saxutils.py

Python Development Environment (Python IDE plugin for Eclipse). Features editor, code completion, re

"""
A library of useful helper classes to the saxlib classes, for the
convenience of application and driver writers.

$Id$
"""

import types, sys, urllib, urlparse, os, string
import handler, _exceptions, xmlreader

try:
    _StringTypes = [types.StringType, types.UnicodeType]
except AttributeError: # 1.5 compatibility:UnicodeType not defined
    _StringTypes = [types.StringType]

def escape(data, entities={}):
    """Escape &, <, and > in a string of data.

    You can escape other strings of data by passing a dictionary as
    the optional entities parameter.  The keys and values must all be
    strings; each key will be replaced with its corresponding value.
    """
    data = string.replace(data, "&", "&amp;")
    data = string.replace(data, "<", "&lt;")
    data = string.replace(data, ">", "&gt;")
    for chars, entity in entities.items():
        data = string.replace(data, chars, entity)
    return data

# --- DefaultHandler

class DefaultHandler(handler.EntityResolver, handler.DTDHandler,
                     handler.ContentHandler, handler.ErrorHandler):
    """Default base class for SAX2 event handlers. Implements empty
    methods for all callback methods, which can be overridden by
    application implementors. Replaces the deprecated SAX1 HandlerBase
    class."""

# --- Location

class Location:
    """Represents a location in an XML entity. Initialized by being passed
    a locator, from which it reads off the current location, which is then
    stored internally."""

    def __init__(self, locator):
        self.__col = locator.getColumnNumber()
        self.__line = locator.getLineNumber()
        self.__pubid = locator.getPublicId()
        self.__sysid = locator.getSystemId()

    def getColumnNumber(self):
        return self.__col

    def getLineNumber(self):
        return self.__line

    def getPublicId(self):
        return self.__pubid

    def getSystemId(self):
        return self.__sysid

# --- ErrorPrinter

class ErrorPrinter:
    "A simple class that just prints error messages to standard out."

    def __init__(self, level=0, outfile=sys.stderr):
        self._level = level
        self._outfile = outfile

    def warning(self, exception):
        if self._level <= 0:
            self._outfile.write("WARNING in %s: %s\n" %
                               (self.__getpos(exception),
                                exception.getMessage()))

    def error(self, exception):
        if self._level <= 1:
            self._outfile.write("ERROR in %s: %s\n" %
                               (self.__getpos(exception),
                                exception.getMessage()))

    def fatalError(self, exception):
        if self._level <= 2:
            self._outfile.write("FATAL ERROR in %s: %s\n" %
                               (self.__getpos(exception),
                                exception.getMessage()))

    def __getpos(self, exception):
        if isinstance(exception, _exceptions.SAXParseException):
            return "%s:%s:%s" % (exception.getSystemId(),
                                 exception.getLineNumber(),
                                 exception.getColumnNumber())
        else:
            return "<unknown>"

# --- ErrorRaiser

class ErrorRaiser:
    "A simple class that just raises the exceptions it is passed."

    def __init__(self, level = 0):
        self._level = level

    def error(self, exception):
        if self._level <= 1:
            raise exception

    def fatalError(self, exception):
        if self._level <= 2:
            raise exception

    def warning(self, exception):
        if self._level <= 0:
            raise exception

# --- AttributesImpl now lives in xmlreader
from xmlreader import AttributesImpl

# --- XMLGenerator is the SAX2 ContentHandler for writing back XML
try:
    import codecs
    def _outputwrapper(stream,encoding):
        writerclass = codecs.lookup(encoding)[3]
        return writerclass(stream)
except ImportError: # 1.5 compatibility: fall back to do-nothing
    def _outputwrapper(stream,encoding):
        return stream

class XMLGenerator(handler.ContentHandler):

    def __init__(self, out=None, encoding="iso-8859-1"):
        if out is None:
            import sys
            out = sys.stdout
        handler.ContentHandler.__init__(self)
        self._out = _outputwrapper(out,encoding)
        self._ns_contexts = [{}] # contains uri -> prefix dicts
        self._current_context = self._ns_contexts[-1]
        self._undeclared_ns_maps = []
        self._encoding = encoding

    # ContentHandler methods

    def startDocument(self):
        self._out.write('<?xml version="1.0" encoding="%s"?>\n' %
                        self._encoding)

    def startPrefixMapping(self, prefix, uri):
        self._ns_contexts.append(self._current_context.copy())
        self._current_context[uri] = prefix
        self._undeclared_ns_maps.append((prefix, uri))

    def endPrefixMapping(self, prefix):
        self._current_context = self._ns_contexts[-1]
        del self._ns_contexts[-1]

    def startElement(self, name, attrs):
        self._out.write('<' + name)
        for (name, value) in attrs.items():
            self._out.write(' %s="%s"' % (name, escape(value)))
        self._out.write('>')

    def endElement(self, name):
        self._out.write('</%s>' % name)

