-;; @(#) xml.el --- XML parser
+;;; xml.el --- XML parser
-;; Copyright (C) 2000 Free Software Foundation, Inc.
+;; Copyright (C) 2000, 2001, 2003 Free Software Foundation, Inc.
;; Author: Emmanuel Briot <briot@gnat.com>
-;; Maintainer: Emmanuel Briot <briot@gnat.com>
-;; Keywords: xml
+;; Maintainer: Mark A. Hershberger <mah@everybody.org>
+;; Keywords: xml, data
;; This file is part of GNU Emacs.
;;; Commentary:
-;; This file contains a full XML parser. It parses a file, and returns a list
-;; that can be used internally by any other lisp file.
-;; See some example in todo.el
+;; This file contains a somewhat incomplete non-validating XML parser. It
+;; parses a file, and returns a list that can be used internally by
+;; any other lisp libraries.
;;; FILE FORMAT
-;; It does not parse the DTD, if present in the XML file, but knows how to
-;; ignore it. The XML file is assumed to be well-formed. In case of error, the
-;; parsing stops and the XML file is shown where the parsing stopped.
+;; The document type declaration may either be ignored or (optionally)
+;; parsed, but currently the parsing will only accept element
+;; declarations. The XML file is assumed to be well-formed. In case
+;; of error, the parsing stops and the XML file is shown where the
+;; parsing stopped.
;;
-;; It also knows how to ignore comments, as well as the special ?xml? tag
-;; in the XML file.
+;; It also knows how to ignore comments and processing instructions.
;;
;; The XML file should have the following format:
-;; <node1 attr1="name1" attr2="name2" ...> value
-;; <node2 attr3="name3" attr4="name4"> value2 </node2>
-;; <node3 attr5="name5" attr6="name6"> value3 </node3>
+;; <node1 attr1="name1" attr2="name2" ...>value
+;; <node2 attr3="name3" attr4="name4">value2</node2>
+;; <node3 attr5="name5" attr6="name6">value3</node3>
;; </node1>
;; Of course, the name of the nodes and attributes can be anything. There can
;; be any number of attributes (or none), as well as any number of children
;; | nil
;; string ::= "..."
;;
-;; Since XML is case insensitive, tag_name is always converted to lower-cases.
-;; tag_name is then converted to a symbol (this is not a string, so that the
-;; list takes less space in memory and is faster to traverse).
-;;
-;; Some macros are provided to ease the parsing of this list
+;; 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.
;;; Code:
+;; Note that {buffer-substring,match-string}-no-properties were
+;; formerly used in several places, but that removes composition info.
+
;;*******************************************************************
;;**
;;** Macros to parse the list
;;**
;;*******************************************************************
-(defmacro xml-node-name (node)
+(defsubst xml-node-name (node)
"Return the tag associated with NODE.
The tag is a lower-case symbol."
- (list 'car node))
+ (car node))
-(defmacro xml-node-attributes (node)
+(defsubst xml-node-attributes (node)
"Return the list of attributes of NODE.
The list can be nil."
- (list 'nth 1 node))
+ (nth 1 node))
-(defmacro xml-node-children (node)
+(defsubst xml-node-children (node)
"Return the list of children of NODE.
This is a list of nodes, and it can be nil."
- (list 'cddr node))
+ (cddr node))
(defun xml-get-children (node child-name)
"Return the children of NODE whose tag is CHILD-NAME.
CHILD-NAME should be a lower case symbol."
- (let ((children (xml-node-children node))
- match)
- (while children
- (if (car children)
- (if (equal (xml-node-name (car children)) child-name)
- (set 'match (append match (list (car children))))))
- (set 'children (cdr children)))
- match))
-
-(defun xml-get-attribute (node attribute)
+ (let ((match ()))
+ (dolist (child (xml-node-children node))
+ (if child
+ (if (equal (xml-node-name child) child-name)
+ (push child match))))
+ (nreverse match)))
+
+(defun xml-get-attribute-or-nil (node attribute)
+ "Get from NODE the value of ATTRIBUTE.
