X-Git-Url: https://git.phdru.name/?p=extfs.d.git;a=blobdiff_plain;f=xml;h=6cf2276f9c3fbcdcd0adcf0dce5fb658ea126644;hp=1c98c86fbc6a97ed0bd7a1d5ecd074736757d85f;hb=af76dda2d3c72d5712ae4559c38da2829398a27b;hpb=2782d4a3a8ed6f10f461935136e0ba7f2d998064

diff --git a/xml b/xml
index 1c98c86..6cf2276 100755
--- a/xml
+++ b/xml
@@ -4,36 +4,69 @@
 The script requires Midnight Commander 3.1+
 (http://www.midnight-commander.org/), Python 2.4+ (http://www.python.org/).
 
-For mc 4.7+ put the script in $HOME/.mc/extfs.d.
+For mc 4.7+ just put the script in $HOME/[.local/share/].mc/extfs.d.
 For older versions put it in /usr/[local/][lib|share]/mc/extfs
 and add a line "xml" to the /usr/[local/][lib|share]/mc/extfs/extfs.ini.
 Make the script executable.
 
-Run this "cd" command in the Midnight Commander (in the "bindings" file the
-command is "%cd"): cd file.xml#xml, where "file.xml" is the name of your xml
-file.
+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.
 
 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
 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
-deliberately ignore a small chance there is a newline character in values). The
-filesystem is read-only.
+deliberately ignore a small chance of newline characters in values); names and
+values are reencoded to the console encoding. Text nodes and comments are
+collected in a file named "text", stripped and reencoded. The filesystem is
+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".
+
+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
+base64-encoded image - just walk down the VFS to the tag's directory and copy
+its text file to a real file.
 
 The VFS was inspired by a FUSE xmlfs: https://github.com/halhen/xmlfs
 
 """
 
-__version__ = "0.2.0"
+__version__ = "0.6.1"
 __author__ = "Oleg Broytman <phd@phdru.name>"
 __copyright__ = "Copyright (C) 2013 PhiloSoft Design"
 __license__ = "GPL"
 
+force_implementation = None  # Can be None for default choice,
+                             # 'lxml', 'elementtree' or 'minidom'
+
+use_minidom = True
+use_elementtree = False
+use_lxml = False
+
 import math
 import sys
 import xml.dom.minidom
 
+try:
+    import xml.etree.ElementTree as ET
+except ImportError:
+    pass
+else:
+    use_elementtree = True
+
+try:
+    import lxml.etree as etree
+except ImportError:
+    pass
+else:
+    use_lxml = True
+
 try:
    import locale
    use_locale = True
@@ -71,7 +104,7 @@ XML Virtual FileSystem for Midnight Commander version %s
 Author: %s
 %s
 
-This is not a program. Put the script in $HOME/.mc/extfs.d or
+This is not a program. Put the script in $HOME/[.local/share/].mc/extfs.d or
 /usr/[local/][lib|share]/mc/extfs. For more information read the source!""",
    __version__, __author__, __copyright__
 )
@@ -80,80 +113,273 @@ This is not a program. Put the script in $HOME/.mc/extfs.d or
 
 locale.setlocale(locale.LC_ALL, '')
 
