Is there any way to take complete manual control over the merge process in Mercurial?
I want to be able to choose direction of file-merge, even for files without conflicts.
Is that possible?
Turn "pre-merge" off in your merge configuration. Then everything counts as a conflict and you can pick "left" or "right" for each and every file change.
[merge-tools]
mymergetool.premerge = False
from MergeToolConfiguration on the Mercurial wiki.
Edit your configuration file this way:
[ui]
merge = kdiff3
[merge-tools]
kdiff3.premerge = false
kdiff3.args=--L1 base --L2 local --L3 other $base $local $other -o $output
By default it puts the --auto argument on kdiff3 so kdiff3 auto merges.
A merge is always performed between the working directory's parent revision and another revision, by default the other head in your repository.
If you want to merge in the other "direction" you can change which branch is in your working directory by checking out a specific revision:
hg update -r [rev]
To see which heads you have in your repository run the following command:
hg heads
Alternatively, if you're using fetch you can use the --switch-parent option to merge in the other direction:
hg fetch --switch-parent
You can't change the direction of the merge on a file-by-file basis as Mercurial works with changesets which affect a whole repository not by tracking changes on to individual files like CVS.
In case of KDiff3 it is vital to add option --qall (see http://kdiff3.sourceforge.net/doc/documentation.html). There will be an automatic merge of some conflicts without this key (like "Automatically Solve Simple Conflicts" from "Merge" menu). Thus the more proper command line is:
[ui]
merge = kdiff3
[merge-tools]
kdiff3.premerge = False
kdiff3.args=$base $local $other -o $output --L1 base --L2 local --L3 other --qall
Related
Whenever I perform an hg rebase and there are merge conflicts, it immediately pulls up an editor for me to resolve the conflict. However, it doesn't give me any information about where in the rebase process I am at. For example, if my history looks as follows:
o 12
|
o 11
|
10 o |
\ /
o 9
Performing hg rebase -s 11 -d 10 may have a conflict trying to apply either 11 or 12. It is difficult to tell at a glance just from the merge conflict where I am stopped, especially when the graph is larger than this. How can I tell where in the rebase process the conflict is?
Very recent Mercurials have two configuration options: [ui] mergemarkertemplate, and [ui] pre-merge-tool-output-template, which can be used to improve this situation a bit.
pre-merge-tool-output-template
pre-merge-tool-output-template is printed before running any external merge tool. This can be used to print something before your editor or kdiff3 pops up; note that if you use a terminal-based merge tool (such as most editors unless they're the gui version), it'll likely be hidden by the merge tool. Depending on OS and what program you're using, you may be able to hit Ctrl-Z to suspend your merge tool to see this output.
Example output:
merging path/to/file
Running merge tool for path/to/file (/usr/bin/vim):
- local (working copy): 10:2d1f533d add binary file (#2) tip default
- base (base): 6:abcd1234 some other description default
- other (merge rev): 9:1e7ad7d7 add binary file (#1) default
... vim runs here ...
See https://www.mercurial-scm.org/repo/hg/file/14589f1989e9/tests/test-merge-tools.t#l1956 for the template that produced that output, hg help config.ui.pre-merge-tool-output-template and hg help templates for more information on that.
mergemarkertemplate
mergemarkertemplate controls the conflict markers you see in your editor. Set [ui] mergemarkers=detailed and see if this is sufficient; if not, you can use [ui] mergemarkertemplate to customize it; this can also be customized on a per-merge-tool basis, so see hg help config.ui.mergemarkers, hg help config.ui.mergemarkertemplate, and hg help config.merge-tools.
Programs with customizable labels
Merge tools like kdiff3 often have the ability to customizable the labels. In the default configuration, this should be the operation-provided name for the base/local/other (in my example above, this would be base, working copy, and merge rev, respectively. I believe if you have [ui] mergemarkers=detailed or [merge-tools] kdiff3.mergemarkers=detailed, these will include additional information. See hg help config.merge-tools for more information on the per-merge-tool configuration options.