+Return `nil' if the attribute was not found.
+
+See also `xml-get-attribute'."
+ (when (xml-node-attributes node)
+ (let ((value (assoc attribute (xml-node-attributes node))))
+ (when value
+ (cdr value)))))
+
+(defsubst xml-get-attribute (node attribute)
"Get from NODE the value of ATTRIBUTE.
-An empty string is returned if the attribute was not found."
- (if (xml-node-attributes node)
- (let ((value (assoc attribute (xml-node-attributes node))))
- (if value
- (cdr value)
- ""))
- ""))
+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) ""))
;;*******************************************************************
;;**
;;**
;;*******************************************************************
-(defun xml-parse-file (file &optional parse-dtd)
- "Parse the well-formed XML FILE.
+;;;###autoload
+(defun xml-parse-file (file &optional parse-dtd parse-ns)
+ "Parse the well-formed XML file FILE.
+If FILE is already visited, use its buffer and don't kill it.
Returns the top node with all its children.
-If PARSE-DTD is non-nil, the DTD is parsed rather than skipped."
- (find-file file)
- (let ((xml (xml-parse-region (point-min)
- (point-max)
- (current-buffer)
- parse-dtd)))
- (kill-buffer (current-buffer))
- xml))
-
-(defun xml-parse-region (beg end &optional buffer parse-dtd)
+If PARSE-DTD is non-nil, the DTD is parsed rather than skipped.
+If PARSE-NS is non-nil, then QNAMES are expanded."
+ (let ((keep))
+ (if (get-file-buffer file)
+ (progn
+ (set-buffer (get-file-buffer file))
+ (setq keep (point)))
+ (let (auto-mode-alist) ; no need for xml-mode
+ (find-file file)))
+
+ (let ((xml (xml-parse-region (point-min)
+ (point-max)
+ (current-buffer)
+ parse-dtd parse-ns)))
+ (if keep
+ (goto-char keep)
+ (kill-buffer (current-buffer)))
+ xml)))
+
+;; 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
+ (dolist (c '(?\t ?\n ?\r)) ; these should be space
+ (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]
+ (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))
+ ;; Fixme: rest of [4]
+ table)
+ "Syntax table used by `xml-parse-region'.")
+
+;; XML [5]
+;; Note that [:alpha:] matches all multibyte chars with word syntax.
+(eval-and-compile
+ (defconst xml-name-regexp "[[:alpha:]_:][[:alnum:]._:-]*"))
+
+;; Fixme: This needs re-writing to deal with the XML grammar properly, i.e.
+;; document ::= prolog element Misc*
+;; prolog ::= XMLDecl? Misc* (doctypedecl Misc*)?
+
+;;;###autoload
+(defun xml-parse-region (beg end &optional buffer parse-dtd parse-ns)
"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
-is not a well-formed XML file.
+is not well-formed XML.
If PARSE-DTD is non-nil, the DTD is parsed rather than skipped,
-and returned as the first element of the list"
- (let (xml result dtd)
- (save-excursion
- (if buffer
- (set-buffer buffer))
- (goto-char beg)
- (while (< (point) end)
- (if (search-forward "<" end t)
- (progn
- (forward-char -1)
- (if (null xml)
- (progn
- (set 'result (xml-parse-tag end parse-dtd))
+and returned as the first element of the list.
+If PARSE-NS is non-nil, then QNAMES are expanded."
+ (save-restriction
+ (narrow-to-region beg end)
+ ;; 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))
+ (goto-char (point-min))
+ (while (not (eobp))
+ (if (search-forward "<" nil t)
+ (progn
+ (forward-char -1)
+ (setq result (xml-parse-tag parse-dtd parse-ns))
+ (if (and xml result)
+ ;; translation of rule [1] of XML specifications
+ (error "XML files can have only one toplevel tag")
(cond
- ((listp (car result))
- (set 'dtd (car result))
- (add-to-list 'xml (cdr result)))
+ ((null result))
+ ((and (listp (car result))
+ parse-dtd)
+ (setq dtd (car result))
+ (if (cdr result) ; possible leading comment
+ (add-to-list 'xml (cdr result))))
(t
- (add-to-list 'xml result))))
-
- ;; translation of rule [1] of XML specifications
- (error "XML files can have only one toplevel tag.")))
