Atul Varma
90d243db16
Refactored test_service and module code so that the weave service constructor is called, and auth is done more accurately.
2008-06-17 19:54:09 -07:00
Atul Varma
945a8b782a
The manage.py test-runner now compares expected results to actual results if a '<test-name>.log.expected' file exists in the test directory; if they don't match, a unified diff is displayed. Note that this is only done via manage.py, it's not currently implemented in the Makefile test framework because makefiles make me barf.
2008-06-17 12:04:40 -07:00
Atul Varma
a6cca49fc3
Added a fake preference service to test_service.
2008-06-17 11:45:13 -07:00
Atul Varma
6b20868829
Fixed a minor typo, though there's a better way to do the unary getter which Myk told me about; will probably substitute that soon.
2008-06-17 11:11:56 -07:00
Atul Varma
f9463209a2
Added a fixme/todo in the code.
2008-06-17 10:23:35 -07:00
Dan Mills
dec04b1bc3
merge upstream changes
2008-06-17 18:04:16 +09:00
Dan Mills
0c3a240876
make RemoteStore a but more Store-like (wrap, wipe methods); fix a missing yield
2008-06-17 18:03:02 +09:00
Dan Mills
5e07cb98d5
Don't allow re-entrance! Fail to lock when we already hold a lock.
2008-06-17 18:01:48 +09:00
Atul Varma
245e6492da
Fixed a strict warning in util.js.
2008-06-16 18:11:41 -07:00
Atul Varma
b0ca747fe2
Added test_service, which currently tests the case in which the server's meta/version and private/privkey files are correct, as well as all authentication information.
2008-06-16 18:08:37 -07:00
Dan Mills
99e6994409
merge upstream changes
2008-06-17 09:51:25 +09:00
Dan Mills
5c196a641a
treat paths beginning with '/' as absolute
2008-06-17 09:51:02 +09:00
Atul Varma
8d40fc280b
Refactored test suites based on recent changes.
2008-06-16 16:53:53 -07:00
Atul Varma
459e94fa97
Factored out fake-timer code into a separate class and moved it to head.js.
2008-06-16 16:42:32 -07:00
Atul Varma
de38af3819
Refactored head.js for unit tests so that we now have an 'initTestLogging()' function.
2008-06-16 16:22:00 -07:00
Atul Varma
2675855ed4
Defined Cu -> Components.utils in head.js so that test files don't have to put it in as boilerplate.
2008-06-16 14:06:05 -07:00
Dan Mills
f8a841ee5e
move more remote-interaction code into RemoteStore (and out of Engine)
2008-06-17 01:04:23 +09:00
Dan Mills
f74b9f910f
catch 404s from initSession and do an initial upload; fix some bugs (missing yields, references to Engine properties) in RemoteStore.initialize
2008-06-16 00:21:44 +09:00
Dan Mills
a8ad7d2804
merge upstream changes
2008-06-15 13:42:10 +09:00
Dan Mills
561b27a92c
merge upstream changes
2008-06-14 17:10:40 +09:00
Dan Mills
049ecf1de3
Move remote init code ('fullUpload') into RemoteStore; make RemoteStore hold an Engine object (tightly coupling them); make the server prefix and identity properties of Engine public
2008-06-14 17:07:06 +09:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
9805cdce23
Merged
2008-06-13 16:22:04 -07:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
835100b96d
Fixed some minor bugs -- the name of the incoming shared folder is parsed correctly, and xmppClient now catches bounce errors that were previously parsed as messages.
2008-06-13 16:20:43 -07:00
Atul Varma
92583fc5ff
Fixed an 'outstanding callbacks' warning and simplified the code a tiny bit.
2008-06-13 15:39:06 -07:00
Myk Melez
b12f3d03db
use nsIXMLHttpRequest::mozBackgroundRequest instead of DummyAuthProvider to suppress authentication dialogs and ensure XMLHttpRequests succeed even when the window that originated the request goes away
2008-06-13 13:08:36 -07:00
Dan Mills
589f3222eb
To avoid some async.js warnings: Turn some 'yield' calls into 'return' ones, and don't call 'self.cb' twice in a row when constructing event listeners for XHRs.
2008-06-13 15:49:18 +09:00
Dan Mills
9812ba557e
Make some 'debug' log calls into 'trace' ones. Use the frame formatter, which will remove long paths to extension dirs from the output. Don't warn about outstanding callbacks if we caught an exception from the generator.
2008-06-13 15:47:41 +09:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
5c8481de3d
Merged
2008-06-12 17:35:57 -07:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
c548d4b940
Made the initialization of the xmppClient an asynchronous call. This included modifying xmppClient.js so that connect() can be passed a callback function that will get called once the connection has succeeded or failed. For most of our purposes this is probably a better API than what we had before where you call waitForConnection() and it busy-waits until the connection has succeeded or failed.
2008-06-12 17:35:44 -07:00
Myk Melez
b4967c8831
work around XmlHttpRequest bug 317600 by pausing for a 0ms timeout before trying to log in
2008-06-12 16:23:59 -07:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
b462aad4a5
Merged changes.
