Python htmlentitydefs.codepoint2name() Examples

The following are 14 code examples of htmlentitydefs.codepoint2name(). You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may also want to check out all available functions/classes of the module htmlentitydefs , or try the search function .
Example #1
Source File: test_codeccallbacks.py    From ironpython2 with Apache License 2.0 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #2
Source File: test_codeccallbacks.py    From BinderFilter with MIT License 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #3
Source File: test_codeccallbacks.py    From oss-ftp with MIT License 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #4
Source File: test_codeccallbacks.py    From gcblue with BSD 3-Clause "New" or "Revised" License 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #5
Source File: test_codeccallbacks.py    From medicare-demo with Apache License 2.0 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #6
Source File: test_codeccallbacks.py    From CTFCrackTools-V2 with GNU General Public License v3.0 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #7
Source File: test_codeccallbacks.py    From CTFCrackTools with GNU General Public License v3.0 6 votes vote down vote up 
def test_xmlcharnamereplace(self):
        # This time use a named character entity for unencodable
        # characters, if one is available.

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                try:
                    l.append(u"&%s;" % htmlentitydefs.codepoint2name[ord(c)])
                except KeyError:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error(
            "test.xmlcharnamereplace", xmlcharnamereplace)

        sin = u"\xab\u211c\xbb = \u2329\u1234\u20ac\u232a"
        sout = "«ℜ» = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("ascii", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ€⟩"
        self.assertEqual(sin.encode("latin-1", "test.xmlcharnamereplace"), sout)
        sout = "\xabℜ\xbb = ⟨ሴ\xa4⟩"
        self.assertEqual(sin.encode("iso-8859-15", "test.xmlcharnamereplace"), sout)
Example #8
Source File: multibytecodec_support.py    From ironpython2 with Apache License 2.0 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            self.skipTest('encoding contains full ISO 10646 map')

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)
Example #9
Source File: test_multibytecodec_support.py    From BinderFilter with MIT License 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            return

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)
Example #10
Source File: test_multibytecodec_support.py    From oss-ftp with MIT License 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            self.skipTest('encoding contains full ISO 10646 map')

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)
Example #11
Source File: test_multibytecodec_support.py    From gcblue with BSD 3-Clause "New" or "Revised" License 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            self.skipTest('encoding contains full ISO 10646 map')

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)
Example #12
Source File: test_multibytecodec_support.py    From medicare-demo with Apache License 2.0 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            return

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)
Example #13
Source File: test_multibytecodec_support.py    From CTFCrackTools-V2 with GNU General Public License v3.0 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            return

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)
Example #14
Source File: test_multibytecodec_support.py    From CTFCrackTools with GNU General Public License v3.0 5 votes vote down vote up 
def test_customreplace_encode(self):
        if self.has_iso10646:
            return

        from htmlentitydefs import codepoint2name

        def xmlcharnamereplace(exc):
            if not isinstance(exc, UnicodeEncodeError):
                raise TypeError("don't know how to handle %r" % exc)
            l = []
            for c in exc.object[exc.start:exc.end]:
                if ord(c) in codepoint2name:
                    l.append(u"&%s;" % codepoint2name[ord(c)])
                else:
                    l.append(u"&#%d;" % ord(c))
            return (u"".join(l), exc.end)

        codecs.register_error("test.xmlcharnamereplace", xmlcharnamereplace)

        if self.xmlcharnametest:
            sin, sout = self.xmlcharnametest
        else:
            sin = u"\xab\u211c\xbb = \u2329\u1234\u232a"
            sout = "«ℜ» = ⟨ሴ⟩"
        self.assertEqual(self.encode(sin,
                                    "test.xmlcharnamereplace")[0], sout)