'NOTATION' S '(' S? Name (S? '|' S? Name)* S? ')'
227 228
  (defvar xml-notation-type-re (concat "\\(?:NOTATION" whitespace "(" whitespace "*" xml-name-re
				       "\\(?:" whitespace "*|" whitespace "*" xml-name-re "\\)*" whitespace "*)\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
229
  ;;[59]   	Enumeration	   ::=   	'(' S? Nmtoken (S? '|' S? Nmtoken)* S? ')'	[VC: Enumeration] [VC: No Duplicate Tokens]
230 231 232
  (defvar xml-enumeration-re (concat "\\(?:(" whitespace "*" xml-nmtoken-re 
				     "\\(?:" whitespace "*|" whitespace "*" xml-nmtoken-re "\\)*"
				     whitespace ")\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
233
  ;;[57]   	EnumeratedType	   ::=   	NotationType | Enumeration
234
  (defvar xml-enumerated-type-re (concat "\\(?:" xml-notation-type-re "\\|" xml-enumeration-re "\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
235 236
  ;;[54]   	AttType	   ::=   	StringType | TokenizedType | EnumeratedType
  ;;[55]   	StringType	   ::=   	'CDATA'
237
  (defvar xml-att-type-re (concat "\\(?:CDATA\\|" xml-tokenized-type-re "\\|" xml-notation-type-re"\\|" xml-enumerated-type-re "\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
238
  ;;[60]   	DefaultDecl	   ::=   	'#REQUIRED' | '#IMPLIED' | (('#FIXED' S)? AttValue)
239
  (defvar xml-default-decl-re (concat "\\(?:#REQUIRED\\|#IMPLIED\\|\\(?:#FIXED" whitespace "\\)*" xml-att-value-re "\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
240
  ;;[53]   	AttDef	   ::=   	S Name S AttType S DefaultDecl
241 242 243
  (defvar xml-att-def-re         (concat "\\(?:" whitespace "*" xml-name-re
					 whitespace "*" xml-att-type-re
					 whitespace "*" xml-default-decl-re "\\)"))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
244 245
  ;;[9] EntityValue ::= '"' ([^%&"] | PEReference | Reference)* '"'
  ;;		   |  "'" ([^%&'] | PEReference | Reference)* "'"
246 247 248 249 250 251 252 253 254 255 256 257
  (defvar xml-entity-value-re    (concat "\\(?:\"\\(?:[^%&\"]\\|" xml-pe-reference-re
					 "\\|" xml-reference-re "\\)*\"\\|'\\(?:[^%&']\\|"
					 xml-pe-reference-re "\\|" xml-reference-re "\\)*'\\)")))
;;[75] ExternalID ::= 'SYSTEM' S SystemLiteral
;;                 | 'PUBLIC' S PubidLiteral S SystemLiteral
;;[76] NDataDecl ::=   	S 'NDATA' S 
;;[73] EntityDef  ::= EntityValue| (ExternalID NDataDecl?)
;;[71] GEDecl     ::= '<!ENTITY' S Name S EntityDef S? '>'
;;[74] PEDef      ::= EntityValue | ExternalID
;;[72] PEDecl     ::= '<!ENTITY' S '%' S Name S PEDef S? '>'
;;[70] EntityDecl ::= GEDecl | PEDecl

Dave Love's avatar
Dave Love committed
258 259 260 261 262 263 264 265 266 267 268 269 270
;; Note that this is setup so that we can do whitespace-skipping with
;; `(skip-syntax-forward " ")', inter alia.  Previously this was slow
;; compared with `re-search-forward', but that has been fixed.  Also
;; note that the standard syntax table contains other characters with
;; whitespace syntax, like NBSP, but they are invalid in contexts in
;; which we might skip whitespace -- specifically, they're not
;; NameChars [XML 4].

(defvar xml-syntax-table
  (let ((table (make-syntax-table)))
    ;; Get space syntax correct per XML [3].
    (dotimes (c 31)
      (modify-syntax-entry c "." table)) ; all are space in standard table
Mark A. Hershberger's avatar
Mark A. Hershberger committed
271
    (dolist (c '(?\t ?\n ?\r))		 ; these should be space
Dave Love's avatar
Dave Love committed
272 273 274 275 276 277 278 279 280
      (modify-syntax-entry c " " table))
    ;; For skipping attributes.
    (modify-syntax-entry ?\" "\"" table)
    (modify-syntax-entry ?' "\"" table)
    ;; Non-alnum name chars should be symbol constituents (`-' and `_'
    ;; are OK by default).
    (modify-syntax-entry ?. "_" table)
    (modify-syntax-entry ?: "_" table)
    ;; XML [89]
Mark A. Hershberger's avatar
Mark A. Hershberger committed
281 282 283 284 285
    (unless (featurep 'xemacs)
      (dolist (c '(#x00B7 #x02D0 #x02D1 #x0387 #x0640 #x0E46 #x0EC6 #x3005
			  #x3031 #x3032 #x3033 #x3034 #x3035 #x309D #x309E #x30FC
			  #x30FD #x30FE))
	(modify-syntax-entry (decode-char 'ucs c) "w" table)))
Dave Love's avatar
Dave Love committed
286 287 288 289 290 291
    ;; Fixme: rest of [4]
    table)
  "Syntax table used by `xml-parse-region'.")

