# XXX TypeErrors on calling handlers, or on bad return values from a
# handler, are obscure and unhelpful.

from io import BytesIO
import os
import sys
import sysconfig
import unittest
import traceback

from xml.parsers import expat
from xml.parsers.expat import errors

from test.support import sortdict


class SetAttributeTest(unittest.TestCase):
    def setUp(self):
        self.parser = expat.ParserCreate(namespace_separator='!')

    def test_buffer_text(self):
        self.assertIs(self.parser.buffer_text, False)
        for x in 0, 1, 2, 0:
            self.parser.buffer_text = x
            self.assertIs(self.parser.buffer_text, bool(x))

    def test_namespace_prefixes(self):
        self.assertIs(self.parser.namespace_prefixes, False)
        for x in 0, 1, 2, 0:
            self.parser.namespace_prefixes = x
            self.assertIs(self.parser.namespace_prefixes, bool(x))

    def test_ordered_attributes(self):
        self.assertIs(self.parser.ordered_attributes, False)
        for x in 0, 1, 2, 0:
            self.parser.ordered_attributes = x
            self.assertIs(self.parser.ordered_attributes, bool(x))

    def test_specified_attributes(self):
        self.assertIs(self.parser.specified_attributes, False)
        for x in 0, 1, 2, 0:
            self.parser.specified_attributes = x
            self.assertIs(self.parser.specified_attributes, bool(x))

    def test_invalid_attributes(self):
        with self.assertRaises(AttributeError):
            self.parser.returns_unicode = 1
        with self.assertRaises(AttributeError):
            self.parser.returns_unicode

        # Issue #25019
        self.assertRaises(TypeError, setattr, self.parser, range(0xF), 0)
        self.assertRaises(TypeError, self.parser.__setattr__, range(0xF), 0)
        self.assertRaises(TypeError, getattr, self.parser, range(0xF))


data = b'''\
<?xml version="1.0" encoding="iso-8859-1" standalone="no"?>
<?xml-stylesheet href="stylesheet.css"?>
<!-- comment data -->
<!DOCTYPE quotations SYSTEM "quotations.dtd" [
<!ELEMENT root ANY>
<!ATTLIST root attr1 CDATA #REQUIRED attr2 CDATA #IMPLIED>
<!NOTATION notation SYSTEM "notation.jpeg">
<!ENTITY acirc "&#226;">
<!ENTITY external_entity SYSTEM "entity.file">
<!ENTITY unparsed_entity SYSTEM "entity.file" NDATA notation>
%unparsed_entity;
]>

<root attr1="value1" attr2="value2&#8000;">
<myns:subelement xmlns:myns="http://www.python.org/namespace">
     Contents of subelements
</myns:subelement>
<sub2><![CDATA[contents of CDATA section]]></sub2>
&external_entity;
&skipped_entity;
\xb5
</root>
'''


# Produce UTF-8 output
class ParseTest(unittest.TestCase):
    class Outputter:
        def __init__(self):
            self.out = []

        def StartElementHandler(self, name, attrs):
            self.out.append('Start element: ' + repr(name) + ' ' +
                            sortdict(attrs))

        def EndElementHandler(self, name):
            self.out.append('End element: ' + repr(name))

        def CharacterDataHandler(self, data):
            data = data.strip()
            if data:
                self.out.append('Character data: ' + repr(data))

        def ProcessingInstructionHandler(self, target, data):
            self.out.append('PI: ' + repr(target) + ' ' + repr(data))

        def StartNamespaceDeclHandler(self, prefix, uri):
            self.out.append('NS decl: ' + repr(prefix) + ' ' + repr(uri))

        def EndNamespaceDeclHandler(self, prefix):
            self.out.append('End of NS decl: ' + repr(prefix))

        def StartCdataSectionHandler(self):
            self.out.append('Start of CDATA section')

        def EndCdataSectionHandler(self):
            self.out.append('End of CDATA section')

        def CommentHandler(self, text):
            self.out.append('Comment: ' + repr(text))

        def NotationDeclHandler(self, *args):
            name, base, sysid, pubid = args
            self.out.append('Notation declared: %s' %(args,))