- (goto-char end)))
- (if parse-dtd
- (cons dtd (reverse xml))
- (reverse xml)))))
-
-
-(defun xml-parse-tag (end &optional parse-dtd)
- "Parse the tag that is just in front of point.
-The end tag must be found before the position END in the current buffer.
+ (add-to-list 'xml result)))))
+ (goto-char (point-max))))
+ (if parse-dtd
+ (cons dtd (nreverse xml))
+ (nreverse xml)))))))
+
+(defun xml-ns-parse-ns-attrs (attr-list &optional xml-ns)
+ "Parse the namespace attributes and return a list of cons in the form:
+\(namespace . prefix)"
+
+ (mapcar
+ (lambda (attr)
+ (let* ((splitup (split-string (car attr) ":"))
+ (prefix (nth 0 splitup))
+ (lname (nth 1 splitup)))
+ (when (string= "xmlns" prefix)
+ (push (cons (if lname
+ lname
+ "")
+ (cdr attr))
+ xml-ns)))) attr-list)
+ xml-ns)
+
+;; expand element names
+(defun xml-ns-expand-el (el xml-ns)
+ "Expand the XML elements from \"prefix:local-name\" to a cons in the form
+\"(namespace . local-name)\"."
+
+ (let* ((splitup (split-string el ":"))
+ (lname (or (nth 1 splitup)
+ (nth 0 splitup)))
+ (prefix (if (nth 1 splitup)
+ (nth 0 splitup)
+ (if (string= lname "xmlns")
+ "xmlns"
+ "")))
+ (ns (cdr (assoc-string prefix xml-ns))))
+ (if (string= "" ns)
+ lname
+ (cons (intern (concat ":" ns))
+ lname))))
+
+;; expand attribute names
+(defun xml-ns-expand-attr (attr-list xml-ns)
+ "Expand the attribute list for a particular element from the form
+\"prefix:local-name\" to the form \"{namespace}:local-name\"."
+
+ (mapcar
+ (lambda (attr)
+ (let* ((splitup (split-string (car attr) ":"))
+ (lname (or (nth 1 splitup)
+ (nth 0 splitup)))
+ (prefix (if (nth 1 splitup)
+ (nth 0 splitup)
+ (if (string= (car attr) "xmlns")
+ "xmlns"
+ "")))
+ (ns (cdr (assoc-string prefix xml-ns))))
+ (setcar attr
+ (if (string= "" ns)
+ lname
+ (cons (intern (concat ":" ns))
+ lname)))))
+ attr-list)
+ attr-list)
+
+(defun xml-intern-attrlist (attr-list)
+ "Convert attribute names to symbols for backward compatibility."
+ (mapcar (lambda (attr)
+ (setcar attr (intern (car attr))))
+ attr-list)
+ attr-list)
+
+(defun xml-parse-tag (&optional parse-dtd parse-ns)
+ "Parse the tag at point.
If PARSE-DTD is non-nil, the DTD of the document, if any, is parsed and
returned as the first element in the list.
+If PARSE-NS is non-nil, then QNAMES are expanded.
Returns one of:
- - a list : the matching node
- - nil : the point is not looking at a tag.
- - a cons cell: the first element is the DTD, the second is the node"
- (cond
- ;; Processing instructions (like the <?xml version="1.0"?> tag at the
- ;; beginning of a document)
- ((looking-at "<\\?")
