Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/ftfy/bad_codecs/sloppy.py: 92%
39 statements
« prev ^ index » next coverage.py v7.3.2, created at 2023-12-08 06:33 +0000
« prev ^ index » next coverage.py v7.3.2, created at 2023-12-08 06:33 +0000
1r"""
2`ftfy.bad_codecs.sloppy` provides character-map encodings that fill their "holes"
3in a messy but common way: by outputting the Unicode codepoints with the same
4numbers.
6This is incredibly ugly, and it's also in the HTML5 standard.
8A single-byte encoding maps each byte to a Unicode character, except that some
9bytes are left unmapped. In the commonly-used Windows-1252 encoding, for
10example, bytes 0x81 and 0x8D, among others, have no meaning.
12Python, wanting to preserve some sense of decorum, will handle these bytes
13as errors. But Windows knows that 0x81 and 0x8D are possible bytes and they're
14different from each other. It just hasn't defined what they are in terms of
15Unicode.
17Software that has to interoperate with Windows-1252 and Unicode -- such as all
18the common Web browsers -- will pick some Unicode characters for them to map
19to, and the characters they pick are the Unicode characters with the same
20numbers: U+0081 and U+008D. This is the same as what Latin-1 does, and the
21resulting characters tend to fall into a range of Unicode that's set aside for
22obsolete Latin-1 control characters anyway.
24These sloppy codecs let Python do the same thing, thus interoperating with
25other software that works this way. It defines a sloppy version of many
26single-byte encodings with holes. (There is no need for a sloppy version of
27an encoding without holes: for example, there is no such thing as
28sloppy-iso-8859-2 or sloppy-macroman.)
30The following encodings will become defined:
32- sloppy-windows-1250 (Central European, sort of based on ISO-8859-2)
33- sloppy-windows-1251 (Cyrillic)
34- sloppy-windows-1252 (Western European, based on Latin-1)
35- sloppy-windows-1253 (Greek, sort of based on ISO-8859-7)
36- sloppy-windows-1254 (Turkish, based on ISO-8859-9)
37- sloppy-windows-1255 (Hebrew, based on ISO-8859-8)
38- sloppy-windows-1256 (Arabic)
39- sloppy-windows-1257 (Baltic, based on ISO-8859-13)
40- sloppy-windows-1258 (Vietnamese)
41- sloppy-cp874 (Thai, based on ISO-8859-11)
42- sloppy-iso-8859-3 (Maltese and Esperanto, I guess)
43- sloppy-iso-8859-6 (different Arabic)
44- sloppy-iso-8859-7 (Greek)
45- sloppy-iso-8859-8 (Hebrew)
46- sloppy-iso-8859-11 (Thai)
48Aliases such as "sloppy-cp1252" for "sloppy-windows-1252" will also be
49defined.
51Five of these encodings (`sloppy-windows-1250` through `sloppy-windows-1254`)
52are used within ftfy.
54Here are some examples, using :func:`ftfy.explain_unicode` to illustrate how
55sloppy-windows-1252 merges Windows-1252 with Latin-1:
57 >>> from ftfy import explain_unicode
58 >>> some_bytes = b'\x80\x81\x82'
59 >>> explain_unicode(some_bytes.decode('latin-1'))
60 U+0080 \x80 [Cc] <unknown>
61 U+0081 \x81 [Cc] <unknown>
62 U+0082 \x82 [Cc] <unknown>
64 >>> explain_unicode(some_bytes.decode('windows-1252', 'replace'))
65 U+20AC € [Sc] EURO SIGN
66 U+FFFD � [So] REPLACEMENT CHARACTER
67 U+201A ‚ [Ps] SINGLE LOW-9 QUOTATION MARK
69 >>> explain_unicode(some_bytes.decode('sloppy-windows-1252'))
70 U+20AC € [Sc] EURO SIGN
71 U+0081 \x81 [Cc] <unknown>
72 U+201A ‚ [Ps] SINGLE LOW-9 QUOTATION MARK
73"""
74from __future__ import annotations
75import codecs
76from encodings import normalize_encoding
77import sys
79REPLACEMENT_CHAR = "\ufffd"
80PY26 = sys.version_info[:2] == (2, 6)
83def make_sloppy_codec(encoding):
84 """
85 Take a codec name, and return a 'sloppy' version of that codec that can
86 encode and decode the unassigned bytes in that encoding.