2008-06-12 14:30:46 -07:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
2b055a3b87
Fixed the server URL for xmpp connections.
2008-06-12 14:30:39 -07:00
Myk Melez
18d8e8fcd7
bug 410550: stop running scheduled sync when the user is not logged into weave
2008-06-12 12:36:58 -07:00
Atul Varma
9d32240087
Changed some of the debug() logging statements I added a few commits ago into trace() statemetns b/c they were drowning the log.
2008-06-12 12:00:19 -07:00
Myk Melez
f47209b9d2
only retrieve score once per engine when doing a scheduled sync, since retrieving the score can be a non-negligible cost for trackers that calculate the score on-demand (like the tab tracker)
2008-06-11 20:07:35 -07:00
Myk Melez
669e57136f
Automated merge with ssh://hg.mozilla.org/labs/weave/
2008-06-11 20:00:59 -07:00
Myk Melez
5d50fffd37
don't sync tab entry IDs, which change with every session, to avoid generating edit commands for every tab on restart even when the tabs haven't actually changed
2008-06-11 20:00:48 -07:00
Atul Varma
4864f07ae2
async.js now keeps track of how many outstanding callbacks it has and uses this information to log warnings about coroutines that may have yielded without an outstanding callback, and coroutines that may have finished while a callback is still outstanding. These are merely 'warnings' rather than certainties because this code assumes that there is a 1:1 correspondence between accesses to self.cb and yields, and also that self.cb's are actually passed to asynchronous functions. It'd be really cool if we could actually keep track of whether a callback got garbage collected before it was called or something, though I don't know how much it'd help in the end.
2008-06-11 19:19:16 -07:00
Atul Varma
7d98a5dbeb
Added a few log messages to hopefully make the debugging of generators easier. Also added an id component to generators, which is part of their name, to help distinguish between concurrent instances of the same generator function. The following debug output represents the new logging infomation:
...
--
Async.Generator DEBUG runTestGenerator-0: self.cb generated at test_async_missing_yield.js:28
Async.Generator DEBUG secondGen-1: self.cb generated at test_async_missing_yield.js:20
Async.Generator DEBUG secondGen-1: done() called.
Async.Generator DEBUG runTestGenerator-0: self.cb() called, resuming coroutine.
Async.Generator DEBUG runTestGenerator-0: done() called.
Async.Generator DEBUG secondGen-1: self.cb() called, resuming coroutine.
Async.Generator DEBUG secondGen-1: done() called.
Async.Generator ERROR Async method 'secondGen-1' is missing a 'yield' call (or called done() after being finalized)
--
As you can see, I've added log messages whenever the Generator's 'cb' property is accessed--this is almost guaranteed to be very close to a 'yield' statement, and therefore provides us with a decently accurate idea of where the generator 'stopped'. We also log a message when the generator continues, and by doing so we get an idea of how the coroutines interleave.
Another idea I had was to actually match calls to self.cb with calls to 'yield' to automatically detect e.g. two yields in a row (which will ordinarily result in a generator 'hanging'), a generator exiting while a self.cb still hasn't been called, but I'm not sure what kinds of reprecussions it may have.
2008-06-11 18:58:30 -07:00
Myk Melez
a850ec7b70
numChanged should be the number of shared items whose data is different, not the same
2008-06-11 18:47:56 -07:00
Atul Varma
28a0971380
Merged changes.
2008-06-11 18:03:11 -07:00
Atul Varma
78c2bba04d
Added test_async_missing_yield. It's very messy right now and duplicates code from other tests, but I've got some ideas about how to write better tests for async ops that I'll commit soon.
2008-06-11 18:02:46 -07:00
Maria Emerson
7da4259fee
bug 438033: implement a better first-run wizard process; r=myk
2008-06-11 17:56:02 -07:00
Myk Melez
f9ef596828
Automated merge with ssh://hg.mozilla.org/labs/weave/
2008-06-11 17:45:54 -07:00
Myk Melez
0a15090c8c
bug 437529: yield after starting to put the status file to the server so we don't finalize the sync until the PUT request completes
2008-06-11 17:44:08 -07:00
Atul Varma
d32ba01952
Modified test_async_exceptions to use a fake nsiTimer.
2008-06-11 17:10:39 -07:00
Atul Varma
561d94aa66
Added a unit test for async exceptions.
2008-06-11 16:38:22 -07:00
jonathandicarlo@jonathan-dicarlos-macbook-pro.local
97f788bb51
Fixed Utils.prefs.getStringPref (should be getCharPref)
2008-06-11 15:43:12 -07:00
Myk Melez
8182b10abf
resetting the score is not an asynchronous operation, so Service::_syncEngine shouldn't yield after calling it
2008-06-11 15:23:54 -07:00
Myk Melez
633afe0864
clarify wording in scheduled sync threshold debug statements
2008-06-11 14:16:03 -07:00