;; XML [5]
;; Note that [:alpha:] matches all multibyte chars with word syntax.
292 293
(eval-and-compile
  (defconst xml-name-regexp "[[:alpha:]_:][[:alnum:]._:-]*"))
Dave Love's avatar
Dave Love committed
294 295 296 297 298 299

;; Fixme:  This needs re-writing to deal with the XML grammar properly, i.e.
;;   document    ::=    prolog element Misc*
;;   prolog    ::=    XMLDecl? Misc* (doctypedecl Misc*)?

;;;###autoload
300
(defun xml-parse-region (beg end &optional buffer parse-dtd parse-ns)
Gerd Moellmann's avatar
Gerd Moellmann committed
301 302 303
  "Parse the region from BEG to END in BUFFER.
If BUFFER is nil, it defaults to the current buffer.
Returns the XML list for the region, or raises an error if the region
304
is not well-formed XML.
Gerd Moellmann's avatar
Gerd Moellmann committed
305
If PARSE-DTD is non-nil, the DTD is parsed rather than skipped,
306 307
and returned as the first element of the list.
If PARSE-NS is non-nil, then QNAMES are expanded."
Mark A. Hershberger's avatar
Mark A. Hershberger committed
308 309 310 311 312 313 314 315 316 317
  ;; Use fixed syntax table to ensure regexp char classes and syntax
  ;; specs DTRT.
  (with-syntax-table (standard-syntax-table)
    (let ((case-fold-search nil)	; XML is case-sensitive.
 	  xml result dtd)
      (save-excursion
 	(if buffer
 	    (set-buffer buffer))
 	(save-restriction
 	  (narrow-to-region beg end)
Dave Love's avatar
Dave Love committed
318 319 320 321 322
	  (goto-char (point-min))
	  (while (not (eobp))
	    (if (search-forward "<" nil t)
		(progn
		  (forward-char -1)
323
		  (setq result (xml-parse-tag parse-dtd parse-ns))
324
		  (if (and xml result (not xml-sub-parser))
Dave Love's avatar
Dave Love committed
325
		      ;;  translation of rule [1] of XML specifications
326
		      (error "XML: (Not Well-Formed) Only one root tag allowed")
Gerd Moellmann's avatar
Gerd Moellmann committed
327
		    (cond
328
		     ((null result))
329 330
		     ((and (listp (car result))
			   parse-dtd)
331
		      (setq dtd (car result))
Dave Love's avatar
Dave Love committed
332 333
		      (if (cdr result)	; possible leading comment
			  (add-to-list 'xml (cdr result))))
Gerd Moellmann's avatar
Gerd Moellmann committed
334
		     (t
Dave Love's avatar
Dave Love committed
335 336 337 338 339
		      (add-to-list 'xml result)))))
	      (goto-char (point-max))))
	  (if parse-dtd
	      (cons dtd (nreverse xml))
	    (nreverse xml)))))))
Gerd Moellmann's avatar
Gerd Moellmann committed
340

341
(defun xml-maybe-do-ns (name default xml-ns)
Stefan Monnier's avatar
Stefan Monnier committed
342 343
  "Perform any namespace expansion.
NAME is the name to perform the expansion on.
344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
DEFAULT is the default namespace.  XML-NS is a cons of namespace
names to uris.  When namespace-aware parsing is off, then XML-NS
is nil.