88 Single-byte encodings in the standard library are defined using some
89 boilerplate classes surrounding the functions that do the actual work,
90 `codecs.charmap_decode` and `charmap_encode`. This function, given an
91 encoding name, *defines* those boilerplate classes.
92 """
93 # Make a bytestring of all 256 possible bytes.
94 all_bytes = bytes(range(256))
96 # Get a list of what they would decode to in Latin-1.
97 sloppy_chars = list(all_bytes.decode("latin-1"))
99 # Get a list of what they decode to in the given encoding. Use the
100 # replacement character for unassigned bytes.
101 if PY26:
102 decoded_chars = all_bytes.decode(encoding, "replace")
103 else:
104 decoded_chars = all_bytes.decode(encoding, errors="replace")
106 # Update the sloppy_chars list. Each byte that was successfully decoded
107 # gets its decoded value in the list. The unassigned bytes are left as
108 # they are, which gives their decoding in Latin-1.
109 for i, char in enumerate(decoded_chars):
110 if char != REPLACEMENT_CHAR:
111 sloppy_chars[i] = char
113 # For ftfy's own purposes, we're going to allow byte 1A, the "Substitute"
114 # control code, to encode the Unicode replacement character U+FFFD.
115 sloppy_chars[0x1A] = REPLACEMENT_CHAR
117 # Create the data structures that tell the charmap methods how to encode
118 # and decode in this sloppy encoding.
119 decoding_table = "".join(sloppy_chars)
120 encoding_table = codecs.charmap_build(decoding_table)
122 # Now produce all the class boilerplate. Look at the Python source for
123 # `encodings.cp1252` for comparison; this is almost exactly the same,
124 # except I made it follow pep8.
125 class Codec(codecs.Codec):
126 def encode(self, input, errors="strict"):
127 return codecs.charmap_encode(input, errors, encoding_table)
129 def decode(self, input, errors="strict"):
130 return codecs.charmap_decode(input, errors, decoding_table) # type: ignore
132 class IncrementalEncoder(codecs.IncrementalEncoder):
133 def encode(self, input, final=False):
134 return codecs.charmap_encode(input, self.errors, encoding_table)[0]
136 class IncrementalDecoder(codecs.IncrementalDecoder):
137 def decode(self, input, final=False):
138 return codecs.charmap_decode(input, self.errors, decoding_table)[0] # type: ignore
140 class StreamWriter(Codec, codecs.StreamWriter):
141 pass
143 class StreamReader(Codec, codecs.StreamReader):
144 pass
146 return codecs.CodecInfo(
147 name="sloppy-" + encoding,
148 encode=Codec().encode,
149 decode=Codec().decode,
150 incrementalencoder=IncrementalEncoder,
151 incrementaldecoder=IncrementalDecoder,
152 streamreader=StreamReader,
153 streamwriter=StreamWriter,
154 )
157# Define a codec for each incomplete encoding. The resulting CODECS dictionary
158# can be used by the main module of ftfy.bad_codecs.
159CODECS = {}
160INCOMPLETE_ENCODINGS = (
161 ["windows-%s" % num for num in range(1250, 1259)]
162 + ["iso-8859-%s" % num for num in (3, 6, 7, 8, 11)]
163 + ["cp%s" % num for num in range(1250, 1259)]
164 + ["cp874"]
165)
167for _encoding in INCOMPLETE_ENCODINGS:
168 _new_name = normalize_encoding("sloppy-" + _encoding)
169 CODECS[_new_name] = make_sloppy_codec(_encoding)