def comment(include_parens=False):
inner = recursive() > named(u'comment', RFC(7230))
inner.rec = '(' + string(ctext | quoted_pair(sensible_for=u'()\\') |
inner) + ')'
if not include_parens:
inner = (lambda s: s[1:-1]) << inner
return inner > named(u'comment', RFC(7230))
def test_parser_edge_cases():
# Our parser implementation is general enough that
# some of its branches are not being exercised by our regular tests,
# so I had to come up with these contrived examples to test them.
p = many(rfc7230.tchar) > named(u'p')
p1 = '1' * p > named(u'p1')
p2 = '11' * p * skip('\n') > named(u'p2')
assert parse(p1 | p2, b'11abc') == (u'1', [u'1', u'a', u'b', u'c'])
assert parse(p1 | p2, b'11abc\n') == (u'11', [u'a', u'b', u'c'])
p = recursive() > named(u'p')
p.rec = (rfc7230.tchar * p | subst(None) << empty)
assert parse(p, b'abc') == (u'a', (u'b', (u'c', None)))
p = literal('ab') > named(u'p')
p0 = subst(u'') << empty | p > named(u'p0')
p1 = 'xab' * p0 > named(u'p1')
p2 = 'x' * string(p0) * '!' > named(u'p2')
assert parse(p1 | p2, b'xabab') == (u'xab', u'ab')
assert parse(p1 | p2, b'xabab!') == (u'x', u'abab', u'!')
p = empty | literal('a') > named(u'p')
p0 = p * 'x' > named(u'x')
assert parse(p0, b'x') == u'x'
def test_parser_edge_cases():
# Our parser implementation is general enough that
# some of its branches are not being exercised by our regular tests,
# so I had to come up with these contrived examples to test them.
p = many(rfc7230.tchar) > named(u'p')
p1 = '1' * p > named(u'p1')
p2 = '11' * p * skip('\n') > named(u'p2')
assert parse(p1 | p2, b'11abc') == (u'1', [u'1', u'a', u'b', u'c'])
assert parse(p1 | p2, b'11abc\n') == (u'11', [u'a', u'b', u'c'])
p = recursive() > named(u'p')
p.rec = (rfc7230.tchar * p | subst(None) << empty)
assert parse(p, b'abc') == (u'a', (u'b', (u'c', None)))
p = literal('ab') > named(u'p')
p0 = subst(u'') << empty | p > named(u'p0')
p1 = 'xab' * p0 > named(u'p1')
p2 = 'x' * string(p0) * '!' > named(u'p2')
assert parse(p1 | p2, b'xabab') == (u'xab', u'ab')
assert parse(p1 | p2, b'xabab!') == (u'x', u'abab', u'!')
p = empty | literal('a') > named(u'p')
p0 = p * 'x' > named(u'x')
assert parse(p0, b'x') == u'x'
def comment(include_parens=False):
inner = recursive() > named(u'comment', RFC(7230))
inner.rec = '(' + string(ctext | quoted_pair(sensible_for=u'()\\')
| inner) + ')'
if not include_parens:
inner = (lambda s: s[1:-1]) << inner
return inner > named(u'comment', RFC(7230))
def preference_parameter(head=False):
# The head (first) ``preference-parameter`` of a ``preference``
# contains the actual preference name, which we want to annotate.
name_cls = Preference if head else CaseInsensitive
return (_normalize_empty_value <<
(parameter(name_cls=name_cls) | name_cls << token)) > named(
u'preference-parameter', RFC(7240, errata=4439), is_pivot=True)
def preference_parameter(head=False):
# The head (first) ``preference-parameter`` of a ``preference``
# contains the actual preference name, which we want to annotate.
name_cls = Preference if head else CaseInsensitive
return (
_normalize_empty_value << (parameter(name_cls=name_cls) |
name_cls << token)
) > named(u'preference-parameter', RFC(7240, errata=4439), is_pivot=True)
def transfer_coding(no_trailers=False, no_q=False):
exclude = _built_in_codings
if no_trailers:
exclude = exclude + ['trailers']
r = transfer_extension(exclude, no_q)
for name in _built_in_codings:
r = r | _empty_params << (TransferCoding << literal(name))
return r > named(u'transfer-coding', RFC(7230), is_pivot=True)
def transfer_coding(no_trailers=False, no_q=False):
exclude = _built_in_codings
if no_trailers:
exclude = exclude + ['trailers']
r = transfer_extension(exclude, no_q)
for name in _built_in_codings:
r = r | _empty_params << (TransferCoding << literal(name))
return r > named(u'transfer-coding', RFC(7230), is_pivot=True)
def quoted_pair(sensible_for):