During namespace-aware parsing, any name without a namespace is
put into the namespace identified by DEFAULT.  nil is used to
specify that the name shouldn't be given a namespace."
  (if (consp xml-ns)
      (let* ((nsp (string-match ":" name))
	     (lname (if nsp (substring name (match-end 0)) name))
	     (prefix (if nsp (substring name 0 (match-beginning 0)) default))
	     (special (and (string-equal lname "xmlns") (not prefix)))
             ;; Setting default to nil will insure that there is not
             ;; matching cons in xml-ns.  In which case we
	     (ns (or (cdr (assoc (if special "xmlns" prefix)
                                 xml-ns))
360
                     "")))
361 362
        (cons ns (if special "" lname)))
    (intern name)))
Gerd Moellmann's avatar
Gerd Moellmann committed
363

364 365 366 367 368 369 370 371 372 373 374 375 376 377
(defun xml-parse-fragment (&optional parse-dtd parse-ns)
  "Parse xml-like fragments."
  (let ((xml-sub-parser t)
	children)
    (while (not (eobp))
      (let ((bit (xml-parse-tag
		  parse-dtd parse-ns)))
	(if children
	    (setq children (append (list bit) children))
	  (if (stringp bit)
	      (setq children (list bit))
	    (setq children bit)))))
    (reverse children)))

378
(defun xml-parse-tag (&optional parse-dtd parse-ns)
Dave Love's avatar
Dave Love committed
379
  "Parse the tag at point.
Gerd Moellmann's avatar
Gerd Moellmann committed
380 381
If PARSE-DTD is non-nil, the DTD of the document, if any, is parsed and
returned as the first element in the list.
382
If PARSE-NS is non-nil, then QNAMES are expanded.
Gerd Moellmann's avatar
Gerd Moellmann committed
383
Returns one of:
Dave Love's avatar
Dave Love committed
384 385 386
 - a list : the matching node
 - nil    : the point is not looking at a tag.
 - a pair : the first element is the DTD, the second is the node."
387 388
  (let ((xml-validating-parser (or parse-dtd xml-validating-parser))
	(xml-ns (if (consp parse-ns)
389 390 391
		    parse-ns
		  (if parse-ns
		      (list
Mark A. Hershberger's avatar
Mark A. Hershberger committed
392
		       ;; Default for empty prefix is no namespace
393
		       (cons ""      "")
394
		       ;; "xml" namespace
395
		       (cons "xml"   "http://www.w3.org/XML/1998/namespace")
396
		       ;; We need to seed the xmlns namespace
397
		       (cons "xmlns" "http://www.w3.org/2000/xmlns/"))))))
398 399 400 401 402 403 404 405 406 407 408
    (cond
     ;; Processing instructions (like the <?xml version="1.0"?> tag at the
     ;; beginning of a document).
     ((looking-at "<\\?")
      (search-forward "?>")
      (skip-syntax-forward " ")
      (xml-parse-tag parse-dtd xml-ns))
     ;;  Character data (CDATA) sections, in which no tag should be interpreted
     ((looking-at "<!\\[CDATA\\[")
      (let ((pos (match-end 0)))
	(unless (search-forward "]]>" nil t)
409
	  (error "XML: (Not Well Formed) CDATA section does not end anywhere in the document"))
410
	(concat
411
	 (buffer-substring-no-properties pos (match-beginning 0))
412
	 (xml-parse-string))))
413 414
     ;;  DTD for the document
     ((looking-at "<!DOCTYPE")
415 416 417 418 419
      (let ((dtd (xml-parse-dtd parse-ns)))
	(skip-syntax-forward " ")
	(if xml-validating-parser
	    (cons dtd (xml-parse-tag nil xml-ns))
	  (xml-parse-tag nil xml-ns))))
420 421 422 423 424 425 426 427 428 429
     ;;  skip comments
     ((looking-at "<!--")
      (search-forward "-->")
      nil)
     ;;  end tag
     ((looking-at "</")
      '())
     ;;  opening tag
     ((looking-at "<\\([^/>[:space:]]+\\)")
      (goto-char (match-end 1))
430 431

      ;; Parse this node
432
      (let* ((node-name (match-string-no-properties 1))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
433 434 435
	     ;; Parse the attribute list.
	     (attrs (xml-parse-attlist xml-ns))
	     children pos)
436

Mark A. Hershberger's avatar
Mark A. Hershberger committed
437 438
	;; add the xmlns:* attrs to our cache
	(when (consp xml-ns)
439 440
	  (dolist (attr attrs)
	    (when (and (consp (car attr))
441 442 443
		       (equal "http://www.w3.org/2000/xmlns/"
			      (caar attr)))
	      (push (cons (cdar attr) (cdr attr))
444 445
		    xml-ns))))

Mark A. Hershberger's avatar
Mark A. Hershberger committed
446
	(setq children (list attrs (xml-maybe-do-ns node-name "" xml-ns)))
447

448 449
	;; is this an empty element ?
	(if (looking-at "/>")
Gerd Moellmann's avatar
Gerd Moellmann committed
450
	    (progn
451
	      (forward-char 2)
452
	      (nreverse children))
453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475