(Not an answer, exactly, but a bit long for a comment...)
When you write:
Performing hg rebase -s 11 -d 10 may have a conflict trying to apply
either 10 or 11.
do you mean to write
may have a conflict trying to apply either 11 or 12
? Because you are trying to rebase those csets to 10, so it doesn't make sense to talk about applying 10. Also, consider using the Evolve extension if you aren't already. It makes everything append-only, which is much better.
Also, test in a clone. And also, try rebasing 12 first, if possible. Anyway, Mercurial is just trying to rebase the changes from both 11 and 12, and I don't think it distinguishes between those changes. And why would you expect it to? Isn't it obvious to you which changes belong to which cset?
Also, consider configuring your merge setup for use with kdiff3, if you aren't already. It makes things much clearer to do things in a merge editor, and you can also see both sides of the merge clearly. See
https://www.mercurial-scm.org/wiki/MergeToolConfiguration and https://www.mercurial-scm.org/wiki/KDiff3
Personally, I have the following lines in my ~/.hgrc, but they've been there a long time, and I don't remember where I got them from. Also, I don't do merges much these days. But for whatever it is worth...
[merge-tools]
kdiff3.args=--auto --L1 base --L2 local --L3 other $base $local $other -o $output
kdiff3.regkey=Software\KDiff3
kdiff3.regappend=\kdiff3.exe
kdiff3.fixeol=True
kdiff3.gui=True
kdiff3.diffargs=--L1 '$plabel1' --L2 '$clabel' $parent $child
Hope that helps.
Consider the following revisions of a file:
branch -1---2---
/
/
main 100--101-102--103---104---
I am currently in my branch and the file is synced to revision 2. I know that p4 has kept the history of the file, because I used merge and not copy, and indeed the history shows revisions previous to 103 (inclusive) when I look at the history of the file and check the option "Follow branch actions".
I would like to roll it back to revision 103 or before. Is this possible?
Answers for p4v preferred, but I'll take command line statements if that's the only way to do it.
The copy command is the best way of doing this:
p4 copy main#103 branch
Another option would be to use integ -f:
p4 integ -f main#103 branch
p4 resolve -at
One benefit of copy is that if you're doing this across a bunch of files (e.g. main/...#2018/01/01 to get all of main as of Jan 1) and some of them are already identical to the rev you're copying from because they haven't changed in that time, copy will leave them alone, which makes the result less noisy.
I find that (unlike some other implementations), Mercurial is over zealous on what it pushes - and a pushed commit is a non-editable commit.
I have trained myself to use hg push -b branchawesome or hg push -r awesomer. However, sometimes my fingers do an "oops" and everything - including transient Draft work - is pushed to the upstream repository.
Is there a Mercurial-way to either prevent hg push outright or require a "force" flag, as in hg push --draft?
I'd prefer not to use Secret Phases. Despite how I've worded the question, the goal is to help/encourage others to be specific about what changes are pushed - without incurring an entirely new concept.
First, a note: Mercurial will not by default let you push a new head/branch without explicitly saying so (you'll get an error message telling you that you need --new-branch or -f to actually push your changes). Accidentally pushing more than you intended can only happen if you have commits on multiple existing branches.
Second, the easy (but problematic) solution is to use an alias for push, e.g. push = push -r . will be more restrictive. However, this overrides the push command, making it difficult to get at the original version (which you can do via rawpush = !$HG --config alias.push=push "$#", for example).
The cleanest way is to use an extension that wraps the push command and provides a different default. e.g.:
from mercurial import extensions, commands
testedwith = "3.5"
default_push_rev = "."