# In RFC 7230, ``<quoted-pair>`` is a single rule,
# but we parametrize it to report no. 1017 depending on the context.
@can_complain
def check_sensible(complain, c):
if c not in sensible_for:
complain(1017, char=c)
return c
return (check_sensible << skip('\\') * (HTAB | SP | VCHAR | obs_text)
> named(u'quoted-pair', RFC(7230)))
def quoted_pair(sensible_for):
# In RFC 7230, ``<quoted-pair>`` is a single rule,
# but we parametrize it to report no. 1017 depending on the context.
@can_complain
def check_sensible(complain, c):
if c not in sensible_for:
complain(1017, char=c)
return c
return (check_sensible << skip('\\') *
(HTAB | SP | VCHAR | obs_text) > named(u'quoted-pair', RFC(7230)))
def media_range(no_q=False):
return Parametrized << (
(
literal('*/*') |
type_ + '/' + '*' |
MediaType << type_ + '/' + subtype
) *
(
MultiDict << many(
skip(OWS * ';' * OWS) *
parameter(exclude=['q'] if no_q else [])
)
)
) > named(u'media-range', RFC(7231), is_pivot=True)
def link_extension(exclude_builtin):
if exclude_builtin:
exclude1 = [name for name in _builtin_params if not name.endswith('*')]
exclude2 = [name.rstrip('*')
for name in _builtin_params if name.endswith('*')]
else: # pragma: no cover
exclude1 = exclude2 = None
return (
(
(CaseInsensitive << parmname__excluding(exclude1)) *
maybe(skip(OWS * '=' * OWS) * (ptoken | quoted_string))
) |
(
(CaseInsensitive << ext_name_star__excluding(exclude2)) *
skip(OWS * '=' * OWS) * ext_value
)
) > named(u'link-extension', RFC(5988), is_pivot=True)
def link_extension(exclude_builtin):
if exclude_builtin:
exclude1 = [name for name in _builtin_params if not name.endswith('*')]
exclude2 = [name.rstrip('*')
for name in _builtin_params if name.endswith('*')]
else:
exclude1 = exclude2 = None
return (
(
(CaseInsensitive << parmname__excluding(exclude1)) *
maybe(skip(OWS * '=' * OWS) * (ptoken | quoted_string))
) |
(
(CaseInsensitive << ext_name_star__excluding(exclude2)) *
skip(OWS * '=' * OWS) * ext_value
)
) > named(u'link-extension', RFC(5988), is_pivot=True)
# RFC 7240 Section 2: "Empty or zero-length values on both
# the preference token and within parameters are equivalent
# to no value being specified at all."
(name, value) = x if isinstance(x, tuple) else (x, None)
return Parametrized(name, None if value == u'' else value)
def preference_parameter(head=False):
# The head (first) ``preference-parameter`` of a ``preference``
# contains the actual preference name, which we want to annotate.
name_cls = Preference if head else CaseInsensitive
return (
_normalize_empty_value << (parameter(name_cls=name_cls) |
name_cls << token)
) > named(u'preference-parameter', RFC(7240, errata=4439), is_pivot=True)
preference = Parametrized << (
preference_parameter(head=True) *
many(skip(OWS * ';') * maybe(skip(OWS) * preference_parameter()))
) > named(u'preference', RFC(7240, errata=4439), is_pivot=True)
Prefer = comma_list1(preference) > pivot
Preference_Applied = comma_list1(preference_parameter(head=True)) > pivot
return_ = CaseInsensitive << (literal('representation') | 'minimal') > pivot
wait = delay_seconds > auto
handling = CaseInsensitive << (literal('strict') | 'lenient') > pivot
fill_names(globals(), RFC(7240))
def transfer_extension(exclude=None, no_q=False):
return Parametrized << (
(TransferCoding << token__excluding(exclude or [])) *
(MultiDict << many(skip(OWS * ';' * OWS) * transfer_parameter(no_q)))
) > named(u'transfer-extension', RFC(7230), is_pivot=True)
def transfer_parameter(no_q=False):
return (
(token__excluding(['q']) if no_q else token) *
skip(BWS * '=' * BWS) * (token | quoted_string)
) > named(u'transfer-parameter', RFC(7230), is_pivot=True)
def ext_name_star__excluding(exclude):
return (parmname__excluding(exclude) + '*'
> named(u'ext-name-star', RFC(5988)))
def comma_list1(element):
return _collect_elements << (
many(subst(None) << ',' * OWS) +
((lambda x: [x]) << group(element)) +
many(skip(OWS * ',') * maybe(skip(OWS) * element))
) > named(u'1#rule', RFC(7230, section=(7,)))
def comma_list(element):
# RFC Errata ID: 5257
return _collect_elements << (maybe(group(element) * skip(OWS)) % many(
skip(literal(',') * OWS) * maybe(group(element) * skip(OWS)))) > named(