	  ;; is this a valid start tag ?
	  (if (eq (char-after) ?>)
	      (progn
		(forward-char 1)
		;;  Now check that we have the right end-tag. Note that this
		;;  one might contain spaces after the tag name
		(let ((end (concat "</" node-name "\\s-*>")))
		  (while (not (looking-at end))
		    (cond
		     ((looking-at "</")
		      (error "XML: (Not Well-Formed) Invalid end tag (expecting %s) at pos %d"
			     node-name (point)))
		     ((= (char-after) ?<)
		      (let ((tag (xml-parse-tag nil xml-ns)))
			(when tag
			  (push tag children))))
		     (t
		      (let ((expansion (xml-parse-string)))
			(setq children
			      (if (stringp expansion)
				  (if (stringp (car children))
				      ;; The two strings were separated by a comment.
Mark A. Hershberger's avatar
Mark A. Hershberger committed
476
				      (setq children (append (list (concat (car children) expansion))
477 478 479 480 481 482 483 484
							     (cdr children)))
				    (setq children (append (list expansion) children)))
				(setq children (append expansion children))))))))

		  (goto-char (match-end 0))
		  (nreverse children)))
	    ;;  This was an invalid start tag (Expected ">", but didn't see it.)
	    (error "XML: (Well-Formed) Couldn't parse tag: %s"
485
		   (buffer-substring-no-properties (- (point) 10) (+ (point) 1)))))))
486 487 488 489 490 491 492 493 494 495
     (t	;; (Not one of PI, CDATA, Comment, End tag, or Start tag)
      (unless xml-sub-parser		; Usually, we error out.
	(error "XML: (Well-Formed) Invalid character"))

      ;; However, if we're parsing incrementally, then we need to deal
      ;; with stray CDATA.
      (xml-parse-string)))))

(defun xml-parse-string ()
  "Parse the next whatever.  Could be a string, or an element."
Mark A. Hershberger's avatar
Mark A. Hershberger committed
496
  (let* ((pos (point))
497
	 (string (progn (skip-chars-forward "^<")
498
			(buffer-substring-no-properties pos (point)))))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
499 500 501 502 503 504 505 506 507 508
    ;; Clean up the string.  As per XML specifications, the XML
    ;; processor should always pass the whole string to the
    ;; application.  But \r's should be replaced:
    ;; http://www.w3.org/TR/2000/REC-xml-20001006#sec-line-ends
    (setq pos 0)
    (while (string-match "\r\n?" string pos)
      (setq string (replace-match "\n" t t string))
      (setq pos (1+ (match-beginning 0))))

    (xml-substitute-special string)))
Gerd Moellmann's avatar
Gerd Moellmann committed
509

510
(defun xml-parse-attlist (&optional xml-ns)
Stefan Monnier's avatar
Stefan Monnier committed
511 512
  "Return the attribute-list after point.
Leave point at the first non-blank character after the tag."
513
  (let ((attlist ())
514
	end-pos name)
Dave Love's avatar
Dave Love committed
515 516 517
    (skip-syntax-forward " ")
    (while (looking-at (eval-when-compile
			 (concat "\\(" xml-name-regexp "\\)\\s-*=\\s-*")))
518
      (setq end-pos (match-end 0))
519
      (setq name (xml-maybe-do-ns (match-string-no-properties 1) nil xml-ns))
520
      (goto-char end-pos)
Gerd Moellmann's avatar
Gerd Moellmann committed
521

522 523
      ;; See also: http://www.w3.org/TR/2000/REC-xml-20001006#AVNormalize

Gerd Moellmann's avatar
Gerd Moellmann committed
524 525
      ;; Do we have a string between quotes (or double-quotes),
      ;;  or a simple word ?
526
      (if (looking-at "\"\\([^\"]*\\)\"")
527
	  (setq end-pos (match-end 0))
528
	(if (looking-at "'\\([^']*\\)'")
529
	    (setq end-pos (match-end 0))
530
	  (error "XML: (Not Well-Formed) Attribute values must be given between quotes")))
Gerd Moellmann's avatar
Gerd Moellmann committed
531 532 533

      ;; Each attribute must be unique within a given element
      (if (assoc name attlist)
534
	  (error "XML: (Not Well-Formed) Each attribute must be unique within an element"))
535