- (search-forward "?>" end)
- (skip-chars-forward " \t\n")
- (xml-parse-tag end))
- ;; Character data (CDATA) sections, in which no tag should be interpreted
- ((looking-at "<!\\[CDATA\\[")
- (let ((pos (match-end 0)))
- (unless (search-forward "]]>" end t)
- (error "CDATA section does not end anywhere in the document"))
- (buffer-substring-no-properties pos (match-beginning 0))))
- ;; DTD for the document
- ((looking-at "<!DOCTYPE")
- (let (dtd)
- (if parse-dtd
- (set 'dtd (xml-parse-dtd end))
- (xml-skip-dtd end))
- (skip-chars-forward " \t\n")
+ - 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."
+ (let ((xml-ns (if (consp parse-ns)
+ parse-ns
+ (if parse-ns
+ (list
+ ;; Default no namespace
+ (cons "" "")
+ ;; We need to seed the xmlns namespace
+ (cons "xmlns" "http://www.w3.org/2000/xmlns/"))))))
+ (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)
+ (error "CDATA section does not end anywhere in the document"))
+ (buffer-substring pos (match-beginning 0))))
+ ;; DTD for the document
+ ((looking-at "<!DOCTYPE")
+ (let (dtd)
+ (if parse-dtd
+ (setq dtd (xml-parse-dtd))
+ (xml-skip-dtd))
+ (skip-syntax-forward " ")
(if dtd
- (cons dtd (xml-parse-tag end))
- (xml-parse-tag end))))
- ;; skip comments
- ((looking-at "<!--")
- (search-forward "-->" end)
- (skip-chars-forward " \t\n")
- (xml-parse-tag end))
- ;; end tag
- ((looking-at "</")
- '())
- ;; opening tag
- ((looking-at "<\\([^/> \t]+\\)")
- (let* ((node-name (match-string 1))
- (children (list (intern (downcase node-name))))
- pos)
+ (cons dtd (xml-parse-tag nil xml-ns))
+ (xml-parse-tag nil xml-ns))))
+ ;; skip comments
+ ((looking-at "<!--")
+ (search-forward "-->")
+ nil)
+ ;; end tag
+ ((looking-at "</")
+ '())
+ ;; opening tag
+ ((looking-at "<\\([^/>[:space:]]+\\)")
(goto-char (match-end 1))
- ;; parses the attribute list
- (set 'children (append children (list (xml-parse-attlist end))))
-
- ;; is this an empty element ?
- (if (looking-at "/>")
- (progn
- (forward-char 2)
- (skip-chars-forward " \t\n")
- (append children '("")))
+ ;; Parse this node
+ (let* ((node-name (match-string 1))
+ (attr-list (xml-parse-attlist))
+ (children (if (consp xml-ns) ;; take care of namespace parsing
+ (progn
+ (setq xml-ns (xml-ns-parse-ns-attrs
+ attr-list xml-ns))
+ (list (xml-ns-expand-attr
+ attr-list xml-ns)
+ (xml-ns-expand-el
+ node-name xml-ns)))
+ (list (xml-intern-attrlist attr-list)
+ (intern node-name))))
+ pos)
+
+ ;; is this an empty element ?
+ (if (looking-at "/>")
+ (progn
+ (forward-char 2)
+ (nreverse children))
;; is this a valid start tag ?
- (if (= (char-after) ?>)
+ (if (eq (char-after) ?>)
(progn
(forward-char 1)
- (skip-chars-forward " \t\n")
- (while (not (looking-at (concat "</" node-name ">")))
- (cond
- ((looking-at "</")
- (error (concat
- "XML: invalid syntax -- invalid end tag (expecting "
- node-name
- ")")))
- ((= (char-after) ?<)
- (set 'children (append children (list (xml-parse-tag end)))))
- (t
- (set 'pos (point))
- (search-forward "<" end)
- (forward-char -1)
- (let ((string (buffer-substring-no-properties pos (point)))
- (pos 0))
-
- ;; Clean up the string (no newline characters)
- ;; Not done, since as per XML specifications, the XML processor
- ;; should always pass the whole string to the application.
