Patch 1503 resolve issue2482: mark-conflicts FILE

Title resolve issue2482: mark-conflicts FILE
Superseder Nosy List bf
Related Issues
Status accepted Assigned To

Created on 2016-05-06.20:26:31 by bf, last changed 2017-10-10.17:03:17 by gh.

File name Status Uploaded Type Edit Remove
patch-preview.txt bf, 2016-05-06.20:26:30 text/x-darcs-patch
resolve-issue2482_-mark_conflicts-file.dpatch bf, 2017-08-11.19:04:30 text/x-darcs-patch
unnamed bf, 2016-05-06.20:26:30 text/plain
wip_-resolve-issue2482-darcs-mark_conflicts-file.dpatch bf, 2016-05-06.20:26:30 application/x-darcs-patch
See mailing list archives for discussion on individual patches.
msg19217 (view) Author: bf Date: 2016-05-06.20:26:30
This is not (yet) for screened; it contains (unconditional) debug output and
I am not sure I have covered all cases. I would love to get some feedback.

The code is more or less a re-write, but it seems to work... The support
code at the end was added mostly because I was getting headaches trying to
reconcile all the different types used for (argument-) path lists, I simply
needed some structure where I could encapsulate things like conversions,
intersection, and lifting. This patch has been lying around here for too
long and I don't want to let the work go to waste just because the result is
not yet perfect. I hope I'll find the time to finish it...

1 patch for repository http://darcs.net/screened:

patch d4ba03a976c89a0f219772be2d6a9e1cb36ffe39
Author: Ben Franksen <benjamin.franksen@helmholtz-berlin.de>
Date:   Tue Mar 15 18:23:47 CET 2016
  * WIP: resolve issue2482 darcs mark-conflicts FILE
msg19552 (view) Author: bf Date: 2017-08-11.19:04:30
This is a rebased version with tamed debug output (using debugDocLn instead
of putDocLn). It doesn't look as if I will find the time to further improve
it, so I think we should use it as is, as suggested by Ganesh.
msg19553 (view) Author: bf Date: 2017-08-11.19:07:23
(Suggestion was by Guillaume, not Ganesh, sorry)
msg19611 (view) Author: gh Date: 2017-08-20.14:38:08
A superficial question: Why add FilePath-related functions in 
Darcs.UI.Commands.MarkConflicts and not in Darcs.Util.Path ?

Isn't this new Only datatype useful also in other commands in which
specifying files restricts the action to them (from a quick glance,
whatsnew, record, amend, revert, unrevert, log) ?
msg19622 (view) Author: bf Date: 2017-08-21.09:00:09
gh: I wanted to avoid adding yet another file path related abstraction
until the new one really proved to be useful enough to warrant such an
addition. I remember having great difficulty to come to grips with the
many existing ones when I started hacking on darcs (RelativePath,
AbsolutePath, + OrURL variants, FileName, AnchoredPath, etc not to
mention the multiple conversion functions between them), some of whom
are of questionable utility if you ask me. It is usually a lot harder to
get rid of such a thing than to introduce it. The path related support
code I added was written so that it can be easily factored to its own
module (or Darcs.Util.Path) if desired. And yes, I do think other
commands may be able to profit from re-using it, but my experience so
far suggests that you don't really know for sure until you actually go
head and do the re-factoring.
msg19726 (view) Author: gh Date: 2017-10-10.17:03:17
Agreed. Accepted.
Date User Action Args
2016-05-06 20:26:31bfcreate
2017-01-11 05:31:12ganeshsetstatus: needs-screening -> followup-in-progress
2017-08-11 19:04:31bfsetfiles: + resolve-issue2482_-mark_conflicts-file.dpatch
messages: + msg19552
2017-08-11 19:07:23bfsetmessages: + msg19553
2017-08-11 19:16:44bfsetstatus: followup-in-progress -> needs-review
2017-08-11 19:17:21bfsettitle: WIP: resolve issue2482 darcs mark-conflicts FILE -> resolve issue2482: mark-conflicts FILE
2017-08-20 14:38:08ghsetmessages: + msg19611
2017-08-21 09:00:10bfsetmessages: + msg19622
2017-10-10 17:03:17ghsetstatus: needs-review -> accepted
messages: + msg19726