536 537
      ;; Multiple whitespace characters should be replaced with a single one
      ;; in the attributes
538
      (let ((string (match-string-no-properties 1))
539
	    (pos 0))
Dave Love's avatar
Dave Love committed
540
	(replace-regexp-in-string "\\s-\\{2,\\}" " " string)
541 542
	(let ((expansion (xml-substitute-special string)))
	  (unless (stringp expansion)
Mark A. Hershberger's avatar
Mark A. Hershberger committed
543 544
					; We say this is the constraint.  It is acctually that
					; external entities nor "<" can be in an attribute value.
545 546
	    (error "XML: (Not Well-Formed) Entities in attributes cannot expand into elements"))
	  (push (cons name expansion) attlist)))
547

548
      (goto-char end-pos)
Dave Love's avatar
Dave Love committed
549
      (skip-syntax-forward " "))
550
    (nreverse attlist)))
Gerd Moellmann's avatar
Gerd Moellmann committed
551 552 553 554 555 556 557 558 559

;;*******************************************************************
;;**
;;**  The DTD (document type declaration)
;;**  The following functions know how to skip or parse the DTD of
;;**  a document
;;**
;;*******************************************************************

Dave Love's avatar
Dave Love committed
560 561 562 563
;; Fixme: This fails at least if the DTD contains conditional sections.

(defun xml-skip-dtd ()
  "Skip the DTD at point.
Gerd Moellmann's avatar
Gerd Moellmann committed
564
This follows the rule [28] in the XML specifications."
565 566
  (let ((xml-validating-parser nil))
    (xml-parse-dtd)))
Gerd Moellmann's avatar
Gerd Moellmann committed
567

568
(defun xml-parse-dtd (&optional parse-ns)
Dave Love's avatar
Dave Love committed
569 570 571
  "Parse the DTD at point."
  (forward-char (eval-when-compile (length "<!DOCTYPE")))
  (skip-syntax-forward " ")
572 573 574
  (if (and (looking-at ">")
	   xml-validating-parser)
      (error "XML: (Validity) Invalid DTD (expecting name of the document)"))
575

576
  ;;  Get the name of the document
Dave Love's avatar
Dave Love committed
577
  (looking-at xml-name-regexp)
578
  (let ((dtd (list (match-string-no-properties 0) 'dtd))
579
	type element end-pos)
Gerd Moellmann's avatar
Gerd Moellmann committed
580 581
    (goto-char (match-end 0))

Dave Love's avatar
Dave Love committed
582 583 584 585 586 587 588 589 590 591
    (skip-syntax-forward " ")
    ;; XML [75]
    (cond ((looking-at "PUBLIC\\s-+")
	   (goto-char (match-end 0))
	   (unless (or (re-search-forward
			"\\=\"\\([[:space:][:alnum:]-'()+,./:=?;!*#@$_%]*\\)\""
			nil t)
		       (re-search-forward
			"\\='\\([[:space:][:alnum:]-()+,./:=?;!*#@$_%]*\\)'"
			nil t))
592
	     (error "XML: Missing Public ID"))
593
	   (let ((pubid (match-string-no-properties 1)))
594
	     (skip-syntax-forward " ")
Dave Love's avatar
Dave Love committed
595 596
	     (unless (or (re-search-forward "\\='\\([^']*\\)'" nil t)
			 (re-search-forward "\\=\"\\([^\"]*\\)\"" nil t))
597
	       (error "XML: Missing System ID"))
598
	     (push (list pubid (match-string-no-properties 1) 'public) dtd)))
Dave Love's avatar
Dave Love committed
599 600 601 602
	  ((looking-at "SYSTEM\\s-+")
	   (goto-char (match-end 0))
	   (unless (or (re-search-forward "\\='\\([^']*\\)'" nil t)
		       (re-search-forward "\\=\"\\([^\"]*\\)\"" nil t))
603
	     (error "XML: Missing System ID"))
604
	   (push (list (match-string-no-properties 1) 'system) dtd)))
Dave Love's avatar
Dave Love committed
605 606 607 608
    (skip-syntax-forward " ")
    (if (eq ?> (char-after))
	(forward-char)
      (if (not (eq (char-after) ?\[))
609
	  (error "XML: Bad DTD")
Dave Love's avatar
Dave Love committed
610 611
	(forward-char)
	;;  Parse the rest of the DTD
612
	;;  Fixme: Deal with NOTATION, PIs.
Dave Love's avatar
Dave Love committed
613 614 615 616 617 618 619 620
	(while (not (looking-at "\\s-*\\]"))
	  (skip-syntax-forward " ")
	  (cond

	   ;;  Translation of rule [45] of XML specifications
	   ((looking-at
	     "<!ELEMENT\\s-+\\([[:alnum:].%;]+\\)\\s-+\\([^>]+\\)>")

621
	    (setq element (match-string-no-properties 1)
Dave Love's avatar
Dave Love committed
622 623 624 625 626 627 628 629 630 631
		  type    (match-string-no-properties 2))
	    (setq end-pos (match-end 0))