- ;; (while (string-match "\\s +" string pos)
- ;; (set 'string (replace-match " " t t string))
- ;; (set 'pos (1+ (match-beginning 0))))
-
- (set 'children (append children
- (list (xml-substitute-special string))))))))
- (goto-char (match-end 0))
- (skip-chars-forward " \t\n")
- (if (> (point) end)
- (error "XML: End tag for %s not found before end of region."
- node-name))
- children
- )
+ ;; 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: 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
+ (setq pos (point))
+ (search-forward "<")
+ (forward-char -1)
+ (let ((string (buffer-substring pos (point)))
+ (pos 0))
+
+ ;; 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
+ (while (string-match "\r\n?" string pos)
+ (setq string (replace-match "\n" t t string))
+ (setq pos (1+ (match-beginning 0))))
+
+ (setq string (xml-substitute-special string))
+ (setq children
+ (if (stringp (car children))
+ ;; The two strings were separated by a comment.
+ (cons (concat (car children) string)
+ (cdr children))
+ (cons string children))))))))
+ (goto-char (match-end 0))
+ (nreverse children))
;; This was an invalid start tag
- (error "XML: Invalid attribute list")
- ))))
- ))
-
-(defun xml-parse-attlist (end)
- "Return the attribute-list that point is looking at.
-The search for attributes end at the position END in the current buffer.
-Leaves the point on the first non-blank character after the tag."
- (let ((attlist '())
- name)
- (skip-chars-forward " \t\n")
- (while (looking-at "\\([a-zA-Z_:][a-zA-Z0-9.-_:]*\\)[ \t\n]*=[ \t\n]*")
- (set 'name (intern (downcase (match-string 1))))
+ (error "XML: Invalid attribute list")))))
+ (t ;; This is not a tag.
+ (error "XML: Invalid character")))))
+
+(defun xml-parse-attlist ()
+ "Return the attribute-list after point. Leave point at the
+first non-blank character after the tag."
+ (let ((attlist ())
+ end-pos name)
+ (skip-syntax-forward " ")
+ (while (looking-at (eval-when-compile
+ (concat "\\(" xml-name-regexp "\\)\\s-*=\\s-*")))
+ (setq name (match-string 1))
(goto-char (match-end 0))
+ ;; See also: http://www.w3.org/TR/2000/REC-xml-20001006#AVNormalize
+
;; Do we have a string between quotes (or double-quotes),
;; or a simple word ?
- (unless (looking-at "\"\\([^\"]+\\)\"")
- (unless (looking-at "'\\([^\"]+\\)'")
- (error "XML: Attribute values must be given between quotes.")))
+ (if (looking-at "\"\\([^\"]*\\)\"")
+ (setq end-pos (match-end 0))
+ (if (looking-at "'\\([^']*\\)'")
+ (setq end-pos (match-end 0))
+ (error "XML: Attribute values must be given between quotes")))
;; Each attribute must be unique within a given element
(if (assoc name attlist)
- (error "XML: each attribute must be unique within an element."))
-
- (set 'attlist (append attlist
- (list (cons name (match-string 1)))))
- (goto-char (match-end 0))
- (skip-chars-forward " \t\n")
- (if (> (point) end)
- (error "XML: end of attribute list not found before end of region."))
- )
- attlist
- ))
+ (error "XML: each attribute must be unique within an element"))
+
+ ;; Multiple whitespace characters should be replaced with a single one
+ ;; in the attributes
+ (let ((string (match-string 1))
+ (pos 0))
+ (replace-regexp-in-string "\\s-\\{2,\\}" " " string)
+ (push (cons name (xml-substitute-special string)) attlist))
+
+ (goto-char end-pos)
+ (skip-syntax-forward " "))
+ (nreverse attlist)))
;;*******************************************************************
;;**
;;**
;;*******************************************************************
-(defun xml-skip-dtd (end)
- "Skip the DTD that point is looking at.
-The DTD must end before the position END in the current buffer.
-The point must be just before the starting tag of the DTD.
+;; Fixme: This fails at least if the DTD contains conditional sections.
+
+(defun xml-skip-dtd ()
+ "Skip the DTD at point.
This follows the rule [28] in the XML specifications."