# alternative choices
# Push the current revision, but only if it is a head
# default_push_rev = ". and head()"
# Push the current revisions and all revisions depending on it
# default_push_rev = "descendants(.)"
# Push the most recent head that follows the current revision
# default_push_rev = "last(descendants(.))"
# Push the tip revision (i.e. the chronologically most recent commit).
# default_push_rev = "tip"
# Push only public change sets
# default_push_rev = "public()"
def override_push(original_cmd, ui, repo, *pats, **opts):
have_rev = False
for opt in ["rev", "branch"]:
if opts.has_key(opt) and opts[opt]:
have_rev = True
if not have_rev:
opts["rev"] = [default_push_rev]
return original_cmd(ui, repo, *pats, **opts)
def uisetup(ui):
extensions.wrapcommand(commands.table, "push", override_push)
Unlike the alias, this will only change the default revisons to be pushed if none is explicitly provided. Note that there are multiple choices that may make sense. The above code is setup to push . by default, but there are alternatives that you may like better.
Note also that the extension does not override the hg outgoing command, but if you wish, that can be easily done by duplicating the wrapcommand line and substituting "outgoing" for "push" in the duplicate line.
To get the original behavior, just use hg push -r 'all()', possibly as an alias:
[alias]
push-all = push -r 'all()'
EDIT: Fixed a bug in the original code which ignored the branch option.
When I want to compute a list of changes in one branch but not an older branch, I do something like this:
diff -u \
<( p4 changes -s submitted -i #OLD_BRANCH_LABEL ) \
<( p4 changes -s submitted -i #NEW_BRANCH_LABEL ) | grep ^+'
NOTE: These are automatic labels containing a view and a revision, but I can also put in the view paths #REVISION
which gives a good answer when both labels are actually in the same view, and good answers generally.
But I find that I get some poor results in this case:
If a commit #A adds a file to MAIN_BRANCH, and commit #D removes the file (and makes some other change), followed by commit #F which forks a NEW_BRANCH, I find that the MAIN_BRANCH contains both commits #A and #D, but NEW_BRANCH contains commits #D and #F
The NEW_BRANCH does not contain commit #A
So my p4 changes recipe above insists that the mainline contains #A which is not in the branch, or in other words was made since the branch, even though it is not in the mainline any more than it is in the branch.
An obvious but unwieldy fix would be to take another fork of MAIN_BRANCH at the point I want to compare, knowing that #A would be excluded again in the same way.
If this were git, I would use git-merge-base or something to find a common point and remove that from the changes of both branches, but perforce branching is too flexible, being just another integration.
Are there any ways that I can convince perforce that NEW_BRANCH does contain #A? (for the branch fork #F occurred after #A was committed).
Or to get perforce to ignore changes whose files are entirely deleted, that WOULD be ignored if a branch were made?
The command p4 interchanges does exactly what you are after, if I read your question correctly.
'p4 interchanges' lists changes that have not been integrated from
a set of source files to a set of target files.
Have a look at p4 help interchanges for the full description. The command does take indirect integrations into account, too.
It is a very simple and stupid question,
I am working on 2 tasks and modified 2 sets of files in the code.
Now when I type 'hg ci', it checks in all the files. Can I remove certain files from the checkin i.e. do checking for only one task?
If I remove the files in the 'check in message' will they be removed from the checkin
Thanks for all the answers
This seems like a big flaw, My use case is very simple and general. Most of the time dev are working on various tasks , some are ready , some are not, bug fixes etc.
Now the only simple solution seems to be multiple local repos or clones
Use hg ci <files>... to commit only certain files in your working directory.
If you want to pick file by file you can use the record command. It ships with mercurial, but you have to turn it on if you want to use it by putting: record= in the [extensions] section of your ~/.hgrc.
It goes hunk by hunk, but you can answer for a whole file:
y - record this change
n - skip this change
s - skip remaining changes to this file
f - record remaining changes to this file
d - done, skip remaining changes and files
a - record all changes to all remaining files
q - quit, recording no changes
? - display help
I'll point out that if you're committing some files but not others it's certain that you've not run your test suite on the one change without the other, but maybe that doesn't apply in your case.