-def _attrs2text(attrs):
-    attrs = [attrs.item(i) for i in range (attrs.length)]
-    return '\n'.join(["%s=%s" %
-        (a.name.encode(default_encoding, "replace"),
-        a.value.encode(default_encoding, "replace"))
-        for a in attrs])
-
-def _list(node, path=''):
-    childNodes = node.childNodes
-    n = 0
-    for element in childNodes:
-        if element.localName:
-            n += 1
-    if n:
-        width = int(math.log10(n))+1
-        template = "%%0%dd" % width
-    else:
-        template = "%d"
-    n = 0
-    for element in childNodes:
-        if element.localName:
+
+class XmlVfs(object):
+    """Abstract base class"""
+
+    supports_namespaces = False
+
+    def __init__(self):
+        self.parse()
+
+    def list(self):
+        self._list(self.getroot())
+
+    def _list(self, node, path=''):
+        n = len(self.getchildren(node))
+        if n:
+            width = int(math.log10(n)) + 1
+            template = "%%0%dd" % width
+        else:
+            template = "%d"
+        n = 0
+        for element in self.getchildren(node):
+            if not self.istag(element):
+                continue
             n += 1
+            tag = self.getlocalname(self.gettag(element))
             if path:
-                subpath = '%s/%s %s' % (path, template % n, element.localName)
+                subpath = '%s/%s %s' % (path, template % n, tag)
             else:
-                subpath = '%s %s' % (template % n, element.localName)
+                subpath = '%s %s' % (template % n, tag)
             subpath_encoded = subpath.encode(default_encoding, "replace")
-            print "dr--r--r-- 1 user group 0 Jan 1 00:00 %s" % subpath_encoded
-            attrs = element.attributes
-            if attrs:
-                attr_text = _attrs2text(attrs)
+            print "dr-xr-xr-x 1 user group 0 Jan 1 00:00 %s" % subpath_encoded
+            if self.getattrs(element):
+                attr_text = self.attrs2text(element)
                 print "-r--r--r-- 1 user group %d Jan 1 00:00 %s/attributes" % (
                     len(attr_text), subpath_encoded)
-            _list(element, subpath)
+            if self.supports_namespaces and self.has_ns(element):
+                ns_text = self.ns2text(element)
+                print "-r--r--r-- 1 user group %d Jan 1 00:00 %s/namespaces" % (
+                    len(ns_text), subpath_encoded)
+            text = self.collect_text(element)
+            if text:
+                print "-r--r--r-- 1 user group %d Jan 1 00:00 %s/text" % (
+                    len(text), subpath_encoded)
+            self._list(element, subpath)
+
+    def get_tag_node(self, node, i):
+        n = 0
+        for element in self.getchildren(node):
+            if self.istag(element):
+                n += 1
+                if n == i:
+                    return element
+        xml_error('There are less than %d nodes' % i)
+
+    def attrs2text(self, node):
+        attr_accumulator = []
+        for name, value in self.getattrs(node):
+            name = self.getlocalname(name).encode(default_encoding, "replace")
+            value = value.encode(default_encoding, "replace")
+            attr_accumulator.append("%s=%s" % (name, value))
+        return '\n'.join(attr_accumulator)
+
+    def has_ns(self, node):
+        return False
+
+
+class MiniDOMXmlVfs(XmlVfs):
+    def parse(self):
+        self.document = xml.dom.minidom.parse(sys.argv[2])
+
+    def getattrs(self, node):
+        attrs = node.attributes
+        attrs = [attrs.item(i) for i in range(attrs.length)]
+        return [(a.name, a.value) for a in attrs]
+
+    def collect_text(self, node):
+        text_accumulator = []
+        for element in node.childNodes:
+            if element.localName:
+                continue
+            elif element.nodeType == element.COMMENT_NODE:
+                text = u"<!--%s-->" % element.nodeValue
+            elif element.nodeType == element.TEXT_NODE:
+                text = element.nodeValue.strip()
+            else:
+                xml_error("Unknown node type %d" % element.nodeType)
+            if text: text_accumulator.append(text)
+        return '\n'.join(text_accumulator).encode(default_encoding, "replace")
+
+    def getroot(self):
+        return self.document
+
+    def getchildren(self, node):
+        return node.childNodes
+
+    def gettag(self, node):
+        return node.localName
+
+    def istag(self, node):
+        return bool(node.localName)
+
+    def getlocalname(self, name):
+        return name
+
+
+if use_elementtree or use_lxml:
+    class CommonEtreeXmlVfs(XmlVfs):
+        def getattrs(self, node):
+            return node.attrib.items()
+
+        def collect_text(self, node):
+            text_accumulator = []
+            if node.text:
+                text = node.text.strip()
+                if text: text_accumulator.append(text)
+            for element in node:
+                if not self.istag(element):
+                    text = u"<!--%s-->" % element.text
+                    text_accumulator.append(text)
+            if node.tail:
+                text = node.tail.strip()
+                if text: text_accumulator.append(text)
+            return '\n'.join(text_accumulator).encode(default_encoding, "replace")
+
+        def getchildren(self, node):
+            return list(node)
+
+        def gettag(self, node):
+            return node.tag
+
+        def istag(self, node):
+            return isinstance(node.tag, basestring)
+
+
+if use_elementtree:
+    class ElementTreeXmlVfs(CommonEtreeXmlVfs):
+        def parse(self):
+            # Copied from http://effbot.org/zone/element-pi.ht
+
+            class PIParser(ET.XMLTreeBuilder):
+
+                def __init__(self):
+                    ET.XMLTreeBuilder.__init__(self)
+                    # assumes ElementTree 1.2.X
+                    self._parser.CommentHandler = self.handle_comment
+                    self._parser.ProcessingInstructionHandler = self.handle_pi
+                    self._target.start("document", {})
+
+                def close(self):
+                    self._target.end("document")
+                    return ET.XMLTreeBuilder.close(self)
+
+                def handle_comment(self, data):
+                    self._target.start(ET.Comment, {})
+                    self._target.data(data)
+                    self._target.end(ET.Comment)
+
+                def handle_pi(self, target, data):
+                    self._target.start(ET.PI, {})
+                    self._target.data(target + " " + data)
+                    self._target.end(ET.PI)
+
+            self.document = ET.parse(sys.argv[2], PIParser())
+
+        def getroot(self):
+            return self.document.getroot()
+
+        def getlocalname(self, name):
+            if name.startswith('{'):
+                name = name.split('}', 1)[1]  # Remove XML namespace
+            return name
+
+
+if use_lxml:
+    class LxmlEtreeXmlVfs(CommonEtreeXmlVfs):
+        supports_namespaces = True
+
+        def parse(self):
+            self.document = etree.parse(sys.argv[2])
+
+        def getroot(self):
+            return [self.document.getroot()]
+
+        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:
+                parents_nsmap = parent.nsmap
+                for key in parents_nsmap:
+                    del this_nsmap[key]
+            return this_nsmap
+
+        def has_ns(self, node):
+            return bool(self._get_local_ns(node))
+
+        def ns2text(self, node):
+            ns_accumulator = []
+            for name, value in self._get_local_ns(node).items():
+                name = name.encode(default_encoding, "replace")
+                value = value.encode(default_encoding, "replace")
+                ns_accumulator.append("%s=%s" % (name, value))
+            return '\n'.join(ns_accumulator)
+
+
+def build_xmlvfs():
+    if force_implementation is None:
+        if use_lxml:
+            return LxmlEtreeXmlVfs()
+        elif use_elementtree:
+            return ElementTreeXmlVfs()
+        else:
+            return MiniDOMXmlVfs()
+    elif force_implementation == 'minidom':
+        return MiniDOMXmlVfs()
+    elif force_implementation == 'elementtree':
+        return ElementTreeXmlVfs()
+    elif force_implementation == 'lxml':
+        return LxmlEtreeXmlVfs()
+    else:
+        raise ValueError('Unknown implementation "%s", expected "minidom", "elementtree" or "lxml"' % force_implementation)
+
 