(forward-char (length "<!DOCTYPE"))
- (if (looking-at "[ \t\n]*>")
+ (if (looking-at "\\s-*>")
(error "XML: invalid DTD (excepting name of the document)"))
(condition-case nil
(progn
- (forward-word 1) ;; name of the document
- (skip-chars-forward " \t\n")
+ (forward-sexp)
+ (skip-syntax-forward " ")
(if (looking-at "\\[")
- (re-search-forward "\\][ \t\n]*>" end)
- (search-forward ">" end)))
+ (re-search-forward "]\\s-*>")
+ (search-forward ">")))
(error (error "XML: No end to the DTD"))))
-(defun xml-parse-dtd (end)
- "Parse the DTD that point is looking at.
-The DTD must end before the position END in the current buffer."
- (let (dtd type element end-pos)
- (forward-char (length "<!DOCTYPE"))
- (skip-chars-forward " \t\n")
- (if (looking-at ">")
- (error "XML: invalid DTD (excepting name of the document)"))
-
- ;; Get the name of the document
- (looking-at "\\sw+")
- (set 'dtd (list 'dtd (match-string-no-properties 0)))
- (goto-char (match-end 0))
-
- (skip-chars-forward " \t\n")
-
- ;; External DTDs => don't know how to handle them yet
- (if (looking-at "SYSTEM")
- (error "XML: Don't know how to handle external DTDs."))
-
- (if (not (= (char-after) ?\[))
- (error "XML: Unknown declaration in the DTD."))
-
- ;; Parse the rest of the DTD
- (forward-char 1)
- (while (and (not (looking-at "[ \t\n]*\\]"))
- (<= (point) end))
- (cond
-
- ;; Translation of rule [45] of XML specifications
- ((looking-at
- "[\t \n]*<!ELEMENT[ \t\n]+\\([a-zA-Z0-9.%;]+\\)[ \t\n]+\\([^>]+\\)>")
-
- (setq element (intern (downcase (match-string-no-properties 1)))
- type (match-string-no-properties 2))
- (set 'end-pos (match-end 0))
-
- ;; Translation of rule [46] of XML specifications
- (cond
- ((string-match "^EMPTY[ \t\n]*$" type) ;; empty declaration
- (set 'type 'empty))
- ((string-match "^ANY[ \t\n]*$" type) ;; any type of contents
- (set 'type 'any))
- ((string-match "^(\\(.*\\))[ \t\n]*$" type) ;; children ([47])
- (set 'type (xml-parse-elem-type (match-string-no-properties 1 type))))
- ((string-match "^%[^;]+;[ \t\n]*$" type) ;; substitution
- nil)
- (t
- (error "XML: Invalid element type in the DTD")))
-
- ;; rule [45]: the element declaration must be unique
- (if (assoc element dtd)
- (error "XML: elements declaration must be unique in a DTD (<%s>)."
- (symbol-name element)))
-
- ;; Store the element in the DTD
- (set 'dtd (append dtd (list (list element type))))
- (goto-char end-pos)
- )
-
-
- (t
- (error "XML: Invalid DTD item"))
- )
- )
+(defun xml-parse-dtd ()
+ "Parse the DTD at point."
