Skip to content

0.9.x#107

Closed
mrawd wants to merge 56 commits into
masterfrom
0.9.x
Closed

0.9.x#107
mrawd wants to merge 56 commits into
masterfrom
0.9.x

Conversation

@mrawd
Copy link
Copy Markdown

@mrawd mrawd commented Jun 5, 2017

No description provided.

rhs and others added 30 commits March 10, 2015 00:37
(cherry picked from commit bc2b630)
cherry-picked from 36e32d2
Remove one of calls to transport.connecting() as this causes
an assert with SslTransport

(cherry picked from commit f8ca35f)
…esponseTest.test_request_response ........ fail

Fix to reactor.py, check for lack of SSL domain.

(cherry picked from commit e31df01)
After commit c65e897 it turned out there were still some issues with
strings containing a codepoint >0xDBFF which was being incorrectly
treated as a surrogate pair in the calculateUTF8Length method.

Fixed this up and added some more test coverage.

Closes #13

(cherry picked from commit 7b9b516)
Having switched to using git-archive, we end up creating a tar that uses
the PAX-based scheme for supporting long pathnames. This causes issues
for a number of extraction tools (ant, 7-zip, etc.) which only support
the GNU tar format. We should stick with GNU tar, as used for previous
qpid-proton releases.

Also update the script to revert to using a suffix of .tar.gz (rather
than .tgz) for the source distribution.

Closes #14
As discussed on the mailing list. Currently proton-j advertises a
provided local idle timeout value as-is, whereas proton-c will halve it
before sending it to the remote side. The two implementations should
match in behaviour.

Closes #15

(cherry picked from commit 0b439c1)
* when receiving an attach with a handle > handle-max, the connection
  should be closed
* when attempting to allocate a new local handle when handle-max are
  already in-use should throw an Exception
* the default of 1024 handle-max is quite small when proton-c appears to
  default to 4294967295 - increase it more modestly to 65536

Closes #16

(cherry picked from commit f937ccd)
Still the encoder wasn't 100% compliant in 0.9. Also fixup the unittest
to run a touch faster.

(cherry picked from commit 836cf27)
The getter was incorrectly dividing the value in half before returning
it.

(cherry picked from commit 677ea33)
Check for no Headers object and create one as needed.
(cherry picked from commit b541ad0)
…ink state in maps, use the references set on the associated Session and Link objects. Update channel+link reference handling to behave more like proton-c in order to resolve the resulting test failures.

This closes #20

(cherry picked from commit f7e7ddd)
…ce it doesnt actually fail without the change.

This reverts commit 7d3063e.

(cherry picked from commit d2262bb)
…losed state, instead create a new object and ensure the old links also get freed. Also fixes similar behaviour as in PROTON-850.

This closes #21

(cherry picked from commit f2d7d66)
…e arrays we typically wont utilise fully

Also reduces handle-max to 65535, giving 65536 usable handles like proton-c currently has.

(cherry picked from commit 74e16df)
… a tutorial to accompany the examples.

(cherry picked from commit 4653cdc)
Please enter the commit message for your changes. Lines starting

(cherry picked from commit df2cd6c)
grs and others added 26 commits April 27, 2015 15:12
* remove out-of-date product information from the README and include the more
  current information from the website
* add a new INSTALL file to contain the build and install instructions
  from the existing README
* move all development and testing related info from that to the
  DEVELOPERS file
* better explain how to run the various test suites and how they work
  across proton-c and proton-j in the developers file
* other misc improvements across the contents of the files
* rename DEVELOPERS and INSTALL to have .md suffixes for better
  rendering on GitHub
* add a symlink from README --> README.md for better landing page
  rendering on GitHub

(cherry picked from commit b532cf2)
Closes #22

(cherry picked from commit c1a6de2)
The system-installed Perl, Python etc. are all built as universal
binaries on Mac and so we need to build libqpid-proton to similarly
contain both architectures. This allows the bindings to link
successfully.

(cherry picked from commit a509195)
(cherry picked from commit da117ea)
(cherry picked from commit 0c1de24)
The SelectableMessengerTest was failing intermittently on Mac builds,
Modify the test slightly to be more resilient.

Closes #23

(cherry picked from commit dd95103)
As of now, it's not possible to install python-qpid-proton if
libqpid-proton is not present in the system. To be more precises, it's
possible to build python-qpid-proton using cmake, upload it and beg to
the gods of OPs that the required (and correct) shared library will be
present in the system.

This patch adds to python-qpid-proton the ability to download, build and
install qpid-proton if the required version is not present in the
system. It does this by checking - using pkg-config - whether the
required version is installed and if not, it goes to downloading the
package from the official apache source and builds it using cmake.

As nasty as it sounds, this process is not strange in the Python
community. Very famous - and way more used - libraries like PyZMQ (from
which this work took lots of inspiration) do this already in a fairly
more complex way.

This first step is quite simple, it checks, downloads and builds using
the standard tools. It's enabled just for linux and it does not use
fancy flags. Future enhancements could take care of improving the
implementation and extending it to support other systems.

(cherry picked from commit 2e2ba9c)
Instead of using cmake to build the downloaded tarball, rely on python
for building and installing the library. This process is not only
cleaner but also plays nicer with the environment we're trying to give
support to.

In addition to the above, this plays nicer with the user and removes a
requirement for having the library installed. Note that the built
library will no longer overwrite system libraries but it'll be placed
along to the python bindings we're building.

(cherry picked from commit 4ed130e)
(cherry picked from commit 70a463f)

Conflicts:
	proton-c/bindings/python/setup.py
@mrawd mrawd closed this Jun 5, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

10 participants