def test_escape(self):
data = "\t\n\r !\"#$%&'()*+,-./0123456789:;<=>?@"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~"
escaped_normal = "%09%0A%0D%20!%22%23%24%25%26'()*%2B%2C-.%2F"\
"0123456789%3A%3B%3C%3D%3E%3F%40"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ%5B%5C%5D%5E_%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D~"
escaped_max = "%09%0A%0D%20%21%22%23%24%25%26%27%28%29%2A%2B%2C%2D"\
"%2E%2F0123456789%3A%3B%3C%3D%3E%3F%40"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ%5B%5C%5D%5E%5F%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D%7E"
escaped_min = "%09%0A%0D%20!%22%23$%25&'()*+,-./0123456789:;%3C=%3E?@"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ[%5C]%5E_%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D~"
escaped_min_2396 = "%09%0A%0D%20!%22%23$%25&'()*+,-./0123456789:;%3C"\
"=%3E?@ABCDEFGHIJKLMNOPQRSTUVWXYZ%5B%5C%5D%5E_%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D~"
self.compare_strings(escaped_normal, uri.escape_data(data),
"Normal escaping")
self.compare_strings(
escaped_max, uri.escape_data(data,
lambda x: not uri.is_alphanum(x)),
"Max escaping")
self.compare_strings(escaped_min,
uri.escape_data(data,
lambda x: False), "Min escaping")
self.compare_strings(
escaped_min_2396,
uri.escape_data(data,
lambda x: False,
allowed_test=uri.is_allowed_2396), "Min escaping")
self.compare_strings(
data,
uri.unescape_data(uri.escape_data(data)).decode('ascii'),
"Round-trip escaping")
try:
uri.escape_data(ul('Caf\xe9'))
self.fail("non-ASCII character for escaping")
except UnicodeEncodeError:
pass
# byte strings can contain any data, no UTF-8 encoding is done
self.assertTrue(uri.escape_data(b'Caf\xe9') == "Caf%E9")
def test_escape(self):
data = "\t\n\r !\"#$%&'()*+,-./0123456789:;<=>?@"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~"
escaped_normal = "%09%0A%0D%20!%22%23%24%25%26'()*%2B%2C-.%2F"\
"0123456789%3A%3B%3C%3D%3E%3F%40"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ%5B%5C%5D%5E_%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D~"
escaped_max = "%09%0A%0D%20%21%22%23%24%25%26%27%28%29%2A%2B%2C%2D"\
"%2E%2F0123456789%3A%3B%3C%3D%3E%3F%40"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ%5B%5C%5D%5E%5F%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D%7E"
escaped_min = "%09%0A%0D%20!%22%23$%25&'()*+,-./0123456789:;%3C=%3E?@"\
"ABCDEFGHIJKLMNOPQRSTUVWXYZ[%5C]%5E_%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D~"
escaped_min_2396 = "%09%0A%0D%20!%22%23$%25&'()*+,-./0123456789:;%3C"\
"=%3E?@ABCDEFGHIJKLMNOPQRSTUVWXYZ%5B%5C%5D%5E_%60"\
"abcdefghijklmnopqrstuvwxyz%7B%7C%7D~"
self.compare_strings(
escaped_normal, uri.escape_data(data), "Normal escaping")
self.compare_strings(escaped_max, uri.escape_data(data,
lambda x: not uri.is_alphanum(x)),
"Max escaping")
self.compare_strings(
escaped_min, uri.escape_data(data, lambda x: False),
"Min escaping")
self.compare_strings(
escaped_min_2396, uri.escape_data(
data, lambda x: False, allowed_test=uri.is_allowed_2396),
"Min escaping")
self.compare_strings(
data, uri.unescape_data(uri.escape_data(data)).decode('ascii'),
"Round-trip escaping")
try:
uri.escape_data(ul('Caf\xe9'))
self.fail("non-ASCII character for escaping")
except UnicodeEncodeError:
pass
# byte strings can contain any data, no UTF-8 encoding is done
self.assertTrue(uri.escape_data(b'Caf\xe9') == "Caf%E9")
def test_basics(self):
"""Tests for basic character classes.