+ (forward-char (eval-when-compile (length "<!DOCTYPE")))
+ (skip-syntax-forward " ")
+ (if (looking-at ">")
+ (error "XML: invalid DTD (excepting name of the document)"))
- ;; Skip the end of the DTD
- (search-forward ">" end)
- dtd
- ))
+ ;; Get the name of the document
+ (looking-at xml-name-regexp)
+ (let ((dtd (list (match-string 0) 'dtd))
+ type element end-pos)
+ (goto-char (match-end 0))
+ (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))
+ (error "XML: missing public id"))
+ (let ((pubid (match-string 1)))
+ (unless (or (re-search-forward "\\='\\([^']*\\)'" nil t)
+ (re-search-forward "\\=\"\\([^\"]*\\)\"" nil t))
+ (error "XML: missing system id"))
+ (push (list pubid (match-string 1) 'public) dtd)))
+ ((looking-at "SYSTEM\\s-+")
+ (goto-char (match-end 0))
+ (unless (or (re-search-forward "\\='\\([^']*\\)'" nil t)
+ (re-search-forward "\\=\"\\([^\"]*\\)\"" nil t))
+ (error "XML: missing system id"))
+ (push (list (match-string 1) 'system) dtd)))
+ (skip-syntax-forward " ")
+ (if (eq ?> (char-after))
+ (forward-char)
+ (skip-syntax-forward " ")
+ (if (not (eq (char-after) ?\[))
+ (error "XML: bad DTD")
+ (forward-char)
+ ;; Parse the rest of the DTD
+ ;; Fixme: Deal with ENTITY, ATTLIST, NOTATION, PIs.
+ (while (not (looking-at "\\s-*\\]"))
+ (skip-syntax-forward " ")
+ (cond
+
+ ;; Translation of rule [45] of XML specifications
+ ((looking-at
+ "<!ELEMENT\\s-+\\([[:alnum:].%;]+\\)\\s-+\\([^>]+\\)>")
+
+ (setq element (match-string 1)
+ 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]*$" type) ;; children ([47])
+ (setq type (xml-parse-elem-type (match-string 1 type))))
+ ((string-match "^%[^;]+;[ \t\n\r]*$" type) ;; substitution
+ nil)
+ (t
+ (error "XML: Invalid element type in the DTD")))
+
+ ;; rule [45]: the element declaration must be unique
+ (if (assoc element dtd)
+ (error "XML: element declarations must be unique in a DTD (<%s>)"
+ element))
+
+ ;; Store the element in the DTD
+ (push (list element type) dtd)
+ (goto-char end-pos))
+ ((looking-at "<!--")
+ (search-forward "-->"))
+
+ (t
+ (error "XML: Invalid DTD item")))
+
+ ;; Skip the end of the DTD
+ (search-forward ">"))))
+ (nreverse dtd)))
(defun xml-parse-elem-type (string)
- "Convert a STRING for an element type into an elisp structure."
+ "Convert element type STRING into a Lisp structure."
(let (elem modifier)
(if (string-match "(\\([^)]+\\))\\([+*?]?\\)" string)
(setq elem (match-string 1 string)
modifier (match-string 2 string))
(if (string-match "|" elem)
- (set 'elem (append '(choice)
+ (setq elem (cons 'choice
(mapcar 'xml-parse-elem-type
(split-string elem "|"))))
(if (string-match "," elem)
- (set 'elem (append '(seq)
+ (setq elem (cons 'seq
(mapcar 'xml-parse-elem-type
- (split-string elem ","))))
- )))
- (if (string-match "[ \t\n]*\\([^+*?]+\\)\\([+*?]?\\)" string)
- (setq elem (match-string 1 string)
+ (split-string elem ",")))))))
+ (if (string-match "[ \t\n\r]*\\([^+*?]+\\)\\([+*?]?\\)" string)
+ (setq elem (match-string 1 string)
modifier (match-string 2 string))))
- (if (and (stringp elem)
- (string= elem "#PCDATA"))
- (set 'elem 'pcdata))
-
- (cond
- ((string= modifier "+")
- (list '+ elem))
- ((string= modifier "*")
- (list '* elem))
- ((string= modifier "?")
- (list '? elem))
- (t
- elem))))
+ (if (and (stringp elem) (string= elem "#PCDATA"))
+ (setq elem 'pcdata))
+ (cond
+ ((string= modifier "+")
+ (list '+ elem))
+ ((string= modifier "*")
+ (list '* elem))
+ ((string= modifier "?")
+ (list '\? elem))
+ (t
+ elem))))
;;*******************************************************************
;;**
;;**
;;*******************************************************************
+(eval-when-compile
+ (defvar str)) ; dynamic from replace-regexp-in-string
+
+;; Fixme: Take declared entities from the DTD when they're available.
