You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucenenet.apache.org by "Van Den Berghe, Vincent" <Vi...@bvdinfo.com> on 2017/06/12 12:23:46 UTC
Pull requests
Hello,
I've done a couple of "pull requests", but for some reason they do not appear. Maybe it's just as well. They are based on "master", which I understand is the most recent version now. I've made them on my current "clone" of the master. Maybe someone can review them and integrate them as they please.
The two pull requests are:
(1)
https://github.com/vvdb/lucenenet/commit/3e3dfd37e4d20107526a911b1e083519659248bf
This implements the more efficient "Put" operation on memory mapped files, analogous to the "Get" operations which was implemented by yours truly as well.
(2) https://github.com/vvdb/lucenenet/commit/6437b37612c575662ececbf2c4b8a0f4c6830f09
This implements "LUCENE-5644: switch to simpler LIFO thread to ThreadState allocator during indexing". Technically, this is something from releases/lucene-solr/4.8.1, but profiling indicates it makes a huge difference in multithreaded scenarios and I had to implement it. Tests have been adapted, and everything runs.
Vincent
RE: Pull requests
Posted by "Van Den Berghe, Vincent" <Vi...@bvdinfo.com>.
Hello Shad,
I've finally managed to do a pull request again. There must be something wrong with my brain.
I've pushed (2). I'll refrain to push (1).
Vincent
-----Original Message-----
From: Shad Storhaug [mailto:shad@shadstorhaug.com]
Sent: Monday, June 12, 2017 11:22 PM
To: dev@lucenenet.apache.org
Subject: RE: Pull requests
Vincent,
(1)
From the documentation of MMapDirectory:
File-based Directory implementation that uses MemoryMappedFile for reading, and FSDirectory.FSIndexOutput for writing.
I didn't see the point in maintaining the various write methods in the MemoryMappedFileByteBuffer, since it is not used for writing anyway, so all of the Put() overloads just throw a NotSupportedException().
(2)
This sounds like a good fix to include.
Note that to make a pull request, you need to first make a branch on your end, push the branch to GitHub, and then make the pull request (targeting master). This way you can have several open pull requests at the same time that each pertain to something different.
Thanks,
Shad Storhaug (NightOwl888)
-----Original Message-----
From: Van Den Berghe, Vincent [mailto:Vincent.VanDenBerghe@bvdinfo.com]
Sent: Monday, June 12, 2017 7:24 PM
To: dev@lucenenet.apache.org
Subject: Pull requests
Hello,
I've done a couple of "pull requests", but for some reason they do not appear. Maybe it's just as well. They are based on "master", which I understand is the most recent version now. I've made them on my current "clone" of the master. Maybe someone can review them and integrate them as they please.
The two pull requests are:
(1)
https://github.com/vvdb/lucenenet/commit/3e3dfd37e4d20107526a911b1e083519659248bf
This implements the more efficient "Put" operation on memory mapped files, analogous to the "Get" operations which was implemented by yours truly as well.
(2) https://github.com/vvdb/lucenenet/commit/6437b37612c575662ececbf2c4b8a0f4c6830f09
This implements "LUCENE-5644: switch to simpler LIFO thread to ThreadState allocator during indexing". Technically, this is something from releases/lucene-solr/4.8.1, but profiling indicates it makes a huge difference in multithreaded scenarios and I had to implement it. Tests have been adapted, and everything runs.
Vincent
RE: Pull requests
Posted by Shad Storhaug <sh...@shadstorhaug.com>.
Vincent,
(1)
From the documentation of MMapDirectory:
File-based Directory implementation that uses
MemoryMappedFile for reading, and
FSDirectory.FSIndexOutput for writing.
I didn't see the point in maintaining the various write methods in the MemoryMappedFileByteBuffer, since it is not used for writing anyway, so all of the Put() overloads just throw a NotSupportedException().
(2)
This sounds like a good fix to include.
Note that to make a pull request, you need to first make a branch on your end, push the branch to GitHub, and then make the pull request (targeting master). This way you can have several open pull requests at the same time that each pertain to something different.
Thanks,
Shad Storhaug (NightOwl888)
-----Original Message-----
From: Van Den Berghe, Vincent [mailto:Vincent.VanDenBerghe@bvdinfo.com]
Sent: Monday, June 12, 2017 7:24 PM
To: dev@lucenenet.apache.org
Subject: Pull requests
Hello,
I've done a couple of "pull requests", but for some reason they do not appear. Maybe it's just as well. They are based on "master", which I understand is the most recent version now. I've made them on my current "clone" of the master. Maybe someone can review them and integrate them as they please.
The two pull requests are:
(1)
https://github.com/vvdb/lucenenet/commit/3e3dfd37e4d20107526a911b1e083519659248bf
This implements the more efficient "Put" operation on memory mapped files, analogous to the "Get" operations which was implemented by yours truly as well.
(2) https://github.com/vvdb/lucenenet/commit/6437b37612c575662ececbf2c4b8a0f4c6830f09
This implements "LUCENE-5644: switch to simpler LIFO thread to ThreadState allocator during indexing". Technically, this is something from releases/lucene-solr/4.8.1, but profiling indicates it makes a huge difference in multithreaded scenarios and I had to implement it. Tests have been adapted, and everything runs.
Vincent