summaryrefslogtreecommitdiff
path: root/test
Commit message (Collapse)AuthorAgeFilesLines
* win32 compatability adjustmentsunknown2010-07-022-2/+10
|
* Removed blob.data property as there is no real reason for an exception to ↵Sebastian Thiel2010-06-292-10/+0
| | | | the rule of trying not to cache possibly heavy data. The data_stream method should be used instead
* Renamed modules utils to util, and errors to exc to be more conforming to ↵Sebastian Thiel2010-06-294-5/+5
| | | | the submodules's naming conventions
* Removed repo tests which for some reason left the 'repos' directory around, ↵Sebastian Thiel2010-06-291-40/+13
| | | | replaced them by a real test which actually executes code, and puts everything into the tmp directory
* Diff: fixed bug that caused a string to end up as a blob modeSebastian Thiel2010-06-292-90/+95
|
* Fixed performance tests which broke in the course of the sha1-20 byte changesSebastian Thiel2010-06-282-7/+9
|
* All tests adjusted to work with the changed internal sha representationSebastian Thiel2010-06-2810-134/+144
|
* index.reset is now partly implemented using python, but in fact it resorts ↵Sebastian Thiel2010-06-251-7/+0
| | | | to using git-read-tree to keep the stat information when merging one tree in. After all this is what needed to be implemented in python as well
* IndexFile.add: writing of the index file can now optionally be turned off. ↵Sebastian Thiel2010-06-241-1/+2
| | | | The default is to write the physical index, which is the behaviour you would expect
* GitCmdStreamReader: fixed terrible bug which only kicked in if the stream ↵Sebastian Thiel2010-06-241-1/+4
| | | | | | was actually empty. This is a rare case that can happen during stream testing. Theoretically there shouldn't be any empty streams of course, but practically they do exist sometimes ;); fixed stream.seek implementation, which previously used seek on standard output Improved GitCmd error handling
* aggressive_tree_merge: fixed incorrect handling of one branch, it was just ↵Sebastian Thiel2010-06-241-0/+30
| | | | | | not implemented causing incorrect merge results. Added test to cover this issue Diff: added NULL_BIN_SHA constant for completeness
* Multiple partly critical bugfixes related to index handlingSebastian Thiel2010-06-242-9/+14
|
* fixed critical bug in traverse_trees_recursive, implemented IndexFile.new ↵Sebastian Thiel2010-06-232-6/+12
| | | | including simple test, it may be simple as the methods it uses are throroughly tested
* Added test for aggressive_tree_mergeSebastian Thiel2010-06-231-5/+106
|
* Implemented simple tree merging and a simple test, more elaborate testing is ↵Sebastian Thiel2010-06-231-7/+57
| | | | in progress
* Initial frame for implementing read_tree using pure python. As git-read-tree ↵Sebastian Thiel2010-06-224-1/+174
| | | | | | | can do much more than we can ( and faster assumably ), the .new method is used to create new index instances from up to 3 trees. Implemented multi-tree traversal to facilitate building a stage list more efficiently ( although I am not sure whether it could be faster to use a dictionary together with some intensive lookup ), including test Added performance to learn how fast certain operations are, and whether one should be preferred over another
* index.write_tree: now uses MemoryDB, making tree handling more efficient as ↵Sebastian Thiel2010-06-221-1/+6
| | | | IO will only be done when required. A possible disadvantage though is that time is spent on compressing the trees, although only the raw data and their shas would theoretically be needed. On the other hand, compressing their data uses less memory. An optimal implementation would just sha the data, check for existance, and compress it to write it to the database right away. This would mean more specialized code though, introducing redundancy. If IStreams would know whether they contain compressed or uncompressed data, and if there was a method to get a sha from data, this would work nicely in the existing framework though
* Tree-Writing now works after fixing an off-by-one errorSebastian Thiel2010-06-222-8/+29
|
* index.write_tree: initial version implemented, although its not yet working ↵Sebastian Thiel2010-06-221-21/+25
| | | | | | | correctly, a test to explicitly compare the git version with the python implementation is still missing Tree and Index internally use 20 byte shas, converting them only as needed to reduce memory footprint and processing time objects: started own 'fun' module containing the most important tree functions, more are likely to be added soon
* index: put serialization methods into new 'fun' module, this makes the calls ↵Sebastian Thiel2010-06-211-6/+6
| | | | faster as it removes one level of indirection, and makes the main file smaller, improving maintainability
* index.add: now uses gitdb.store functionality instead of git-hash-file. The ↵Sebastian Thiel2010-06-213-594/+621
| | | | python version is about as fast, but could support multithreading using async
* Moved LockedFD and its test into the gitdb projectSebastian Thiel2010-06-151-72/+1
|
* Reimplemented Lock handling to be conforming to the git lock protocol, which ↵Sebastian Thiel2010-06-151-17/+41
| | | | | | is actually more efficient than the previous implementation Index now locks its file for reading, and properly uses LockedFD when writing
* tree: added TreeModifier, allowing to adjust existing trees safely and or ↵Sebastian Thiel2010-06-141-5/+58
| | | | fast, while staying compatible with serialization which requires it to be sorted
* Implemented initial version of tree serialization which appears to work ↵Sebastian Thiel2010-06-142-55/+86
| | | | | | | according to a simple test ( presort still needs implementation ) submodule: added stub to allow the tree to return something, its not implemented though
* Removed odb from project, it is now used as a submodule named gitdb, which ↵Sebastian Thiel2010-06-128-370/+6
| | | | | | was added instead Adjusted all imports to deal with the changed package names
* Removed async from treeSebastian Thiel2010-06-128-955/+0
|
* task: improved naming of task types, improved pool test to be less dependent ↵Sebastian Thiel2010-06-123-19/+26
| | | | on starting with just the main thread
