Suppose you have commit some changes, and then somehow deleted/modified some files by accident.
Which would be better? Using revert oder update?
There are two big differences between running hg update -C and doing hg revert -a
Update will move your parent up to the tip of the head of the branch
Update will no create any backup files
The revert command on the other hand
Creates backups of all reverted files (unless you give --no-backup command)
Does not change your working directories parent changeset.
Now which is better? Depends in which of the things listed above you want.
In your case you want revert -- it alters your working directory without altering the output of the hg parents command. Your parent revision is the "currently checked out revision" and will become the "parent" of your next commit. You don't need to alter that pointer, so just revert.
Related
I've searched the docs of Mercurial and still am confused. What I'm wanting to do is just reinstate the last commit I made i.e. I want my project to go back to being exactly the same as it was when I made the last commit. I see hg revert, rollback, etc. and still am not understanding which is correct for this situation. Which should I use?
The hg rollback command is used to undo the last action that modified Mercurial's internal store, usually a pull or commit. So, if you want to undo your last commit hg rollback will work.
But it sounds like you want to undo all your uncommitted changes. You have two options. The hg revert --all command will undo all uncommitted changes. Each changed file is saved/backed-up with a .orig extension before being reverted.
If you don't need to preserve your changes in .orig files, run hg update -C. This clears out all uncomitted changes, without preserving anything.
If you have commited changesets and wish to remove them, I like the strip extension, strip extension. With it, you can remove explicit changesets from your history.
However, strip is an unforgiving command, i.e. if you get it wrong there is no retrieval unless you have a backup of the repo. You might prefer the prune command which comes with the evolve extension. Using prune, you can mark changesets as obsolete and they will no longer normally be visible in logs or tortoise. [You can make them visible by adding --hidden on an hg log command line, or in tortoise by enabling the Filter toolbar (from the view menu) and selecting 'Show/Hide hidden changesets'.]
Say I type hg add in Mercurial, and there a bunch of untracked files in my working directory that are not ignored. What is the easiest way to un-add all those files without explicitly typing the name of each file?
Can I just un-add them all with one command?
Preface
You must always ask questions, which contain as much information as possible. Because now your question, depending from some conditions, may have totally different answers.
Case One - no local modifications in already versioned files, only added (and not committed) files
hg revert will return your working directory to the state after the last commit, undoing all changes it it.
Case One - local edits, which you want to save and occasionally added files
Read about filesets in Mercurial.
Use fileset in the hg forget command, something like hg forget "set:added()".
Use hg revert or hg forget on the files (both do the same for a file you ran hg add on). To avoid typing out the filenames, you can use a fileset like this:
$ hg revert "set:added()"
This will revert the file back to how it looked in the working copy parent revision, i.e., it will become unknown again.
hg revert -r .^ path-to-file will revert the commit from the commit-set.
then commit and submit (if using jelly fish) and you'll see the files removed from the changeset. I don't know why .^ works yet, but somebody will probably know.
You could always just re-clone your repository and then replace (delete existing and then copy new) the .hg directory in your working folder with the one from the fresh clone... (assuming you have no pending commits..)
what would be an equivalent mercurial command (or workflow) for
git reset --mixed HEAD^
or
git reset --soft HEAD^
i.e. I want leave the working tree intact but get the repository back into the state it was before the last commit. Surprisingly I did not find anything useful on stackoverflow or with google.
Note that I cannot use
hg rollback
as I've done some history rewriting using HistEdit after the last commit.
Added to clarify:
After some rebasing and history editing I had ended up with A<--B<--C. Then I used HistEdit to squash B and C together, obtaining A<--C'. Now I want to split up the commit C' (I committed the wrong files in B). I figured the easiest way to do this was to get the repository back to state A (which technically never existed in the repository because of all the rebasing and history editing before hand) and the working tree to the state of C' and then doing two commits.
The right way to replicate git reset --soft HEAD^ (undo the current commit but keep changes in the working copy) is:
hg strip --keep -r .
-1 will only work if the commit you want to strip is the very last commit that entered the repository. . refers to the currently checked out commit, which is the closest equivalent Mercurial has to Git's HEAD.
Note that if . has descendants, those will get stripped away too. If you'd like to keep the commit around, then once you have the commit ID, you can instead:
hg update .^
hg revert --all -r <commit id>
This will update to the commit's parent and then replace the files in the working copy with the versions at that commit.
I believe the more modern and simpler way to do this now is hg uncommit. Note this leaves behind an empty commit which can be useful if you want to reuse the commit message later. If you don't, use hg uncommit --no-keep to not leave the empty commit.
hg uncommit [OPTION]... [FILE]...
uncommit part or all of a local changeset
This command undoes the effect of a local commit, returning the affected
files to their uncommitted state. This means that files modified or
deleted in the changeset will be left unchanged, and so will remain
modified in the working directory.
If no files are specified, the commit will be left empty, unless --no-keep
I renamed a few directories in my mercurial repository [not by hg rename command i was not aware of that at that time] just by doing a mv on my linux box , and then I realized that I want the things the way they were before so I did a hg rollback. But it didnt' restore the things to the state they were , how can i get my last commit?
The hg rollback did remove your the latest commit¹. You need to move the directories back with mv, and redo you last commit.
When you encounter the same situation again, you can use hg revert to replace the changes in the working copy by the content of the current revision. But since your current revision is the revision before your last commit, using revert now would also undo the changes of your last commit.
As a rule of thumb never use rollback², since you get the effect of rollback also with the much more safe mq extension.
¹Technically it removed the latest transaction, but it is very likely that your last transaction was a commit.
You want revert. Try
hg revert -a
or specify a directory if you want to revert just that directory.
Suppose that:
I have a repo called MyRepo.
I have uncommitted changes in my working directory.
I do a pull from Repo1 and that creates a branch in MyRepo
I want to do a merge of what I already had in my repo with what I have just pulled.
As described here, the merge process changes the state of the working directory.
In my scenario, I don't want to loose the uncommitted changes that are in my working directory because at that point, I'm interested in changing the state of MyRepo; not the state of my working directory.
Is there a way to go through the merging process, including resolving conflicts manually, without affecting the content my working directory? Can this process be done in temporary files only? Or should I shelve my changes, do the merge and then unshelve to restore my working dir to the state it was before the merge?
Use shelve or attic extensions to temporarily stash your changes while you merge.
You could clone your repository to your latest checkin, merge the changes with the clone, commit, and then pull the new changeset from your cloned repository back into your current one.
You can't do that. As the documentation says, merge really is a working tree operation. Resolving conflicts without testing the result is crazy. Either shelve or commit the changes and then do the merge. If you don't want the commit to be visible anywhere, you can commit the change, update to the previous revision, merge the new branch, apply the temporarily committed patch and strip that temporary branch.
Just do it in a clone.
Clone MyRepo to MyRepo-merger, which will pull over all the committed changes, but not your uncommitted changes
Inside MyRepo-merger do a pull from Repo1.
Inside MyRepo-merger do all the hg merges you want
Inside MyRepo-merger push to either Repo1 or the original MyRepo
Pushing back to MyRepo won't alter the working dir of MyRepo, so that's a safe action.
Remember in Mercurial that clones are incredibly fast and cheap -- on modern file systems they use hardlinks under the covers so they're not even taking up much space on disk.
I'm pretty new to mercurial, but couldn't you clone from your local repository and copy your working copy over to the clone? You could then run your merge in the original? You'd preserve the state of your working copy in the clone while being free to allow the change of the original's working copy.
Test this first. I've never done it in hg.