 def mcxml_list():
     """List the entire VFS"""
 
-    dom = xml.dom.minidom.parse(sys.argv[2])
-    _list(dom)
+    xmlvfs = build_xmlvfs()
+    xmlvfs.list()
 
 
-def _get_child_node(node, i):
-    n = 0
-    for element in node.childNodes:
-        if element.localName:
-            n += 1
-            if n == i:
-                return element
-    xml_error('There are less than %d nodes' % i)
-
 def mcxml_copyout():
     """Extract a file from the VFS"""
 
-    node = xml.dom.minidom.parse(sys.argv[2])
+    xmlvfs = build_xmlvfs()
     xml_filename = sys.argv[3]
     real_filename = sys.argv[4]
 
+    node = xmlvfs.getroot()
     for path_comp in xml_filename.split('/'):
         if ' ' in path_comp:
             i = int(path_comp.split(' ', 1)[0])
-            node = _get_child_node(node, i)
-        elif path_comp == 'attributes':
+            node = xmlvfs.get_tag_node(node, i)
+        elif path_comp in ('attributes', 'namespaces', 'text'):
             break
         else:
             xml_error('Unknown file')
 
     if path_comp == 'attributes':
-        attrs = node.attributes
-        if attrs:
-            text = _attrs2text(attrs)
+        if xmlvfs.getattrs(node):
+            text = xmlvfs.attrs2text(node)
         else:
             xml_error('There are no attributes')
 
+    elif path_comp == 'namespaces':
+        if xmlvfs.supports_namespaces and xmlvfs.has_ns(element):
+            text = xmlvfs.ns2text(node)
+        else:
+            xml_error('There are no namespaces')
+
+    elif path_comp == 'text':
+        text = xmlvfs.collect_text(node)
+
+    else:
+        xml_error('Unknown file')
+
     outfile = open(real_filename, 'w')
     outfile.write(text)
     outfile.close()