        def UnparsedEntityDeclHandler(self, *args):
            entityName, base, systemId, publicId, notationName = args
            self.out.append('Unparsed entity decl: %s' %(args,))

        def NotStandaloneHandler(self):
            self.out.append('Not standalone')
            return 1

        def ExternalEntityRefHandler(self, *args):
            context, base, sysId, pubId = args
            self.out.append('External entity ref: %s' %(args[1:],))
            return 1

        def StartDoctypeDeclHandler(self, *args):
            self.out.append(('Start doctype', args))
            return 1

        def EndDoctypeDeclHandler(self):
            self.out.append("End doctype")
            return 1

        def EntityDeclHandler(self, *args):
            self.out.append(('Entity declaration', args))
            return 1

        def XmlDeclHandler(self, *args):
            self.out.append(('XML declaration', args))
            return 1

        def ElementDeclHandler(self, *args):
            self.out.append(('Element declaration', args))
            return 1

        def AttlistDeclHandler(self, *args):
            self.out.append(('Attribute list declaration', args))
            return 1

        def SkippedEntityHandler(self, *args):
            self.out.append(("Skipped entity", args))
            return 1

        def DefaultHandler(self, userData):
            pass

        def DefaultHandlerExpand(self, userData):
            pass

    handler_names = [
        'StartElementHandler', 'EndElementHandler', 'CharacterDataHandler',
        'ProcessingInstructionHandler', 'UnparsedEntityDeclHandler',
        'NotationDeclHandler', 'StartNamespaceDeclHandler',
        'EndNamespaceDeclHandler', 'CommentHandler',
        'StartCdataSectionHandler', 'EndCdataSectionHandler', 'DefaultHandler',
        'DefaultHandlerExpand', 'NotStandaloneHandler',
        'ExternalEntityRefHandler', 'StartDoctypeDeclHandler',
        'EndDoctypeDeclHandler', 'EntityDeclHandler', 'XmlDeclHandler',
        'ElementDeclHandler', 'AttlistDeclHandler', 'SkippedEntityHandler',
        ]

    def _hookup_callbacks(self, parser, handler):
        """
        Set each of the callbacks defined on handler and named in
        self.handler_names on the given parser.
        """
        for name in self.handler_names:
            setattr(parser, name, getattr(handler, name))

    def _verify_parse_output(self, operations):
        expected_operations = [
            ('XML declaration', ('1.0', 'iso-8859-1', 0)),
            'PI: \'xml-stylesheet\' \'href="stylesheet.css"\'',
            "Comment: ' comment data '",
            "Not standalone",
            ("Start doctype", ('quotations', 'quotations.dtd', None, 1)),
            ('Element declaration', ('root', (2, 0, None, ()))),
            ('Attribute list declaration', ('root', 'attr1', 'CDATA', None,
                1)),
            ('Attribute list declaration', ('root', 'attr2', 'CDATA', None,
                0)),
            "Notation declared: ('notation', None, 'notation.jpeg', None)",
            ('Entity declaration', ('acirc', 0, '\xe2', None, None, None, None)),
            ('Entity declaration', ('external_entity', 0, None, None,
                'entity.file', None, None)),
            "Unparsed entity decl: ('unparsed_entity', None, 'entity.file', None, 'notation')",
            "Not standalone",
            "End doctype",
            "Start element: 'root' {'attr1': 'value1', 'attr2': 'value2\u1f40'}",
            "NS decl: 'myns' 'http://www.python.org/namespace'",
            "Start element: 'http://www.python.org/namespace!subelement' {}",
            "Character data: 'Contents of subelements'",
            "End element: 'http://www.python.org/namespace!subelement'",
            "End of NS decl: 'myns'",
            "Start element: 'sub2' {}",
            'Start of CDATA section',
            "Character data: 'contents of CDATA section'",
            'End of CDATA section',
            "End element: 'sub2'",
            "External entity ref: (None, 'entity.file', None)",
            ('Skipped entity', ('skipped_entity', 0)),
            "Character data: '\xb5'",
            "End element: 'root'",
        ]
        for operation, expected_operation in zip(operations, expected_operations):
            self.assertEqual(operation, expected_operation)