+(defun xml-substitute-entity (match)
+ "Subroutine of xml-substitute-special."
+ (save-match-data
+ (let ((match1 (match-string 1 str)))
+ (cond ((string= match1 "lt") "<")
+ ((string= match1 "gt") ">")
+ ((string= match1 "apos") "'")
+ ((string= match1 "quot") "\"")
+ ((string= match1 "amp") "&")
+ ((and (string-match "#\\([0-9]+\\)" match1)
+ (let ((c (decode-char
+ 'ucs
+ (string-to-number (match-string 1 match1)))))
+ (if c (string c))))) ; else unrepresentable
+ ((and (string-match "#x\\([[:xdigit:]]+\\)" match1)
+ (let ((c (decode-char
+ 'ucs
+ (string-to-number (match-string 1 match1) 16))))
+ (if c (string c)))))
+ ;; Default to asis. Arguably, unrepresentable code points
+ ;; might be best replaced with U+FFFD.
+ (t match)))))
+
(defun xml-substitute-special (string)
- "Return STRING, after subsituting special XML sequences."
- (while (string-match "&" string)
- (set 'string (replace-match "&" t nil string)))
- (while (string-match "<" string)
- (set 'string (replace-match "<" t nil string)))
- (while (string-match ">" string)
- (set 'string (replace-match ">" t nil string)))
- (while (string-match "'" string)
- (set 'string (replace-match "'" t nil string)))
- (while (string-match """ string)
- (set 'string (replace-match "\"" t nil string)))
- string)
+ "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;" or
+ ;; "&amp;" won't DTRT.
+ (replace-regexp-in-string "&\\([^;]+\\);"
+ #'xml-substitute-entity string t t))
;;*******************************************************************
;;**
;;*******************************************************************
(defun xml-debug-print (xml)
- (while xml
- (xml-debug-print-internal (car xml) "")
- (set 'xml (cdr xml)))
- )
+ (dolist (node xml)
+ (xml-debug-print-internal node "")))
-(defun xml-debug-print-internal (xml &optional indent-string)
+(defun xml-debug-print-internal (xml indent-string)
"Outputs the XML tree in the current buffer.
-The first line indented with INDENT-STRING."
+The first line is indented with INDENT-STRING."
(let ((tree xml)
attlist)
- (unless indent-string
- (set 'indent-string ""))
-
- (insert indent-string "<" (symbol-name (xml-node-name tree)))
-
+ (insert indent-string ?< (symbol-name (xml-node-name tree)))
+
;; output the attribute list
- (set 'attlist (xml-node-attributes tree))
+ (setq attlist (xml-node-attributes tree))
(while attlist
- (insert " ")
- (insert (symbol-name (caar attlist)) "=\"" (cdar attlist) "\"")
- (set 'attlist (cdr attlist)))
-
- (insert ">")
-
- (set 'tree (xml-node-children tree))
+ (insert ?\ (symbol-name (caar attlist)) "=\"" (cdar attlist) ?\")
+ (setq attlist (cdr attlist)))
+
+ (insert ?>)
+
+ (setq tree (xml-node-children tree))
;; output the children
- (while tree
+ (dolist (node tree)
(cond
- ((listp (car tree))
- (insert "\n")
- (xml-debug-print-internal (car tree) (concat indent-string " "))
- )
- ((stringp (car tree))
- (insert (car tree))
- )
+ ((listp node)
+ (insert ?\n)
+ (xml-debug-print-internal node (concat indent-string " ")))
+ ((stringp node) (insert node))
(t
- (error "Invalid XML tree")))
- (set 'tree (cdr tree))
- )
+ (error "Invalid XML tree"))))
- (insert "\n" indent-string
- "</" (symbol-name (xml-node-name xml)) ">")
- ))
+ (insert ?\n indent-string
+ ?< ?/ (symbol-name (xml-node-name xml)) ?>)))
(provide 'xml)
+;;; arch-tag: 5864b283-5a68-4b59-a20d-36a72b353b9b
;;; xml.el ends here