I have a project using Perforce. The scenario I am looking at is like below :
Multiple users submit their changes to Perforce repository and I would like to take a snapshot before syncing. I want to be able to revert if new release has some problems.
Basically I want to create a snapshot of current file which I can view with 'p4 have' and then revert to 'p4 have' files after syncing to head revision! How can I do that !
I tried to use 'p4 tag -l rel1 //depot/', but this command creates a label for head revision. I am looking for a way to do some command like this 'p4 label have' ! Is it possible ?
You need to specify that you want the label to include just the files currently in your workspace:
p4 tag -l rel1 //depot/...#have
Or better yet use this shortcut:
p4 label myws (define the label)
p4 labelsync (labels what you currently have in your workspace)
Then you can restore with:
p4 sync #myws
Related
We are using the following P4 command for snapshot versioning.
p4 changes -m1 /path/to/files/...#have
I noticed that the above command doesn't capture CLs of deleted files if the CL is the head CL.
So it ends up having two artifacts with the same version but different files.
I was wondering how we can cover this edge case?
This is an artifact of the fact that deleted files aren't synced to the client and will therefore never be included in #have.
If possible, the best way to address this IMO is to capture the changelist at the time you sync to the head revision. Do:
p4 changes -m1 -ssubmitted /path/to/files/...
and then use that changelist to do:
p4 sync /path/to/files/...#CHANGE
Since submitted changelists are immutable (outside of edge cases like obliterate and +S files) you now have a guarantee that anyone else syncing to that same path#change will get the same set of revisions that's in your workspace.
Note that the p4 changes command will capture changelists with deleted revisions, even if those revisions are then skipped by the p4 sync!
I am very new to Perforce and just started using it.
After syncing my code using $p4 sync command i started editing a few files.
$p4 edit file1
$p4 edit file2
$p4 edit file3
These files were getting added to my default changelist. For collaborating with my team i wanted to create a changelist. For creating a changelist i ran $p4 change and removed "file1" from the changelist description. Now when i run $p4 opened. It gives an output similar to this.
//depot/... /file1 edit default change (text)
//depot/... /file2 edit default 111 (text+k)
//depot/... /file3 edit default 111 (text+k)
Now i have the following questions in my mind:
Which changelist am i currently working on is it default or 111 ??
what is the meaning of (test + k)?
When i make changes to file2 and file3 it is getting synced to changelist 111 and when i make changes to file1 it gets synced to default changelist . I am highly confused how is this happening ?
Also one more thing i am confused about is if do $p4 shelve will my changes disappear and will be reapplied only when i run unshelve command for the changelist created ?? Is this similar to git stash and git stash apply ?
It will be great if someone can answer these questions in detail. Any tutorial suggestions for Perforce will also be of great help.
Thanks in advance.
My recommendation for a Perforce tutorial is the Perforce User's Guide. Here's the section on changelists:
https://www.perforce.com/perforce/r15.1/manuals/intro/chapter.working_in_perforce.html#working_in_perforce.working_with_files.changelists
Which changelist am i currently working on is it default or 111 ??
Both! Both of these are pending changelists in your workspace.
what is the meaning of (text+k)?
The thing in parentheses is the "filetype". +k is a "filetype modifier" meaning that keywords (special words like $Id$ and $Revision$ and $Author$ in this file will be automatically expanded to appropriate values when you submit.
When i make changes to file2 and file3 it is getting synced to changelist 111 and when i make changes to file1 it gets synced to default changelist . I am highly confused how is this happening ?
Nothing is getting "synced" anywhere yet -- the pending changelists are just containers that reference the different files. When you shelve or submit, then the files associated with those changelists will get sent to the server and will be accessible by other clients. Since shelve and submit are changelist-level operations, only the files in those changelists are affected -- that's the point of having different pending changelists. You have all of the pending files in your workspace, but you can split up which ones get sent to the server at which times (it's a little like pushing different branches in git, but not -- you can do this on every operation in Perforce even when you aren't branching because each file is versioned individually rather than having the entire tree versioned as one atomic blob).
