43ccbf0459
issue #19 : second attempt at import scanner
...
This version is based on the modulefinder standard library module,
pruned back just to handle modules we know have been loaded already, and
to scan module-level imports only, rather than imports occurring in
class and function scope (crappy heuristic, but assume they are lazy
imports).
The ast and compiler modules were far too slow, whereas this version can
bytecode compile and scan all the imports for django.db.models (58
modules) in around 200ms.. 3.4ms per dependency, it's probably not going
to get much faster than that.
7 years ago
40e2e6eb4c
Add README.md to tests subdir.
7 years ago
175abee2c9
Fix inverted logic.
7 years ago
2fed3613e9
Syntax error.
7 years ago
8054fa867a
Only import compiler when necessary, it's not available in 3.x.
7 years ago
aa817f1b83
Handle failure to import the ast module
7 years ago
dd8271f061
Fix stale references to m.master.connect & m.ssh.connect
...
I'm fairly sure these updates are correct, but I'm not able to run the
test suite to confirm.
7 years ago
d706b7d6b8
Update lingering references to with_broker & run_with_broker
...
A previous commit renamed run_with_broker() and with_broker() to
run_with_router() and with_router() respctively. Some references were
missed.
7 years ago
dc26460a28
Add missing LOG import
...
Fixes an undefined variable error found by pylint
7 years ago
3b24314a8b
Use fully qualified Context
...
Fixes an undefined variable error found by pylint
7 years ago
57a0aa26bf
Add missing import of AnsibleError
...
Found with pylint
7 years ago
f2ea38a963
Add missing auth_incorrect_msg used to in exception message
...
Found using pylint
7 years ago
1f955fc655
Declare self-less method as static
...
Found using pylint
7 years ago
36754254b2
Add missing imports: ast, compiler
...
Found using pylint
7 years ago
01729b18a5
core: use an output deque rather than string to improve worst case perf
...
This probably worsens performance in the common case, but it prevents
runaway producers (see e.g. issue #36 ) from spending all their CPU
copying around huge strings.
It's also a small step towards a solution to issue #6 , which will
replace the output buffer with some sort of fancier queue anyway.
This reduces a particular 40 second run of rsync to 1.5 seconds.
7 years ago
eb060e1a0e
Document Router.profiling.
7 years ago
effe4117e1
Treat EPIPE as disconnect too; needed for fakessh.
7 years ago
9c4bf37cfc
Remove final vestiges of context.key.
7 years ago
0ff65c9780
docs: Use sphinx-programoutput plugin for preamble_size.py
7 years ago
2af1830323
Update preamble_size output.
7 years ago
05cc74d142
core: Support profiling
7 years ago
b827ee1bc7
Delete mitogen.utils.log_to_tmp()
7 years ago
400c971f4a
fakessh: SSH server uses user's shell to parse command line.
7 years ago
b12433695c
fakessh: add to preamble_size.py
7 years ago
04ea5306e9
Document a bunch of mitogen.master and move more docstrings into Sphinx.
7 years ago
1510b72a25
fakessh: Add integration test for inter-child rsync.
7 years ago
b2f13f1fa4
master: propagate routes for IDs allocated via ALLOCATE_ID
...
needed for inter-child rsync.
7 years ago
f01e457d70
Allow running without py.test.
7 years ago
750e9fab24
Make log_to_file accept string level rather than logging package constant
...
Saves an import, pain in the ass to type all the time.
7 years ago
a9387b0504
core: remove pointless eval() of ARGV0 environment variable.
7 years ago
1411790f56
master: make write_all() handle O_NONBLOCK and deadlines.
7 years ago
2ee7309378
master: send_await() always unpickles now.
7 years ago
769ece2514
fakessh: improve logging slightly.
...
Makes it easier to match up reads/writes across log files by using
message byte count
7 years ago
fb9ce1054c
core: set O_NONBLOCK on every side. Closes #33
...
The last time I tested set_nonblock() as a fix for the rsync hang, I
used F_SETFD rather than F_SETFL, which resulted in no error, but also
did not set O_NONBLOCK. Turns out missing O_NONBLOCK was the problem.
The rsync hang was due to every context blocking in os.write() waiting
for either a parent or child buffer to empty, which was exacerbated by
rsync's own pipelining, that allows writes from both sides to proceed
even while reads aren't progressing. The hang was due to os.write() on a
blocking fd blocking until buffer space is available to complete the
write. Partial writes are only supported when O_NONBLOCK is enabled.
7 years ago
db225638f0
core: Make iter_read() handle deadline (and non-blocking IO) properly
7 years ago
092bbef0b3
Reorder table of contents.
7 years ago
caa1fde4fb
Add serialization rules section.
7 years ago
3c8c1f71a6
Update test.
7 years ago
7a60b20dc6
core: Generalize/duplicate the call/send_await code using Receiver.
7 years ago
76d35df889
master: use decorators rather than call_with_deadline(with_context=..)
7 years ago
5dff221842
fakessh: cleanup master test.
7 years ago
76d1e66790
fakessh: use decorators rather than call_with_deadline(with_context=..)
7 years ago
e4c832685d
core: synchronize Stream._output_buf by deferring send()
...
Previously _output_buf was racy. This may or may not be cheaper than
simply using a lock, but it requires much less code, so I prefer it for
now.
7 years ago
ead67de883
core: make Side.write() return None rather than crash if side already closed.
7 years ago
74b31bbe47
core: better Message.__repr__.
7 years ago
2a365aa9b0
Replace `with_context` parameter with mitogen.core.takes_econtext decorator
7 years ago
8d4005c264
fakessh: fix non-master context ID allocation.
7 years ago
bd8001d481
fakessh: Fix NameError.
7 years ago
e7c70127bf
fakessh: return child exit status
7 years ago
77b8d28c68
fakessh: Use local process's sys.executable.
7 years ago
David Wilson