* Cleaned up channel design, Reader and Writer bases don't require a channel ↵Sebastian Thiel2010-06-121-3/+17
| | | | | | anymore, but are abstract. Added IteratorReader, implementing the reader interface from an iterator. The implementation moved from the TaskIterator to the channel
* Added performance test, improved iterator task which will now be usable by ↵Sebastian Thiel2010-06-112-13/+76
| | | | default. It shows that there must be the notion of a producer, which can work if there are no items read
* test_task: fixed import error, made all modules from x import * safeSebastian Thiel2010-06-111-1/+4
|
* Removed commented-out debug code and additional debug printings. Verified it ↵Sebastian Thiel2010-06-111-1/+0
| | | | works on py2.4, 2.5 and 2.6
* Improved shutdown handling - although its impossible to prevent some stderr ↵Sebastian Thiel2010-06-111-5/+3
| | | | printing thanks to the underlying threading implementation, we can at least make sure that the interpreter doesn't block during shutdown. Now it appears to be running smoothly
* Finished dependent task testing according to the features we would currently ↵Sebastian Thiel2010-06-111-6/+42
| | | | like to see
* test.async: split test_pool up into task implenetations and related ↵Sebastian Thiel2010-06-112-180/+196
| | | | utilities, as well as the tests themselves. File became too large
* IMPORTANT: sometimes, when notifying waiters by releasing their lock, the ↵Sebastian Thiel2010-06-111-18/+8
| | | | lock is not actually released or they are not actually notifyied, staying in a beautysleep. This glitch is probably caused by some detail not treated correctly in the thread python module, which is something we cannot fix. It works most of the time as expected though - maybe some cleanup is not done correctly which causes this
* Added dependency-task tests, and fixed plenty of ref-count related bugs, as ↵Sebastian Thiel2010-06-102-34/+111
| | | | well as concurrency issues. Now it works okay, but the thread-shutdown is still an issue, as it causes incorrect behaviour making the tests fail. Its good, as it hints at additional issues that need to be solved. There is just a little more left on the feature side, but its nearly there
* Now tracking the amount of concurrent writers to assure the channel is ↵Sebastian Thiel2010-06-101-1/+2
| | | | closed only when there is no one else writing to it. This assures that all tasks can continue working, and put their results accordingly. Shutdown is still not working correctly, but that should be solvable as well. Its still not perfect though ...
* channel: Changed design to be more logical - a channel now has any amount of ↵Sebastian Thiel2010-06-102-5/+6
| | | | | | readers and writers, a ready is not connected to its writer anymore. This changes the refcounting of course, which is why the auto-cleanup for the pool is currently broken. The benefit of this are faster writes to the channel, reading didn't improve, refcounts should be clearer now
* Added more dependency task tests, especially the single-reads are not yet ↵Sebastian Thiel2010-06-101-19/+110
| | | | fully deterministic as tasks still run into the problem that they try to write into a closed channel, it was closed by one of their task-mates who didn't know someone else was still computing
* InputChannelTask now has interface for properly handling the reading from ↵Sebastian Thiel2010-06-101-3/+4
| | | | the same and different pools
* messy first version of a properly working depth-first graph method, which ↵Sebastian Thiel2010-06-102-40/+37
| | | | allows the pool to work as expected. Many more tests need to be added, and there still is a problem with shutdown as sometimes it won't kill all threads, mainly because the process came up with worker threads started, which cannot be
* test: prepared task dependency test, which already helped to find bug in the ↵Sebastian Thiel2010-06-091-31/+128
| | | | reference counting mechanism, causing references to the pool to be kepts via cycles
* Channel: Callbacks reviewed - they are now part of Subclasses of the default ↵Sebastian Thiel2010-06-091-0/+27
| | | | channel implementation, one of which is used as base by the Pool Read channel, releasing it of the duty to call these itself. The write channel with callback subclass allows the transformation of the item to be written
* Channel: removed pseudoconstructor, which clearly improves the design and ↵Sebastian Thiel2010-06-091-4/+2
| | | | | | makes it easier to constomize pool: in serial mode, created channels will be serial-only, which brings 15% of performance
* Channel: Read method revised - now it really really doesn't block anymore, ↵Sebastian Thiel2010-06-092-6/+8
| | | | and it runs faster as well, about 2/3 of the performance we have when being in serial mode
* HSCondition: Fixed terrible bug which it inherited from its default python ↵Sebastian Thiel2010-06-091-2/+15
| | | | | | Condition implementation, related to the notify method not being treadsafe. Although I was aware of it, I missed the first check which tests for the size - the result could be incorrect if the whole method wasn't locked. Testing runs stable now, allowing to move on \!
* HSCondition: now deriving from deque, as the AsyncQeue does, to elimitate ↵Sebastian Thiel2010-06-091-4/+4
| | | | one more level of indirection. Clearly this not good from a design standpoint, as a Condition is no Deque, but it helps speeding things up which is what this is about. Could make it a hidden class to indicate how 'special' it is
* thread: fixed initialization problem if an empty iterable was handed inSebastian Thiel2010-06-091-4/+13
| | | | | queue: Queue now derives from deque directly, which safes one dict lookup as the queue does not need to be accessed through self anymore pool test improved to better verify threads are started correctly
* queue: fixed critical bug in the notify method, as it was not at all ↵Sebastian Thiel2010-06-091-6/+5
| | | | | | thread-safe, causing locks to be released multiple times. Now it runs very fast, and very stable apparently. Now its about putting previous features back in, and studying their results, before more complex task graphs can be examined
View Lorry Controller Status