ENH: remove all prior existing citation files first #1567
Conversation
- makes it friendly to the cases where all output is under git-annex. .write_text() just opens a file in "w" mode so it would fail on annex protected file. - if previous run did succeed generating e.g. .html file, if subsequent run fails, updated version of .html would not be generated and provided CITATION files could get out of sync. Removing all CITATION* files prior generation resolves both of the issues.
They are suggestsions from code review by Oscar Co-Authored-By: yarikoptic <debian@onerussian.com>
da3cf26 to
19c7e71
Compare
|
@yarikoptic, do you want to add yourself to the |
The order is not really clear, so added myself all the way at the end before Russ P.
|
done... wasn't sure where to place myself so 38cfd77 |
|
That is correct, will merge in soon. Thanks a lot!
…On Thu, Apr 4, 2019, 19:53 Yaroslav Halchenko ***@***.***> wrote:
done... wasn't sure where to place myself so 38cfd77
<38cfd77>
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#1567 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAkhxg2TqHlN4m3QSSa7lItOT6gcG4nRks5vdrqegaJpZM4cbBTe>
.
|
|
BTW, this made me think whether we should also delete other top-level files (e.g. |
|
well, if you keep thinking then may be you will even converge on my thinking that those files should not be just removed but loaded/compared to and if they are different - at least a warning or info message should be displayed like "you are now using different preprocessing settings from what you have used before - you might arrive to an incongruent collection of results (may be across or even within subjects!)" ;-) or that is not possible? ;-) |
|
Well, that would apply to the boilerplate too, since it is dynamic and reflects the methodological choices. |
… fully run. This PR addresses the memory peak caused by pandoc, running the process at the very end of fMRIPrep. We probably want to explicitly clear up as many objects as we can from memory and force gc before the subprocess opens pandoc, but that's left for a future PR. This PR just closes nipreps#1709 with the delayed pandoc run. I'm not implementing the flag to avoid the boilerplate generation because I'm not sure we want to introduce an additioinal argument for this. Coincidentally, it also closes nipreps#1667 - now the deletion of citation files is done inside a try..catch where ``FileNotFoundError`` is dismissed. That should avert the issue, while keeping the functionality of nipreps#1567.
makes it friendly to the cases where all output is under git-annex.
.write_text() just opens a file in "w" mode so it would fail on
annex protected file.
if previous run did succeed generating e.g. .html file, if subsequent run
fails, updated version of .html would not be generated and provided
CITATION files could get out of sync.
Removing all relevant
CITATION*files prior generation resolves both of the issues.No pressure -- if you don't like it just close it.
Comes not tested whatsoever ;-) (are there unittests to run for it?)