    def startElementNS(self, name, qname, attrs):
        if name[0] is None:
            name = name[1]
        elif self._current_context[name[0]] is None:
            # default namespace
            name = name[1]
        else:
            name = self._current_context[name[0]] + ":" + name[1]
        self._out.write('<' + name)

        for k,v in self._undeclared_ns_maps:
            if k is None:
                self._out.write(' xmlns="%s"' % v)
            else:
                self._out.write(' xmlns:%s="%s"' % (k,v))
        self._undeclared_ns_maps = []

        for (name, value) in attrs.items():
            name = self._current_context[name[0]] + ":" + name[1]
            self._out.write(' %s="%s"' % (name, escape(value)))
        self._out.write('>')

    def endElementNS(self, name, qname):
        # XXX: if qname is not None, we better use it.
        # Python 2.0b2 requires us to use the recorded prefix for
        # name[0], though
        if name[0] is None:
            qname = name[1]
        elif self._current_context[name[0]] is None:
            qname = name[1]
        else:
            qname = self._current_context[name[0]] + ":" + name[1]
        self._out.write('</%s>' % qname)

    def characters(self, content):
        self._out.write(escape(content))

    def ignorableWhitespace(self, content):
        self._out.write(content)

    def processingInstruction(self, target, data):
        self._out.write('<?%s %s?>' % (target, data))

# --- ContentGenerator is the SAX1 DocumentHandler for writing back XML
class ContentGenerator(XMLGenerator):

    def characters(self, str, start, end):
        # In SAX1, characters receives start and end; in SAX2, it receives
        # a string. For plain strings, we may want to use a buffer object.
        return XMLGenerator.characters(self, str[start:start+end])

# --- XMLFilterImpl
class XMLFilterBase(xmlreader.XMLReader):
    """This class is designed to sit between an XMLReader and the
    client application's event handlers.  By default, it does nothing
    but pass requests up to the reader and events on to the handlers
    unmodified, but subclasses can override specific methods to modify
    the event stream or the configuration requests as they pass
    through."""

    # ErrorHandler methods

    def error(self, exception):
        self._err_handler.error(exception)

    def fatalError(self, exception):
        self._err_handler.fatalError(exception)

    def warning(self, exception):
        self._err_handler.warning(exception)

    # ContentHandler methods

    def setDocumentLocator(self, locator):
        self._cont_handler.setDocumentLocator(locator)

    def startDocument(self):
        self._cont_handler.startDocument()

    def endDocument(self):
        self._cont_handler.endDocument()

    def startPrefixMapping(self, prefix, uri):
        self._cont_handler.startPrefixMapping(prefix, uri)

    def endPrefixMapping(self, prefix):
        self._cont_handler.endPrefixMapping(prefix)

    def startElement(self, name, attrs):
        self._cont_handler.startElement(name, attrs)

    def endElement(self, name):
        self._cont_handler.endElement(name)

    def startElementNS(self, name, qname, attrs):
        self._cont_handler.startElementNS(name, qname, attrs)

    def endElementNS(self, name, qname):
        self._cont_handler.endElementNS(name, qname)

    def characters(self, content):
        self._cont_handler.characters(content)

    def ignorableWhitespace(self, chars):
        self._cont_handler.ignorableWhitespace(chars)

    def processingInstruction(self, target, data):
        self._cont_handler.processingInstruction(target, data)

    def skippedEntity(self, name):
        self._cont_handler.skippedEntity(name)

    # DTDHandler methods

    def notationDecl(self, name, publicId, systemId):
        self._dtd_handler.notationDecl(name, publicId, systemId)

    def unparsedEntityDecl(self, name, publicId, systemId, ndata):
        self._dtd_handler.unparsedEntityDecl(name, publicId, systemId, ndata)

    # EntityResolver methods

    def resolveEntity(self, publicId, systemId):
        self._ent_handler.resolveEntity(publicId, systemId)

    # XMLReader methods

    def parse(self, source):
        self._parent.setContentHandler(self)
        self._parent.setErrorHandler(self)
        self._parent.setEntityResolver(self)
        self._parent.setDTDHandler(self)
        self._parent.parse(source)

    def setLocale(self, locale):
        self._parent.setLocale(locale)

    def getFeature(self, name):
        return self._parent.getFeature(name)

    def setFeature(self, name, state):
        self._parent.setFeature(name, state)

    def getProperty(self, name):
        return self._parent.getProperty(name)

    def setProperty(self, name, value):
        self._parent.setProperty(name, value)

# FIXME: remove this backward compatibility hack when not needed anymore
XMLFilterImpl = XMLFilterBase

# --- BaseIncrementalParser

class BaseIncrementalParser(xmlreader.IncrementalParser):
    """This class implements the parse method of the XMLReader
    interface using the feed, close and reset methods of the
    IncrementalParser interface as a convenience to SAX 2.0 driver
    writers."""

    def parse(self, source):
        source = prepare_input_source(source)
        self.prepareParser(source)