    def test_parse_bytes(self):
        out = self.Outputter()
        parser = expat.ParserCreate(namespace_separator='!')
        self._hookup_callbacks(parser, out)

        parser.Parse(data, 1)

        operations = out.out
        self._verify_parse_output(operations)
        # Issue #6697.
        self.assertRaises(AttributeError, getattr, parser, '\uD800')

    def test_parse_str(self):
        out = self.Outputter()
        parser = expat.ParserCreate(namespace_separator='!')
        self._hookup_callbacks(parser, out)

        parser.Parse(data.decode('iso-8859-1'), 1)

        operations = out.out
        self._verify_parse_output(operations)

    def test_parse_file(self):
        # Try parsing a file
        out = self.Outputter()
        parser = expat.ParserCreate(namespace_separator='!')
        self._hookup_callbacks(parser, out)
        file = BytesIO(data)

        parser.ParseFile(file)

        operations = out.out
        self._verify_parse_output(operations)

    def test_parse_again(self):
        parser = expat.ParserCreate()
        file = BytesIO(data)
        parser.ParseFile(file)
        # Issue 6676: ensure a meaningful exception is raised when attempting
        # to parse more than one XML document per xmlparser instance,
        # a limitation of the Expat library.
        with self.assertRaises(expat.error) as cm:
            parser.ParseFile(file)
        self.assertEqual(expat.ErrorString(cm.exception.code),
                          expat.errors.XML_ERROR_FINISHED)

class NamespaceSeparatorTest(unittest.TestCase):
    def test_legal(self):
        # Tests that make sure we get errors when the namespace_separator value
        # is illegal, and that we don't for good values:
        expat.ParserCreate()
        expat.ParserCreate(namespace_separator=None)
        expat.ParserCreate(namespace_separator=' ')

    def test_illegal(self):
        try:
            expat.ParserCreate(namespace_separator=42)
            self.fail()
        except TypeError as e:
            self.assertEqual(str(e),
                "ParserCreate() argument 'namespace_separator' must be str or None, not int")

        try:
            expat.ParserCreate(namespace_separator='too long')
            self.fail()
        except ValueError as e:
            self.assertEqual(str(e),
                'namespace_separator must be at most one character, omitted, or None')

    def test_zero_length(self):
        # ParserCreate() needs to accept a namespace_separator of zero length
        # to satisfy the requirements of RDF applications that are required
        # to simply glue together the namespace URI and the localname.  Though
        # considered a wart of the RDF specifications, it needs to be supported.
        #
        # See XML-SIG mailing list thread starting with
        # http://mail.python.org/pipermail/xml-sig/2001-April/005202.html
        #
        expat.ParserCreate(namespace_separator='') # too short


class InterningTest(unittest.TestCase):
    def test(self):
        # Test the interning machinery.
        p = expat.ParserCreate()
        L = []
        def collector(name, *args):
            L.append(name)
        p.StartElementHandler = collector
        p.EndElementHandler = collector
        p.Parse(b"<e> <e/> <e></e> </e>", 1)
        tag = L[0]
        self.assertEqual(len(L), 6)
        for entry in L:
            # L should have the same string repeated over and over.
            self.assertTrue(tag is entry)

    def test_issue9402(self):
        # create an ExternalEntityParserCreate with buffer text
        class ExternalOutputter:
            def __init__(self, parser):
                self.parser = parser
                self.parser_result = None

            def ExternalEntityRefHandler(self, context, base, sysId, pubId):
                external_parser = self.parser.ExternalEntityParserCreate("")
                self.parser_result = external_parser.Parse(b"", 1)
                return 1

        parser = expat.ParserCreate(namespace_separator='!')
        parser.buffer_text = 1
        out = ExternalOutputter(parser)
        parser.ExternalEntityRefHandler = out.ExternalEntityRefHandler
        parser.Parse(data, 1)
        self.assertEqual(out.parser_result, 1)


class BufferTextTest(unittest.TestCase):