alpha = lowalpha | upalpha
lowalpha = "a" | ... | "z"
upalpha = "A" | ... | "Z"
digit = "0" | ... | "9"
alphanum = alpha | digit
reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" | "$" | ","
unreserved = alphanum | mark
mark = "-" | "_" | "." | "!" | "~" | "*" | "'" | "(" | ")"
"""
# UPALPHA = <any US-ASCII uppercase letter "A".."Z">
upalpha = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_upalpha(c) == (c in upalpha),
"is_upalpha(chr(%i))" % i)
lowalpha = "abcdefghijklmnopqrstuvwxyz"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_lowalpha(c) == (c in lowalpha),
"is_lowalpha(chr(%i))" % i)
alpha = upalpha + lowalpha
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_alpha(c) == (c in alpha),
"is_alpha(chr(%i))" % i)
digit = "0123456789"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_digit(c) == (c in digit),
"is_digit(chr(%i))" % i)
alphanum = alpha + digit
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_alphanum(c) == (c in alphanum),
"is_alphanum(chr(%i))" % i)
reserved = ";/?:@&=+$,"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_reserved_2396(c) == (c in reserved),
"is_reserved_2396(chr(%i))" % i)
reserved = ";/?:@&=+$,[]"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_reserved(c) == (c in reserved),
"is_reserved(chr(%i))" % i)
mark = "-_.!~*'()"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_mark(c) == (c in mark),
"is_mark(chr(%i))" % i)
unreserved = alphanum + mark
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_unreserved(c) == (c in unreserved),
"is_unreserved(chr(%i))" % i)
control = "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B\x0C\x0D"\
"\x0E\x0F\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C"\
"\x1D\x1E\x1F\x7F"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_control(c) == (c in control), "is_control(chr(%i))" % i)
space = " "
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_space(c) == (c in space),
"is_space(chr(%i))" % i)
delims = "<>#%\""
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_delims(c) == (c in delims), "is_delims(chr(%i))" % i)
unwise_2396 = "{}|\\^[]`"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_unwise_2396(c) == (c in unwise_2396),
"is_unwise_2396(chr(%i))" % i)
unwise = "{}|\\^`"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_unwise(c) == (c in unwise), "is_unwise(chr(%i))" % i)
authority_reserved = ";:@?/"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_authority_reserved(c) == (
c in authority_reserved), "is_authority_reserved(chr(%i))" % i)
path_segment_reserved = "/;=?"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_path_segment_reserved(c) ==
(c in path_segment_reserved),
"is_path_segment_reserved(chr(%i))" % i)
query_reserved = ";/?:@&=+,$"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_query_reserved(c) == (c in query_reserved),
"is_query_reserved(chr(%i))" % i)
def test_basics(self):
"""Tests for basic character classes.
alpha = lowalpha | upalpha
lowalpha = "a" | ... | "z"
upalpha = "A" | ... | "Z"
digit = "0" | ... | "9"
alphanum = alpha | digit
reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" | "$" | ","
unreserved = alphanum | mark
mark = "-" | "_" | "." | "!" | "~" | "*" | "'" | "(" | ")"
"""
# UPALPHA = <any US-ASCII uppercase letter "A".."Z">
upalpha = "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_upalpha(c) == (c in upalpha),
"is_upalpha(chr(%i))" % i)
lowalpha = "abcdefghijklmnopqrstuvwxyz"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_lowalpha(c) == (c in lowalpha),
"is_lowalpha(chr(%i))" % i)
alpha = upalpha + lowalpha
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_alpha(c) == (c in alpha),
"is_alpha(chr(%i))" % i)
digit = "0123456789"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_digit(c) == (c in digit),
"is_digit(chr(%i))" % i)
alphanum = alpha + digit
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_alphanum(c) == (c in alphanum),
"is_alphanum(chr(%i))" % i)
reserved = ";/?:@&=+$,"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_reserved_2396(c) == (c in reserved),
"is_reserved_2396(chr(%i))" % i)
reserved = ";/?:@&=+$,[]"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_reserved(c) == (c in reserved),
"is_reserved(chr(%i))" % i)
mark = "-_.!~*'()"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_mark(c) == (c in mark),
"is_mark(chr(%i))" % i)
unreserved = alphanum + mark
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_unreserved(c) == (c in unreserved),
"is_unreserved(chr(%i))" % i)
control = "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B\x0C\x0D"\
"\x0E\x0F\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C"\
"\x1D\x1E\x1F\x7F"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_control(c) == (c in control), "is_control(chr(%i))" % i)
space = " "
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_space(c) == (c in space),
"is_space(chr(%i))" % i)
delims = "<>#%\""
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_delims(c) == (c in delims), "is_delims(chr(%i))" % i)
unwise_2396 = "{}|\\^[]`"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_unwise_2396(c) == (c in unwise_2396),
"is_unwise_2396(chr(%i))" % i)
unwise = "{}|\\^`"
for i in range3(0, 256):
c = character(i)
self.assertTrue(
uri.is_unwise(c) == (c in unwise), "is_unwise(chr(%i))" % i)
authority_reserved = ";:@?/"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_authority_reserved(c) == (
c in authority_reserved), "is_authority_reserved(chr(%i))" % i)
path_segment_reserved = "/;=?"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_path_segment_reserved(c) ==
(c in path_segment_reserved),
"is_path_segment_reserved(chr(%i))" % i)
query_reserved = ";/?:@&=+,$"
for i in range3(0, 256):
c = character(i)
self.assertTrue(uri.is_query_reserved(c) == (c in query_reserved),
"is_query_reserved(chr(%i))" % i)