This isn't possible with mercurial out of the box. As have been suggested there are several ways of selecting what files you want to commit. To get this function via the commit message template, you would need an extension or a shell script wrapping the commit command. Here's one way to do that:
[alias]
ci = ! hg-partial-commit
hg-partial-commit:
#!/bin/sh
# partial commit
edit=$(mktemp ${TMPDIR:-/tmp}/$(basename $0).XXXXXXXXXXXX)
filelist=$(mktemp ${TMPDIR:-/tmp}/$(basename $0).XXXXXXXXXXXX)
logmessage=$(mktemp ${TMPDIR:-/tmp}/$(basename $0).XXXXXXXXXXXX)
cleanup="rm -f '$edit' '$filelist' '$logmessage'"
trap "$cleanup" 0 1 2 3 15
(
echo user: $(hg debugconfig ui.username)
echo branch: $(hg branch)
hg parents --template 'parent: {rev}:{node|short} {author} {date|isodate}\n'
echo
echo 'Enter commit message. Select files to commit by deleting lines:'
hg status 'set:not unknown()' | sed -e 's/^/#/'
) | sed -e 's/^/HG: /' >"$edit"
${VISUAL:-${EDITOR:-vi}} "$edit"
egrep -v '^HG:' "$edit" >"$logmessage"
egrep '^HG: #' "$edit" | cut -c8- >"$filelist"
hg commit -l "$logmessage" "listfile:$filelist"
$cleanup
The real problem here is the fact that you're doing changes related to different tasks jumbled together. Mercurial has a few ways you can keep things separate.
Task Branches
Suppose you've been working on a task and you've checked in a few times since you last pulled, but things aren't ready to share yet.
o----o----B----o----o----o
Here, B is the revision where you started your changes. What we do is (after making sure our current changes are checked in):
> hg update -r B
<do our work on the other task>
> hg commit
We've now created a new branch with the changes for this task separated from the changes for our original task.
o----o----B----o----o----o
\
----o
We can do this for as many different tasks as we want. The only problem is that sometimes remembering which branch is which can be awkward. This is where features like bookmarks come in useful. A bookmark is a tag which moves forward as commits are made so that it always points at the head of a branch.
Mercurial Queues
MQ adds the ability to work on changes incrementally and move between them by pushing and poping then off a stack (or "Queue" I guess). So if I had a set of uncommitted changes that I needed to split up I'd:
> hg qrecord taska
> hg qrecord taskb
> hg qrecord taskc
I'd use the record extension (or more likely the crecord extension) to select which parts of files I want to select.
If I needed to go back to taska and make some changes:
> hg qpop; hg qpop # pop two off the queue to go back to task a
<Do changes>
> hg qrefresh # update task a with the new changes
When I want to turn the queue into normal changesets:
> hg qpush or hg qpop # get the changes I want converted onto the queue
> hg qfinish -a # converts mq changes to normal changesets
There's other methods too, but that will do for now.
You will unavoidably have to either specify the files that you want to add or the files you want to leave out. If you have a lot of files, as you indicate above, the following steps will simplify the procedure (I'm assuming a UNIX-ish system here, Windows would be slightly different).
First, generate a list of changed files:
hg status -mard -n >/tmp/changedlist.txt
The -mard options will list all files that were modified, added, removed, or delated. The -n option will list them without the status prefix so that all you have is a raw list of files.
Second, edit /tmp/changedlist.txt with your favorite text editor so that it contains only the files you wish to commit.
Third, commit these files:
hg commit `cat /tmp/changedlist.txt`
You will be able to review the files to be committed before actually performing the commit.
Alternatively, you can add more files to a commit incrementally by using
`hg commit --amend file...`
Here, hg commit --amend will not create a new commit, but add the new files to the existing commit. So, you start out with committing just a couple of files, then incrementally adding more until you are done. This still requires you to type all of them in.
For yet another alternative, you can use Mercurial Queues to split a commit in more sophisticated ways, but that's a bit more of an advanced topic.