	    ;;  Translation of rule [46] of XML specifications
	    (cond
	     ((string-match "^EMPTY[ \t\n\r]*$" type) ;; empty declaration
	      (setq type 'empty))
	     ((string-match "^ANY[ \t\n\r]*$" type) ;; any type of contents
	      (setq type 'any))
	     ((string-match "^(\\(.*\\))[ \t\n\r]*$" type) ;; children ([47])
632
	      (setq type (xml-parse-elem-type (match-string-no-properties 1 type))))
Dave Love's avatar
Dave Love committed
633 634 635
	     ((string-match "^%[^;]+;[ \t\n\r]*$" type)	;; substitution
	      nil)
	     (t
636
	      (if xml-validating-parser 
637
		  (error "XML: (Validity) Invalid element type in the DTD"))))
638

Dave Love's avatar
Dave Love committed
639
	    ;;  rule [45]: the element declaration must be unique
640 641 642
	    (if (and (assoc element dtd)
		     xml-validating-parser)
		(error "XML: (Validity) Element declarations must be unique in a DTD (<%s>)"
643
		       element))
Dave Love's avatar
Dave Love committed
644 645 646 647

	    ;;  Store the element in the DTD
	    (push (list element type) dtd)
	    (goto-char end-pos))
648 649 650 651 652 653 654 655 656

	   ;; Translation of rule [52] of XML specifications
	   ((looking-at (concat "<!ATTLIST[ \t\n\r]*\\(" xml-name-re
				"\\)[ \t\n\r]*\\(" xml-att-def-re
				"\\)*[ \t\n\r]*>"))

	    ;; We don't do anything with ATTLIST currently
	    (goto-char (match-end 0)))

Dave Love's avatar
Dave Love committed
657 658
	   ((looking-at "<!--")
	    (search-forward "-->"))
659 660 661
	   ((looking-at (concat "<!ENTITY[ \t\n\r]*\\(" xml-name-re
				"\\)[ \t\n\r]*\\(" xml-entity-value-re
				"\\)[ \t\n\r]*>"))
662 663 664
	    (let ((name  (match-string-no-properties 1))
		  (value (substring (match-string-no-properties 2) 1
				    (- (length (match-string-no-properties 2)) 1))))
665
	      (goto-char (match-end 0))
666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683
	      (setq xml-entity-alist
		    (append xml-entity-alist
			    (list (cons name 
					(with-temp-buffer
					  (insert value)
					  (goto-char (point-min))
					  (xml-parse-fragment
					   xml-validating-parser
					   parse-ns))))))))
	   ((or (looking-at (concat "<!ENTITY[ \t\n\r]+\\(" xml-name-re
				    "\\)[ \t\n\r]+SYSTEM[ \t\n\r]+"
				    "\\(\"[^\"]*\"\\|'[^']*'\\)[ \t\n\r]*>"))
		(looking-at (concat "<!ENTITY[ \t\n\r]+\\(" xml-name-re
				    "\\)[ \t\n\r]+PUBLIC[ \t\n\r]+"
				    "\"[- \r\na-zA-Z0-9'()+,./:=?;!*#@$_%]*\""
				    "\\|'[- \r\na-zA-Z0-9()+,./:=?;!*#@$_%]*'"
				    "[ \t\n\r]+\\(\"[^\"]*\"\\|'[^']*'\\)"
				    "[ \t\n\r]*>")))
684 685 686
	    (let ((name  (match-string-no-properties 1))
		  (file  (substring (match-string-no-properties 2) 1
				    (- (length (match-string-no-properties 2)) 1))))
687
	      (goto-char (match-end 0))
688 689 690 691 692 693 694 695
	      (setq xml-entity-alist
		    (append xml-entity-alist
			    (list (cons name (with-temp-buffer
					       (insert-file-contents file)
					       (goto-char (point-min))
					       (xml-parse-fragment
						xml-validating-parser
						parse-ns))))))))
696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711
	   ;; skip parameter entity declarations
	   ((or (looking-at (concat "<!ENTITY[ \t\n\r]+%[ \t\n\r]+\\(" xml-name-re
				    "\\)[ \t\n\r]+SYSTEM[ \t\n\r]+"
				    "\\(\"[^\"]*\"\\|'[^']*'\\)[ \t\n\r]*>"))
		(looking-at (concat "<!ENTITY[ \t\n\r]+"
				    "%[ \t\n\r]+"
				    "\\(" xml-name-re "\\)[ \t\n\r]+"
				    "PUBLIC[ \t\n\r]+"
				    "\\(\"[- \r\na-zA-Z0-9'()+,./:=?;!*#@$_%]*\""
				    "\\|'[- \r\na-zA-Z0-9()+,./:=?;!*#@$_%]*'\\)[ \t\n\r]+"
				    "\\(\"[^\"]+\"\\|'[^']+'\\)"
				    "[ \t\n\r]*>")))
	    (goto-char (match-end 0)))
	   ;; skip parameter entities
	   ((looking-at (concat "%" xml-name-re ";"))
	    (goto-char (match-end 0)))
Dave Love's avatar
Dave Love committed
712
	   (t
713
	    (when xml-validating-parser
714
	      (error "XML: (Validity) Invalid DTD item"))))))
715
      (if (looking-at "\\s-*]>")
716
	  (goto-char (match-end 0))))
717
    (nreverse dtd)))
Gerd Moellmann's avatar
Gerd Moellmann committed
718 719

