(top level comment): Updated to reflect the fact that

white spaces are relevant in the XML file
(xml-parse-file): Do not kill an existing Emacs buffer if the file
to parse was already edited. This allows for on-the-fly analysis
of XML files
(xml-parse-tag): Check that the casing is the same in the start
tag and end tag, since XML is case-sensitive.  Allows for spaces
in the end tag, after the name of the tag.
(xml-parse-attlist): Allow for the character '-' in the name of
attributes, as in the standard http-equiv attribute Do not save
the properties in the XML tree, since they are not relevant

diff --git a/lisp/ChangeLog b/lisp/ChangeLog
index bd9b2e314060d14e5fab91afc276951ed6043ff2..645f106e73f91a3cdd6be0fe42c17068aae243be 100644 (file)
--- a/lisp/ChangeLog
+++ b/lisp/ChangeLog
@@ -1,3 +1,17 @@
+2000-08-22  Emmanuel Briot  <briot@gnat.com>
+
+       * xml.el (top level comment): Updated to reflect the fact that
+       white spaces are relevant in the XML file
+       (xml-parse-file): Do not kill an existing Emacs buffer if the file
+       to parse was already edited. This allows for on-the-fly analysis
+       of XML files
+       (xml-parse-tag): Check that the casing is the same in the start
+       tag and end tag, since XML is case-sensitive.  Allows for spaces
+       in the end tag, after the name of the tag.
+       (xml-parse-attlist): Allow for the character '-' in the name of
+       attributes, as in the standard http-equiv attribute Do not save
+       the properties in the XML tree, since they are not relevant
+
 2000-12-21  Stefan Monnier  <monnier@cs.yale.edu>
 
        * generic.el (generic-read-type): Undo last change, inline into

diff --git a/lisp/xml.el b/lisp/xml.el
index a7d2ba48ef536d61ef6e8561e0e2aee0a9b53bd9..25851e2a9ea174d9eb7cf7f99b4134a94fe133cb 100644 (file)
--- a/lisp/xml.el
+++ b/lisp/xml.el
@@ -39,9 +39,9 @@
 ;; in the XML file.
 ;;
 ;; 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
@@ -118,15 +118,24 @@ An empty string is returned if the attribute was not found."
 
 (defun xml-parse-file (file &optional parse-dtd)
   "Parse the well-formed XML FILE.
+If FILE is already edited, this will keep the buffer alive.
 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))
+  (let ((keep))
+    (if (get-file-buffer file)
+       (progn
+         (set-buffer (get-file-buffer file))
+         (setq keep (point)))
+      (find-file file))
+    
+    (let ((xml (xml-parse-region (point-min)
+                                (point-max)
+                                (current-buffer)
+                                parse-dtd)))
+      (if keep
+         (goto-char keep)
+       (kill-buffer (current-buffer)))
+      xml)))
 
 (defun xml-parse-region (beg end &optional buffer parse-dtd)
   "Parse the region from BEG to END in BUFFER.
@@ -206,6 +215,7 @@ Returns one of:
    ((looking-at "<\\([^/> \t\n]+\\)")
     (let* ((node-name (match-string 1))
           (children (list (intern node-name)))
+          (case-fold-search nil) ;; XML is case-sensitive
           pos)
       (goto-char (match-end 1))
 
@@ -224,13 +234,15 @@ Returns one of:
            (progn
              (forward-char 1)
              (skip-chars-forward " \t\n")
-             (while (not (looking-at (concat "</" node-name ">")))
+             ;;  Now check that we have the right end-tag. Note that this one might
+             ;;  contain spaces after the tag name
+             (while (not (looking-at (concat "</" node-name "[ \t\n]*>")))
                (cond
                 ((looking-at "</")
                  (error (concat
                          "XML: invalid syntax -- invalid end tag (expecting "
                          node-name
-                         ")")))
+                         ") at pos " (number-to-string (point)))))
                 ((= (char-after) ?<)
                  (set 'children (append children (list (xml-parse-tag end)))))
                 (t
@@ -269,7 +281,7 @@ 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]*")
+    (while (looking-at "\\([a-zA-Z_:][-a-zA-Z0-9._:]*\\)[ \t\n]*=[ \t\n]*")
       (set 'name (intern (match-string 1)))
       (goto-char (match-end 0))
 
@@ -284,7 +296,7 @@ Leaves the point on the first non-blank character after the tag."
          (error "XML: each attribute must be unique within an element."))
       
       (set 'attlist (append attlist
-                           (list (cons name (match-string 1)))))
+                           (list (cons name (match-string-no-properties 1)))))
       (goto-char (match-end 0))
       (skip-chars-forward " \t\n")
       (if (> (point) end)