Move the trustme import from global scope into the nosan_server fixture
in order to make it possible to deselect the test via command-line
when trustme is not available.
Fix the listen() invocation for the test server not to pass a backlog
value of zero. The value of zero means no backlog which effectively
means that the socket can not accept any connections. This does not
matter for the majority of platforms since the value is only advisory
and the platform tends to go with a bigger backlog anyway. However,
a few platforms (e.g. alpha or riscv Linux) do take the value literally,
and therefore the tests fail since they are unable to connect to
the server.
pytest-httpbin<1.0 ships with a server certificate with a commonName but
no subjectAltName. urllib3 2.0 will stop supporting those in the future,
so we want to upgrade pytest-httpbin.
Unfortunately, `test_https_warnings` was relying on this broken
certificate. With this change, we use `trustme` to create a broken
certificate specifically for this test, so that we can upgrade
pytest-httpbin and make sure that other tests relying on httpbin TLS
support will continue to work with urllib3 2.0.
* Add support for Brotli decoding
When the brotli or brotlicffi packages are installed,
urllib3.util.make_headers() inserts ',br' in the Accept-Encoding header
and decodes br from the answers.
* Create the default Accept-Encoding header once
* Preserve the previous delimiter behavior
* Update prose in quickstart.rst
Co-authored-by: Seth Michael Larson <sethmichaellarson@gmail.com>
Extract also creates the folder hierarchy, however we do not need that,
the file itself being extracted to a temporary folder is good enough.
Instead we read the content of the zip and then write it. The write is
not locked but it's OK to update the same file multiple times given the
update operation will not alter the content of the file. By not creating
the folder hierarchy (default via extract) we no longer can run into the
problem of two parallel extracts both trying to create the folder
hierarchy without exists ok flag, and one must fail.
Resolves#5223.
Signed-off-by: Bernát Gábor <bgabor8@bloomberg.net>
Although using the (non-vendored) chardet library is fine for requests
itself, but using a LGPL dependency the story is a lot less clear
for downstream projects, particularly ones that might like to bundle
requests (and thus chardet) in to a single binary -- think something
similar to what docker-compose is doing. By including an LGPL'd module
it is no longer clear if the resulting artefact must also be LGPL'd.
By changing out this dependency for one under MIT we remove all
license ambiguity.
As an "escape hatch" I have made the code so that it will use chardet
first if it is installed, but we no longer depend upon it directly,
although there is a new extra added, `requests[lgpl]`. This should
minimize the impact to users, and give them an escape hatch if
charset_normalizer turns out to be not as good. (In my non-exhaustive
tests it detects the same encoding as chartdet in every case I threw at
it)
Co-authored-by: Jarek Potiuk <jarek@potiuk.com>
Co-authored-by: Jarek Potiuk <jarek@potiuk.com>
* disallow nan values in json serialize
* test nan value in json post
* added exception for invalid json in request
* use invalid json exception
* invalid json test
Nate Prewitt