diff -r c4e6e560e6f5 Lib/test/test_sax.py --- a/Lib/test/test_sax.py Mon Feb 04 18:29:47 2013 +0200 +++ b/Lib/test/test_sax.py Mon Feb 04 21:34:11 2013 +0200 @@ -9,15 +9,15 @@ # don't try to test this module if we cannot create a parser raise ImportError("no XML parsers available") from xml.sax.saxutils import XMLGenerator, escape, unescape, quoteattr, \ - XMLFilterBase + XMLFilterBase, prepare_input_source from xml.sax.expatreader import create_parser from xml.sax.handler import feature_namespaces from xml.sax.xmlreader import InputSource, AttributesImpl, AttributesNSImpl -from io import StringIO +from io import BytesIO, StringIO import os.path import shutil from test import support -from test.support import findfile, run_unittest +from test.support import findfile, run_unittest, TESTFN import unittest TEST_XMLFILE = findfile("test.xml", subdir="xmltestdata") @@ -171,6 +171,81 @@ p = make_parser(['xml.parsers.no_such_parser']) +class PrepareInputSourceTest(unittest.TestCase): + + # Fixture methods + def setUp(self): + self.file = TESTFN + with open(self.file, "w") as tmp: + tmp.write("This was read from a file.") + + def tearDown(self): + support.unlink(self.file) + + def make_byte_stream(self): + return BytesIO(b"This is a byte stream.") + + def make_character_stream(self): + return StringIO("This is a character stream.") + + def checkContent(self, stream, content): + self.assertIsNotNone(stream) + self.assertEqual(stream.read(), content) + stream.close() + + # The tests + def test_character_stream(self): + '''If the source is an InputSource with a character stream, use it.''' + src = InputSource(self.file) + src.setCharacterStream(self.make_character_stream()) + prep = prepare_input_source(src) + self.assertIsNone(prep.getByteStream()) + self.checkContent(prep.getCharacterStream(), + "This is a character stream.") + + def test_byte_stream(self): + '''If the source is an InputSource that does not have a character + stream but does have a byte stream, use the byte stream.''' + src = InputSource(self.file) + src.setByteStream(self.make_byte_stream()) + prep = prepare_input_source(src) + self.assertIsNone(prep.getCharacterStream()) + self.checkContent(prep.getByteStream(), + b"This is a byte stream.") + + def test_system_id(self): + '''If the source is an InputSource that has neither a character + stream nor a byte stream, open the system ID.''' + src = InputSource(self.file) + prep = prepare_input_source(src) + self.assertIsNone(prep.getCharacterStream()) + self.checkContent(prep.getByteStream(), + b"This was read from a file.") + + def test_string(self): + '''If the source is a string, use it as a system ID and open it.''' + prep = prepare_input_source(self.file) + self.assertIsNone(prep.getCharacterStream()) + self.checkContent(prep.getByteStream(), + b"This was read from a file.") + + def test_binary_file(self): + '''If the source is a binary file-like object, use it as a byte + stream.''' + prep = prepare_input_source(self.make_byte_stream()) + self.assertIsNone(prep.getCharacterStream()) + self.checkContent(prep.getByteStream(), + b"This is a byte stream.") + + def test_text_file(self): + '''If the source is a text file-like object, use it as a character + stream.''' + prep = prepare_input_source(self.make_character_stream()) + self.assertIsNone(prep.getByteStream()) + self.checkContent(prep.getCharacterStream(), + "This is a character stream.") + + # ===== XMLGenerator start = '\n' @@ -478,26 +553,37 @@ # # =========================================================================== -with open(TEST_XMLFILE_OUT) as f: +with open(TEST_XMLFILE_OUT, encoding='iso-8859-1') as f: xml_test_out = f.read() class ExpatReaderTest(XmlTestBase): # ===== XMLReader support - def test_expat_file(self): + def test_expat_binary_file(self): parser = create_parser() result = StringIO() xmlgen = XMLGenerator(result) parser.setContentHandler(xmlgen) - with open(TEST_XMLFILE) as f: + with open(TEST_XMLFILE, 'rb') as f: + parser.parse(f) + + self.assertEqual(result.getvalue(), xml_test_out) + + def test_expat_text_file(self): + parser = create_parser() + result = StringIO() + xmlgen = XMLGenerator(result) + + parser.setContentHandler(xmlgen) + with open(TEST_XMLFILE, 'rt', encoding='iso-8859-1') as f: parser.parse(f) self.assertEqual(result.getvalue(), xml_test_out) @requires_nonascii_filenames - def test_expat_file_nonascii(self): + def test_expat_binary_file_nonascii(self): fname = support.TESTFN_UNICODE shutil.copyfile(TEST_XMLFILE, fname) self.addCleanup(support.unlink, fname) @@ -507,7 +593,7 @@ xmlgen = XMLGenerator(result) parser.setContentHandler(xmlgen) - parser.parse(open(fname)) + parser.parse(open(fname, 'rb')) self.assertEqual(result.getvalue(), xml_test_out) @@ -547,7 +633,7 @@ def resolveEntity(self, publicId, systemId): inpsrc = InputSource() - inpsrc.setByteStream(StringIO("")) + inpsrc.setByteStream(BytesIO(b"")) return inpsrc def test_expat_entityresolver(self): @@ -665,19 +751,32 @@ self.assertEqual(result.getvalue(), xml_test_out) - def test_expat_inpsource_stream(self): + def test_expat_inpsource_byte_stream(self): parser = create_parser() result = StringIO() xmlgen = XMLGenerator(result) parser.setContentHandler(xmlgen) inpsrc = InputSource() - with open(TEST_XMLFILE) as f: + with open(TEST_XMLFILE, 'rb') as f: inpsrc.setByteStream(f) parser.parse(inpsrc) self.assertEqual(result.getvalue(), xml_test_out) + def test_expat_inpsource_character_stream(self): + parser = create_parser() + result = StringIO() + xmlgen = XMLGenerator(result) + + parser.setContentHandler(xmlgen) + inpsrc = InputSource() + with open(TEST_XMLFILE, 'rt', encoding='iso-8859-1') as f: + inpsrc.setCharacterStream(f) + parser.parse(inpsrc) + + self.assertEqual(result.getvalue(), xml_test_out) + # ===== IncrementalParser support def test_expat_incremental(self): @@ -766,7 +865,7 @@ parser = create_parser() parser.setContentHandler(ContentHandler()) # do nothing source = InputSource() - source.setByteStream(StringIO("")) #ill-formed + source.setByteStream(BytesIO(b"")) #ill-formed name = "a file name" source.setSystemId(name) try: @@ -857,6 +956,7 @@ def test_main(): run_unittest(MakeParserTest, SaxutilsTest, + PrepareInputSourceTest, XmlgenTest, ExpatReaderTest, ErrorReportingTest, diff -r c4e6e560e6f5 Lib/test/xmltestdata/test.xml --- a/Lib/test/xmltestdata/test.xml Mon Feb 04 18:29:47 2013 +0200 +++ b/Lib/test/xmltestdata/test.xml Mon Feb 04 21:34:11 2013 +0200 @@ -1,4 +1,4 @@ - + Introduction to XSL

Introduction to XSL

@@ -110,6 +110,6 @@ - +µ diff -r c4e6e560e6f5 Lib/test/xmltestdata/test.xml.out --- a/Lib/test/xmltestdata/test.xml.out Mon Feb 04 18:29:47 2013 +0200 +++ b/Lib/test/xmltestdata/test.xml.out Mon Feb 04 21:34:11 2013 +0200 @@ -110,6 +110,6 @@ - +µ \ No newline at end of file diff -r c4e6e560e6f5 Lib/xml/sax/expatreader.py --- a/Lib/xml/sax/expatreader.py Mon Feb 04 18:29:47 2013 +0200 +++ b/Lib/xml/sax/expatreader.py Mon Feb 04 21:34:11 2013 +0200 @@ -219,9 +219,14 @@ self._parsing = 0 # break cycle created by expat handlers pointing to our methods self._parser = None - bs = self._source.getByteStream() - if bs is not None: - bs.close() + try: + file = self._source.getCharacterStream() + if file is not None: + file.close() + finally: + file = self._source.getByteStream() + if file is not None: + file.close() def _reset_cont_handler(self): self._parser.ProcessingInstructionHandler = \ diff -r c4e6e560e6f5 Lib/xml/sax/saxutils.py --- a/Lib/xml/sax/saxutils.py Mon Feb 04 18:29:47 2013 +0200 +++ b/Lib/xml/sax/saxutils.py Mon Feb 04 21:34:11 2013 +0200 @@ -311,11 +311,14 @@ elif hasattr(source, "read"): f = source source = xmlreader.InputSource() - source.setByteStream(f) + if isinstance(f.read(0), str): + source.setCharacterStream(f) + else: + source.setByteStream(f) if hasattr(f, "name"): source.setSystemId(f.name) - if source.getByteStream() is None: + if source.getCharacterStream() is None and source.getByteStream() is None: sysid = source.getSystemId() basehead = os.path.dirname(os.path.normpath(base)) sysidfilename = os.path.join(basehead, sysid) diff -r c4e6e560e6f5 Lib/xml/sax/xmlreader.py --- a/Lib/xml/sax/xmlreader.py Mon Feb 04 18:29:47 2013 +0200 +++ b/Lib/xml/sax/xmlreader.py Mon Feb 04 21:34:11 2013 +0200 @@ -117,7 +117,9 @@ source = saxutils.prepare_input_source(source) self.prepareParser(source) - file = source.getByteStream() + file = source.getCharacterStream() + if file is None: + file = source.getByteStream() buffer = file.read(self._bufsize) while buffer: self.feed(buffer)