u'#rule', RFC(7230, section=u'7'))
def comma_list(element):
# RFC Errata ID: 5257
return _collect_elements << (
maybe(group(element) * skip(OWS)) %
many(skip(literal(',') * OWS) * maybe(group(element) * skip(OWS)))
) > named(u'#rule', RFC(7230, section=u'7'))
def parmname__excluding(exclude):
return (string_excluding(attr_char, [''] + exclude)
> named(u'parmname', RFC(5987), is_pivot=True))
def parmname__excluding(exclude):
return (string_excluding(attr_char, [''] + exclude) > named(
u'parmname', RFC(5987), is_pivot=True))
def transfer_extension(exclude=None, no_q=False):
return Parametrized << (
(TransferCoding << token__excluding(exclude or [])) *
(MultiDict << many(skip(OWS * ';' * OWS) * transfer_parameter(no_q)))
) > named(u'transfer-extension', RFC(7230), is_pivot=True)
def transfer_parameter(no_q=False):
return ((token__excluding(['q']) if no_q else token) *
skip(BWS * '=' * BWS) * (token | quoted_string)) > named(
u'transfer-parameter', RFC(7230), is_pivot=True)
def ext_name_star__excluding(exclude):
return (parmname__excluding(exclude) + '*'
> named(u'ext-name-star', RFC(5988)))
def comma_list1(element):
return _collect_elements << (many(subst(None) << ',' * OWS) + (
(lambda x: [x]) << group(element)) + many(
skip(OWS * ',') * maybe(skip(OWS) * element))) > named(
u'1#rule', RFC(7230, section=u'7'))
def comma_list(element):
return _collect_elements << maybe(
(subst([None, None]) << literal(',') |
(lambda x: [x]) << group(element)) +
many(skip(OWS * ',') * maybe(skip(OWS) * element))
) > named(u'#rule', RFC(7230, section=(7,)))
def media_range(no_q=False):
return Parametrized << (
(literal('*/*') | type_ + '/' + '*' | _check_media_type <<
(MediaType << type_ + '/' + subtype)) * (MultiDict << many(
skip(OWS * ';' * OWS) * parameter(exclude=['q'] if no_q else [])))
) > named(u'media-range', RFC(7231), is_pivot=True)
def comma_list(element):
return _collect_elements << maybe(
(subst([None, None]) << literal(',') |
(lambda x: [x]) << group(element)) +
many(skip(OWS * ',') * maybe(skip(OWS) * element))) > named(
u'#rule', RFC(7230, section=(7, )))
def parameter(exclude=None):
return ((CaseInsensitive << token__excluding(exclude or [])) * skip('=') *
(token | quoted_string)) > named(
u'parameter', RFC(7231), is_pivot=True)
def parameter(exclude=None):
return (
(CaseInsensitive << token__excluding(exclude or [])) *
skip('=') * (token | quoted_string)
) > named(u'parameter', RFC(7231), is_pivot=True)
def extension_pragma(exclude_no_cache=False):
return Parametrized << (
(token__excluding(['no-cache']) if exclude_no_cache else token) *
maybe(skip('=') * (token | quoted_string))) > named(
u'extension-pragma', RFC(7234), is_pivot=True)
def extension_pragma(exclude_no_cache=False):
return Parametrized << (
(token__excluding(['no-cache']) if exclude_no_cache else token) *
maybe(skip('=') * (token | quoted_string))
) > named(u'extension-pragma', RFC(7234), is_pivot=True)
def _normalize_empty_value(x):
# RFC 7240 Section 2: "Empty or zero-length values on both
# the preference token and within parameters are equivalent
# to no value being specified at all."
(name, value) = x if isinstance(x, tuple) else (x, None)
return Parametrized(name, None if value == u'' else value)
def preference_parameter(head=False):
# The head (first) ``preference-parameter`` of a ``preference``
# contains the actual preference name, which we want to annotate.
name_cls = Preference if head else CaseInsensitive
return (_normalize_empty_value <<
(parameter(name_cls=name_cls) | name_cls << token)) > named(
u'preference-parameter', RFC(7240, errata=4439), is_pivot=True)
preference = Parametrized << (preference_parameter(head=True) * many(
skip(OWS * ';') * maybe(skip(OWS) * preference_parameter()))) > named(
u'preference', RFC(7240, errata=4439), is_pivot=True)
Prefer = comma_list1(preference) > pivot
Preference_Applied = comma_list1(preference_parameter(head=True)) > pivot
return_ = CaseInsensitive << (literal('representation') | 'minimal') > pivot
wait = delay_seconds > auto
handling = CaseInsensitive << (literal('strict') | 'lenient') > pivot
fill_names(globals(), RFC(7240))