(defun xml-parse-elem-type (string)
Dave Love's avatar
Dave Love committed
720
  "Convert element type STRING into a Lisp structure."
Gerd Moellmann's avatar
Gerd Moellmann committed
721 722 723 724

  (let (elem modifier)
    (if (string-match "(\\([^)]+\\))\\([+*?]?\\)" string)
	(progn
725 726
	  (setq elem     (match-string-no-properties 1 string)
		modifier (match-string-no-properties 2 string))
Gerd Moellmann's avatar
Gerd Moellmann committed
727
	  (if (string-match "|" elem)
728
	      (setq elem (cons 'choice
Gerd Moellmann's avatar
Gerd Moellmann committed
729 730 731
			       (mapcar 'xml-parse-elem-type
				       (split-string elem "|"))))
	    (if (string-match "," elem)
732
		(setq elem (cons 'seq
Gerd Moellmann's avatar
Gerd Moellmann committed
733
				 (mapcar 'xml-parse-elem-type
Dave Love's avatar
Dave Love committed
734
					 (split-string elem ",")))))))
735
      (if (string-match "[ \t\n\r]*\\([^+*?]+\\)\\([+*?]?\\)" string)
736 737
	  (setq elem	 (match-string-no-properties 1 string)
		modifier (match-string-no-properties 2 string))))
Gerd Moellmann's avatar
Gerd Moellmann committed
738

739 740
    (if (and (stringp elem) (string= elem "#PCDATA"))
	(setq elem 'pcdata))
741

742 743 744 745 746 747
    (cond
     ((string= modifier "+")
      (list '+ elem))
     ((string= modifier "*")
      (list '* elem))
     ((string= modifier "?")
748
      (list '\? elem))
749 750
     (t
      elem))))
Gerd Moellmann's avatar
Gerd Moellmann committed
751 752 753 754 755 756 757 758

;;*******************************************************************
;;**
;;**  Substituting special XML sequences
;;**
;;*******************************************************************

(defun xml-substitute-special (string)
Dave Love's avatar
Dave Love committed
759 760 761 762
  "Return STRING, after subsituting entity references."
  ;; This originally made repeated passes through the string from the
  ;; beginning, which isn't correct, since then either "&amp;amp;" or
  ;; "&#38;amp;" won't DTRT.
Gerd Moellmann's avatar
Gerd Moellmann committed
763

764 765
  (let ((point 0)
	children end-point)
766
    (while (string-match "&\\([^;]*\\);" string point)
767
      (setq end-point (match-end 0))
768
      (let* ((this-part (match-string-no-properties 1 string))
769 770 771 772 773 774
	     (prev-part (substring string point (match-beginning 0)))
	     (entity (assoc this-part xml-entity-alist))
	     (expansion 
	      (cond ((string-match "#\\([0-9]+\\)" this-part)
		     (let ((c (decode-char
			       'ucs
775
			       (string-to-number (match-string-no-properties 1 this-part)))))
776 777 778 779
		       (if c (string c))))
		    ((string-match "#x\\([[:xdigit:]]+\\)" this-part)
		     (let ((c (decode-char
			       'ucs
780
			       (string-to-number (match-string-no-properties 1 this-part) 16))))
781 782 783
		       (if c (string c))))
		    (entity
		     (cdr entity))
784
		    ((eq (length this-part) 0)
785
		     (error "XML: (Not Well-Formed) No entity given"))
786
		    (t
787
		     (if xml-validating-parser
788
			 (error "XML: (Validity) Undefined entity `%s'"
789 790
				this-part)
		       xml-undefined-entity)))))
791 792