Also one more thing i am confused about is if do $p4 shelve will my changes disappear and will be reapplied only when i run unshelve command for the changelist created ?? Is this similar to git stash and git stash apply ?
No, p4 shelve only syncs the shelved change on the server with the local files in your workspace -- it doesn't in itself change your workspace. The equivalent of "stashing" would be to p4 shelve and then to p4 revert to wipe out the workspace changes. shelve on its own is a little more like doing a git push to a branch -- you keep your local copy but now it's also on the server (but not part of the "master" history). (It's not exactly the same though -- to be honest if you're brand new to Perforce I'd stick to regular old "submit" in a regular old branch since that's the basic workflow. Sharing work via shelves requires a lot more manual work since each shelf is like its own little mini-branch with no versioning.)
I have deleted a file from a perforce branch and I am trying to recover it. I thought of syncing to its previous version and resubmitting it. Is it the better way to do and how?
P.S: Using command line
You've got it.
p4 sync file#version
p4 add file
p4 submit
Another option (only available with server versions after ~2013):
p4 copy file#version file
p4 submit
Another other option (only available with server versions after ~2016):
p4 undo file#head
p4 submit
Any of these ways you do it, the history will record that the new version came from the older one (you'll see "file#4 add from file#2" or something along those lines), and that will do things like connect up the history in Revision Graph and allow merge operations to find a common base.
I have made several checkins using perforce. I have no realized that all of them are unnecessary. I would like to revert all the changes for the last x revisions in the working directory, update the version number, and check in.
I am familiar with Mercurial. The way that I would it for that would be:
$ hg revert -r last_good_changeset .
$ edit version-number.txt
$ hg ci
Is there a way to do something similar in perforce?
In Perforce, a revert refers to restoring a file to the state it was in before it was checked out. What you're looking to do is back out a submitted changelist. This Perforce KB article has a few methods to do what you're trying to do, depending on your particular circumstance.
For example, if you have revisions #1 - #6 of a particular file, and you want to roll back to revision #3, you'd do this:
p4 sync myfilename#3
p4 edit myfilename
p4 sync myfilename
You're telling Perforce to get revision #3 from the depot, check it out for edit, then try to sync it back up to #head (the latest version in the depot). Since the file is checked out from an earlier revision, Perforce schedules a resolve so you need to tell it what you want to do with the file: accept the version in the depot, accept your local changes, or try to merge the two. You'll want to tell Perforce to accept the local version (or in Perforce parlance "yours"):
p4 resolve -ay myfilename
Now that it's resolved, you can submit it with:
p4 submit
If you have a series of files you want to do this with (for example, you've edited a bunch of files in a given directory and have checked them all in together several times, and you want to back out all of those), you can use changelist syntax as well. For example, if you want to roll everything back in a given directory to changelist 123, you can do this:
p4 sync //depot/some/path/*#123
p4 edit //depot/some/path/*
p4 sync //depot/some/path/*
p4 resolve -ay //depot/some/path/*
p4 submit
This will work for any revision modifier (see p4 help revisions for alternate methods of specifying the version you want).
The rollback function is specifically designed to do this. It goes back to a certain date/time or change list # and reverts all changes in the window you give it.
Simply right click on the file in question (P4V obviously) and select rollback. It will bring up this box. Not sure how to execute from command line...Ill see if I can figure it out and add that info.
Is there a way to duplicate the change list description using p4 command line when integrating from the source change list across branches ?
Details:
For e.g.
p4 integ -t -b branch_a_to_branch_b -s //my/tree/...#2425,#2425
integrates files from changelist 2425 to the default changelist in the target branch branch_b from source branch branch_a
Is it possible to create a new changelist which has the same description as the source changelist ?
Then the target integration can be done on this new changelist.
I don't know of a way to do this in one command, but it seems like it would be easy enough to script up. You might want to look at the commands p4v runs when it integrates, as it has the option to integrate to a new changelist. Then it would simply be a matter of changing the description of the new changelist to match the old one.