aboutsummaryrefslogtreecommitdiffstats
path: root/fusearchive.py
Commit message (Collapse)AuthorAgeFilesLines
* Turned off profilingSteve Slaven2009-08-071-1/+1
|
* Switched to cProfileSteve Slaven2009-08-071-5/+4
|
* Kill the logging.debug function if we're not debugging, otherwise it showsSteve Slaven2009-08-071-0/+4
| | | | | up in the top 3 functions in profiling, making it mostly nop speeds things up a bit
* Moved everything to main() so we can run the profiler againSteve Slaven2009-08-071-34/+38
|
* Changed the way we look for the debug flag and actually turn the flag onSteve Slaven2009-08-071-2/+1
| | | | ONLY when we see debug enabled
* Turn on debugging with -d instead of having to edit the scriptSteve Slaven2009-08-071-9/+13
|
* Update usage, took out xmp.py sample copyrights since almost all of xmp.pySteve Slaven2009-08-071-4/+4
| | | | has been replaced
* Fix namespace problemSteve Slaven2009-08-041-2/+2
|
* Split up main classesSteve Slaven2009-08-041-982/+34
|
* Use some more printf like % stuff instead of str()before-class-splitSteve Slaven2009-07-291-6/+6
|
* Eliminated some more str() stuffSteve Slaven2009-07-291-6/+9
|
* Changed debug to use param args instead of way too much + str()Steve Slaven2009-07-291-76/+72
|
* Use "best" cpickle formatSteve Slaven2009-07-291-1/+1
|
* Changed default log levelSteve Slaven2009-07-291-1/+1
|
* Be more explicit about saving chunks, since some other parts of the codeSteve Slaven2009-07-291-2/+8
| | | | look at len(self.chunks) and this is NOT extended until we do a _save_chunk
* Catch when size == truncate size, some more debugging stuffSteve Slaven2009-07-291-2/+4
|
* Fixed some off-by-one errors in write when extending the blocks with nullsSteve Slaven2009-07-291-3/+6
|
* Added some more debug info, assert that our calculated size when savingSteve Slaven2009-07-291-6/+20
| | | | matches what we thought it would
* More fixes to ftruncate to expand a file which was causing the "fake" mmapSteve Slaven2009-07-281-16/+17
| | | | errors in fsx, fixed loading the right block during write()
* This is a workaround for a bug with lseek causing a getattr instead ofSteve Slaven2009-07-281-5/+30
| | | | | | fgetattr on an already opened file, so it reports the wrong file size if you use seek_end for whence by reading the on-disk info instead of the open-file info
* self.offset was never usedSteve Slaven2009-07-281-3/+0
|
* Only add O_CREAT if it was passed initiallySteve Slaven2009-07-281-1/+5
|
* Do create new fh if neededSteve Slaven2009-07-281-1/+1
|
* Fix not incrementing blocks while fillingSteve Slaven2009-07-281-1/+2
|
* Bug fixes to extend via seek(), properly setting dirty flag and not tryingSteve Slaven2009-07-281-4/+12
| | | | to pull up block -1
* Always open the main fh for serialization in rw modeSteve Slaven2009-07-281-4/+8
|
* Fixed log message and added log message for serializationSteve Slaven2009-07-281-3/+6
|
* Added an exception printer on the file setup since it seems to crash thereSteve Slaven2009-07-281-12/+10
| | | | a lot when there are file serialization problems
* Fix flag detection (rdonly,rdrw,etc), fix some variables shadowingSteve Slaven2009-07-281-39/+50
| | | | | functions/classes, modified write() to extend the file with nulls if needed, like seeking out to a larger length than the file
* Fix checking for debug level to output lots of data to use the logger levelSteve Slaven2009-07-281-7/+7
|
* Changed to log to stderrSteve Slaven2009-07-281-1/+1
|
* Use the logging package instead of dmsgSteve Slaven2009-07-281-108/+110
|
* Modify ftruncate to allow growing or shrinking to a size other than 0Steve Slaven2009-07-281-5/+64
|
* Modified to open a fh at the init using the mode requested, and justSteve Slaven2009-07-281-26/+45
| | | | | shuttling the pickle data in and out of this fh so the cp -a on files with perm of 444 should work correctly
* A workaround attempt to fix opening a file for writing then changing it toSteve Slaven2009-07-281-3/+22
| | | | R/O later (happens with cp -a, etc)
* Allow switching chunk stylesSteve Slaven2009-07-241-2/+13
|
* Testing out using zip files for chunk storage to eliminate some of the hugeSteve Slaven2009-07-241-3/+103
| | | | overhead of having thousands of tiny files and directories
* Fixed some pylint/pychecker/pyflakes errorsSteve Slaven2009-07-241-10/+7
|
* Re-enable gzipfinal-demand-chunking-before-mergeSteve Slaven2009-07-241-2/+2
|
* Fix read bug on buffer boundry, make psyco a flagSteve Slaven2009-07-241-13/+40
|
* Fix a bug with writing partial blocksSteve Slaven2009-07-241-2/+10
|
* Made stats a flag, moved around flags to group related ones togetherSteve Slaven2009-07-241-4/+6
|
* Added a flag that basically disables all IO to test for "perfect"Steve Slaven2009-07-241-5/+20
| | | | performance to see how much fuse and gzip is hurting things
* Fix a huge bug with reading data blocks sometimes returning more data thanSteve Slaven2009-07-241-8/+10
| | | | | the caller requested, renamed fusearechivewriter to fusearchivestream because it's for both reading and writing
* Switched to cpickle and added wrappers for file open/dump/load stuff soSteve Slaven2009-07-241-13/+32
| | | | it's easier to experiment with different methods
* Added binary flag and gzip compression levelSteve Slaven2009-07-231-7/+8
|
* Bumped block size a bit, initialize current block to -1 so that we don'tSteve Slaven2009-07-231-3/+3
| | | | think block zero is already loaded
* Make file ready handle cached dirty pagesSteve Slaven2009-07-231-2/+6
|
* Added caching some dirty blocksSteve Slaven2009-07-231-10/+40
|
* Added some un-commentables for profilingSteve Slaven2009-07-231-2/+15
|