	(cond ((null children)
793 794 795
	       ;; FIXME: If we have an entity that expands into XML, this won't work.
	       (setq children
		     (concat prev-part expansion)))
796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816
	      ((stringp children)
	       (if (stringp expansion)
		   (setq children (concat children prev-part expansion))
		 (setq children (list expansion (concat prev-part children)))))
	      ((and (stringp expansion)
		    (stringp (car children)))
	       (setcar children (concat prev-part expansion (car children))))
	      ((stringp expansion)
	       (setq children (append (concat prev-part expansion)
				      children)))
	      ((stringp (car children))
	       (setcar children (concat (car children) prev-part))
	       (setq children (append expansion children)))
	      (t
	       (setq children (list expansion
				    prev-part
				    children))))
	(setq point end-point)))
    (cond ((stringp children)
	   (concat children (substring string point)))
	  ((stringp (car (last children)))
817
	   (concat (car (last children)) (substring string point)))
818 819 820
	  ((null children)
	   string)
	  (t
821 822 823 824 825
	   (concat (mapconcat 'identity
			      (nreverse children)
			      "")
		   (substring string point))))))

Gerd Moellmann's avatar
Gerd Moellmann committed
826 827 828 829 830 831 832
;;*******************************************************************
;;**
;;**  Printing a tree.
;;**  This function is intended mainly for debugging purposes.
;;**
;;*******************************************************************

833 834 835 836 837
(defun xml-debug-print (xml &optional indent-string)
  "Outputs the XML in the current buffer.
XML can be a tree or a list of nodes.
The first line is indented with the optional INDENT-STRING."
  (setq indent-string (or indent-string ""))
838
  (dolist (node xml)
839 840 841
    (xml-debug-print-internal node indent-string)))

(defalias 'xml-print 'xml-debug-print)
Gerd Moellmann's avatar
Gerd Moellmann committed
842

Mark A. Hershberger's avatar
Mark A. Hershberger committed
843
(defun xml-escape-string (string)
844 845
  "Return the string with entity substitutions made from
xml-entity-alist."
Mark A. Hershberger's avatar
Mark A. Hershberger committed
846 847 848 849 850
  (mapconcat (lambda (byte)
               (let ((char (char-to-string byte)))
                 (if (rassoc char xml-entity-alist)
                     (concat "&" (car (rassoc char xml-entity-alist)) ";")
                   char)))
851 852 853
             ;; This differs from the non-unicode branch.  Just
             ;; grabbing the string works here.
             string ""))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
854

855
(defun xml-debug-print-internal (xml indent-string)
Gerd Moellmann's avatar
Gerd Moellmann committed
856
  "Outputs the XML tree in the current buffer.
Dave Love's avatar
Dave Love committed
857
The first line is indented with INDENT-STRING."
Gerd Moellmann's avatar
Gerd Moellmann committed
858 859
  (let ((tree xml)
	attlist)
Dave Love's avatar
Dave Love committed
860
    (insert indent-string ?< (symbol-name (xml-node-name tree)))
861

Gerd Moellmann's avatar
Gerd Moellmann committed
862
    ;;  output the attribute list
863
    (setq attlist (xml-node-attributes tree))
Gerd Moellmann's avatar
Gerd Moellmann committed
864
    (while attlist
Mark A. Hershberger's avatar
Mark A. Hershberger committed
865 866
      (insert ?\  (symbol-name (caar attlist)) "=\""
              (xml-escape-string (cdar attlist)) ?\")
867
      (setq attlist (cdr attlist)))
868

869
    (setq tree (xml-node-children tree))
Gerd Moellmann's avatar
Gerd Moellmann committed
870

871 872 873 874 875 876 877 878 879 880
    (if (null tree)
	(insert ?/ ?>)
      (insert ?>)

      ;;  output the children
      (dolist (node tree)
	(cond
	 ((listp node)
	  (insert ?\n)
	  (xml-debug-print-internal node (concat indent-string "  ")))
Mark A. Hershberger's avatar
Mark A. Hershberger committed
881 882
	 ((stringp node)
          (insert (xml-escape-string node)))
883 884 885 886 887 888 889
	 (t
	  (error "Invalid XML tree"))))

      (when (not (and (null (cdr tree))
		      (stringp (car tree))))
	(insert ?\n indent-string))
      (insert ?< ?/ (symbol-name (xml-node-name xml)) ?>))))
Gerd Moellmann's avatar
Gerd Moellmann committed
890 891 892

(provide 'xml)

893
;; arch-tag: 5864b283-5a68-4b59-a20d-36a72b353b9b
Gerd Moellmann's avatar
Gerd Moellmann committed
894
;;; xml.el ends here