xml version 1.1.3: fix a minor bug in handling namespaces

[extfs.d.git] / xml

diff --git a/xml b/xml
index 8bc536f24e9a6a87839479514e42d8535c20aa7f..4822ab8c6005edc39c7995afe1fd856adcc62883 100755 (executable)
--- a/xml
+++ b/xml
@@ -13,8 +13,11 @@ For mc 4.7+ run this "cd" command in the Midnight Commander (in the "bindings"
 file the command is "%cd"): cd file/xml://; in older versions it is
 cd file#xml, where "file" is the name of your XML file.
 
+See detailed installation instructions at
+http://phdru.name/Software/mc/xml_INSTALL.html.
+
 The VFS represents tags as directories; the directories are numbered to
-distinguish tags with the same name; also numbering helps to sort tags by their
+distinguish tags with the same name; numbering also helps to sort tags by their
 order in XML instead of sorting them by name. Attributes, text nodes and
 comments are represented as text files; attributes are shown in a file named
 "attributes", attributes are listed in the file as name=value lines (I
@@ -26,7 +29,7 @@ read-only.
 Implementation based on minidom doesn't understand namespaces, it just shows
 them among other attributes. ElementTree-based implementation doesn't show
 namespaces at all. Implementation based on lxml.etree shows namespaces in a
-separate file "namespaces"; every child tag includes its parent's namespaces.
+separate file "namespaces".
 
 It is useful to have a top-down view on an XML structure but it's especially
 convenient to extract text values from tags. One can get, for example, a
@@ -37,9 +40,9 @@ The VFS was inspired by a FUSE xmlfs: https://github.com/halhen/xmlfs
 
 """
 
-__version__ = "0.6.0"
+__version__ = "1.1.3"
 __author__ = "Oleg Broytman <phd@phdru.name>"
-__copyright__ = "Copyright (C) 2013 PhiloSoft Design"
+__copyright__ = "Copyright (C) 2013-2015 PhiloSoft Design"
 __license__ = "GPL"
 
 force_implementation = None  # Can be None for default choice,
@@ -123,6 +126,9 @@ class XmlVfs(object):
         self.parse()
 
     def list(self):
+        root_comments = self.get_root_comments()
+        if root_comments:
+            print "-r--r--r-- 1 user group %d Jan 1 00:00 text" % (len(root_comments))
         self._list(self.getroot())
 
     def _list(self, node, path=''):
@@ -205,6 +211,9 @@ class MiniDOMXmlVfs(XmlVfs):
     def getroot(self):
         return self.document
 
+    def get_root_comments(self):
+        return self.collect_text(self.document)
+
     def getchildren(self, node):
         return node.childNodes
 
@@ -250,7 +259,7 @@ if use_elementtree or use_lxml:
 if use_elementtree:
     class ElementTreeXmlVfs(CommonEtreeXmlVfs):
         def parse(self):
-            # Copied from http://effbot.org/zone/element-pi.ht
+            # Copied from http://effbot.org/zone/element-pi.htm
 
             class PIParser(ET.XMLTreeBuilder):
 
@@ -280,6 +289,14 @@ if use_elementtree:
         def getroot(self):
             return self.document.getroot()
 
+        def get_root_comments(self):
+            text_accumulator = []
+            for element in self.getroot():
+                if not self.istag(element):
+                    text = u"<!--%s-->" % element.text
+                    text_accumulator.append(text)
+            return '\n'.join(text_accumulator).encode(default_encoding, "replace")
+
         def getlocalname(self, name):
             if name.startswith('{'):
                 name = name.split('}', 1)[1]  # Remove XML namespace
@@ -296,16 +313,36 @@ if use_lxml:
         def getroot(self):
             return [self.document.getroot()]
 
+        def get_root_comments(self):
+            text_accumulator = []
+            for element in self.document.getroot().itersiblings(tag=etree.Comment, preceding=True):
+                text = u"<!--%s-->" % element.text
+                text_accumulator.append(text)
+            return '\n'.join(text_accumulator).encode(default_encoding, "replace")
+
         def getlocalname(self, name):
             return etree.QName(name).localname
 
+        def _get_local_ns(self, node):
+            this_nsmap = node.nsmap
+            parent = node.getparent()
+            if parent is not None:
+                parent_nsmap = parent.nsmap
+                for key in parent_nsmap:
+                    if this_nsmap[key] == parent_nsmap[key]:
+                        del this_nsmap[key]
+            return this_nsmap
+
         def has_ns(self, node):
-            return bool(node.nsmap)
+            return bool(self._get_local_ns(node))
 
         def ns2text(self, node):
             ns_accumulator = []
-            for name, value in node.nsmap.items():
-                name = name.encode(default_encoding, "replace")
+            for name, value in self._get_local_ns(node).items():
+                if name:
+                    name = name.encode(default_encoding, "replace")
+                else:
+                    name = 'xmlns'
                 value = value.encode(default_encoding, "replace")
                 ns_accumulator.append("%s=%s" % (name, value))
             return '\n'.join(ns_accumulator)
@@ -360,13 +397,16 @@ def mcxml_copyout():
             xml_error('There are no attributes')
 
     elif path_comp == 'namespaces':
-        if xmlvfs.has_ns(node):
+        if xmlvfs.supports_namespaces and xmlvfs.has_ns(node):
             text = xmlvfs.ns2text(node)
         else:
-            xml_error('There are no attributes')
+            xml_error('There are no namespaces')
 
     elif path_comp == 'text':
-        text = xmlvfs.collect_text(node)
+        if '/' in xml_filename:
+            text = xmlvfs.collect_text(node)
+        else:
+            text = xmlvfs.get_root_comments()
 
     else:
         xml_error('Unknown file')