Avoid to output non-errors on stderr in tests/packed_mphf_tests.c
Yeah, stdout would be ok, but this has had this behavior for a long time. Not worth changing.
Merge /u/aerostitch/cmph2/ branch typo into master
Fix typo: Partioning -> Partitioning
Merge /u/aerostitch/cmph2/ branch chmod into master
Change permissions on examples files as C and AM files are not supposed to be executables
add fast HW crc32, cleanup docs
fix and re-enable hashes, various fixes
I agree with the maintainer not to print it to stdout, but stderr
This was fixed upstream with bbf77c63c9e87b0675209fadf53a36528e38bb03
It is generally discouraged in C to cast pointers with small alignment requirements to pointers with bigger alignment requirements. To do so is generally considered a beginner's mistake. For that reason, it is generally recommended to properly align all data. Unaligned data (such as packed structs) can also decrease performance. There are compiler extensions to deal with unaligned access. Packed structs and GCC's aligned attribute can be used, for example. But that will not be portable, and still...
It is generally discouraged in C to cast pointers with small alignment requirements to pointers with bigger alignment requirements. To do so is generally considered a beginner's mistake. For that reason, it is generally recommended to properly align all data. Unaligned data (such as packed structs) can also decrease performance. There are compiler extensions to deal with unaligned access. Packed structs and GCC's aligned attribute can be used, for example. But that will not be portage, and still...
Set theme jekyll-theme-cayman
Minor version bump.
Add docs directory for github.
Apply some of debian patches.
Partially apply https://sourceforge.net/p/cmph/patches/3/
Generating minimal lookup-code for a given algorithm
This has been done before: http://lua-gtk.luaforge.net/reference/modules/src/hash/hash-cmph-bdz.c.html, http://lua-gtk.luaforge.net/reference/modules/src/hash/hash-cmph-fch.c.html. You can just borrow the code from there.
wrong include directive for win32
Adding an ebuild for 0.9
Allow compile under vs2010
Compile under mingw
I failed to get the parts that require struct rusage to work, hence only adopted the patch partially.
Merged, but branch is broken, mostly to simple stuff.
All tests pass.
forget
Need a more robust strategy for implementing slow insert.
All tests pass.
Merge branch 'master' into perfect_cuckoo
More on dummy.
Some progress, but still looping.
More work on dummy, tests seem disabled for some reason.
Tests for power map compile.
Some progress, but still looping.
small fixes
reinventing it
Merge branch 'master' into perfect_cuckoo
Add flat_hash_map to benchmark.
Add a swap function.
Decrease macro pollution.
Add hopscotch to the benchmark baseline.
Rollback confusing ACLOCAL_AMFLAGS.
Make benchmarks optional and dependent on hopscotch_map.
Avoid to output non-errors on stderr in tests/packed_mphf_tests.c
Fix typo: Partioning -> Partitioning
Note: this merge request only removes the executable flags on the source files in the examples directory. No code change. Thanks! :)
Note: this merge request only removes the executable flags on the source files in the examples directory. No code change.
Change permissions on examples files as C and AM files are not supposed to be executables
The 2nd part doesn't seem to be valid. It has been specified in the BRZ spec that this has been done as the memory is limited. Kindly address the first part. Also, change the priority to 5.
BRZ Algorithm segmentation fault and wastage of disk resources without using -m parameter.
Generating minimal lookup-code for a given algorithm
Please mark this as resolved. The issue was due to the length argument of the search...
The issue was with the lenght argument in search() function. It was considering the...
The outputs generated by the cmph application and the program (using the value returned...
The outputs generated by the cmph application and the program (using the value returned...
The MPHFs for CHD algorithms do not generate unique ids even though the parameters 't' and 'b' are set to 1.
FYI the code has been moved to https://github.com/thomasmueller/minperf. Now the...
FYI the code has been moved to https://github.com/thomasmueller/minperf
This is normal and expected for a perfect hash function.
I found that the hash function description is smaller if I use ./src/cmph -a chd...
Hi, I tried to use the command line tool to generate a minimal perfect hash for 1...
Hi, I tried to use the command line tool to generate a minimal perfect hash for 1...
I have implemented a perfect hash / minimal perfect hash algorithm, and I'm not quite...
I would also like to know the answer to this question.
Simple fix to gendocs.
Merge branch 'master' of ssh://git.code.sf.net/...
Merge branch 'master' of ssh://git.code.sf.net/...
First tentative of minor version bump.
Add c++11 m4 check.
Tentative fix to m4 stuff.
A couple test cases.
Merge commit 'c85b8f8ecd23039aa9794098735f896c2...
Merge commit '2e797796a545748ea815f39113088c701...
Style fixes to https://sourceforge.net/p/cmph/g...
Fixed initialization order of test framework.
Merge remote-tracking branch 'sf/githubmaster'
Fixed initialization order of test framework.
A couple test cases.
Merge commit 'efe08b8080d0696bf388b21'
Revert "Fix memory leak problem for bmz8"
Small syntatic fixes.
Dropped inline from Murmur for gcc 4.8 compatib...
Static code generation
Hi Huang-Ming, I have incorporated the fix for disposing keys in FCH, but the patch...
Thanks a lot Joseph.
Hi Thomas, Thanks a lot for the patch. I think it is a nice proof of concept, but...
Hi Joseph, Thanks a lot for the patch. We appreciate. Let us know if you need anything...
After having constructed a MPH, is lookup (cmph_search()) thread safe? In principal...