[BART-devel] fileexistopen() Refactor

[aj-foster]

The following is an archived message from the BART-devel mailing list, which has now closed.

From: aj***@kn*** (AJ Foster)
Date: Wed, 17 Jun 2015 18:09:31 +0000
Subject: [BART-devel] fileexistopen() Refactor

Hello all,

At your convenience, please take a lo***@th***ollowing commit:

https://github.com/aj-foster/transit/commit/5c847b5d57c81a829351ba272f485cc4cf47216b

The purpose of the change is to allow us to check for a file's existence without automatically loading the file. For example, when shared memory is being used, we want transit to check for the opacity file's existence but not load it until we're sure that it hasn't already been copied into the shared memory.

Do you agree with this change? Do you foresee any issues with it? I want to be sure, since this function's scope is larger than the code I'm currently looking at.

Thanks,

• AJ Foster

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://physics.ucf.edu/pipermail/bart-devel/attachments/20150617/2f539bd9/attachment.html

[aj-foster]

From: rc***@kn*** (Ryan Challener)
Date: Wed, 17 Jun 2015 18:30:16 +0000
Subject: [BART-devel] fileexistopen() Refactor

That should be fine. I can't think of any issues off the top of my head, since I imagine the only case it would currently receive a NULL pointer would be if the file was not specified, in which case it certainly won't exist and the function would return an error.

I have been using the access() function from unistd.h to check for file existence, but it maybe it would be preferred to use our own function.

-Ryan

---

From: BART-devel <ba***@ph***> on behalf of AJ Foster <aj***@kn***>
Sent: Wednesday, June 17, 2015 2:09 PM
To: ba***@ph***
Subject: [BART-devel] fileexistopen() Refactor

Hello all,

At your convenience, please take a lo***@th***ollowing commit:

https://github.com/aj-foster/transit/commit/5c847b5d57c81a829351ba272f485cc4cf47216b

The purpose of the change is to allow us to check for a file's existence without automatically loading the file. For example, when shared memory is being used, we want transit to check for the opacity file's existence but not load it until we're sure that it hasn't already been copied into the shared memory.

Do you agree with this change? Do you foresee any issues with it? I want to be sure, since this function's scope is larger than the code I'm currently looking at.

Thanks,

• AJ Foster

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://physics.ucf.edu/pipermail/bart-devel/attachments/20150617/7076d651/attachment.html