After searching the googles for couple hours I found an answer to my question. I know this post Undo checkout TFS answers my question, however it doesn't answer all the questions I have. I want to achieve the same objective that the post asked about. How to only revert files that have been checked out if nothing was modified in that file? The answer to my question shouldn't be too hard to answer.
So what I'm doing is copying files from a server and overwriting them in my local workspace. I am checking out all the files being copied. However, if a file that was copied is not modified in anyway(server file and destination file are exact same), I'd like to undo the checkout of that file.
I know I'm to use the workspace.Undo() method and the gentleman said it worked for him. However he didn't show how he implemented it.
Here is the code I have with help from the link:
public static void CheckOutFromTFS(string filepath)
{
var workspaceInfo = Workstation.Current.GetLocalWorkspaceInfo(filepath);
if (workspaceInfo == null)
{
return;
}
var server = new TfsTeamProjectCollection(workspaceInfo.ServerUri);
var workspace = workspaceInfo.GetWorkspace(server);
workspace.PendEdit(filepath);
}
The answer given was to use the workspace.Undo() method. Do I add this method as the last line in CheckOutFromTFS() like so?
public static void CheckOutFromTFS(string filepath)
{
var workspaceInfo = Workstation.Current.GetLocalWorkspaceInfo(filepath);
if (workspaceInfo == null)
{
return;
}
var server = new TfsTeamProjectCollection(workspaceInfo.ServerUri);
var workspace = workspaceInfo.GetWorkspace(server);
workspace.PendEdit(filepath);
workspace.Undo();
}
Or is it done differently? I'm not sure if this Undo() will only revert files if there are no changes or just revert the checkout entirely and render the PendEdit() useless. Can someone help clarify this for me?
If you use a local workspace then all file that have no changes will automatically revert to not checked-out. You don't need to do anything at all. This works with VS 2012 or better with TFS 2012 or better. You'll need to convert you workspace to a local workspace first like this
So I found the answer to my question in various posts. I kinda took bits an pieces and combined them together to get my working solution. The use of the Undo() function with passing in the filepath actually does uncheckout the file regardless if it was modified or not. My workspace was also local but VS and TFS couldn't automatically revert those unmodified files for me so I took the below approach.
So what I decided to do was to just use the Team Foundation Power Tools "uu" command to undo the changes to unchanged files in the workspace. I created a batch file and entered the following command: echo y | tfpt uu . /noget /recursive. Since we will not show the shell during execution, I used the "echo y" command to automatically answer the question, "Do you wish to undo these redundant pending changes? (Y/N)". Including /noget is highly recommended since it prevents a forced 'get latest' of all your project's files which depending on the total number can take a extremely long time.
var startInfo = new System.Diagnostics.ProcessStartInfo
{
WorkingDirectory = projectRoot,
FileName = projectRoot + #"\undoUnchanged.bat",
UseShellExecute = false,
CreateNoWindow = true
};
Process process = Process.Start(startInfo);
process.WaitForExit();
process.Close();
After the script runs and the process.Close() executes you and double check if your unmodified files actually were unchecked out by hitting the refresh button on the Team Explorer window in your project. Hope someone else can find some use in this.
If I understand the question well and you actually need undo through C# code behind, I believe this shoul help you:
Undo checkout TFS
Related
I'm trying to rename a branch programaticly using "PendRename" method from Microsoft.TeamFoundation.VersionControl.Client.
Moves working, so, if I use this to move everything to a new location, it works, but what I need is to rename a branch.
It is possible to do this by commandline "tfs.exe rename " (even this is given me errors if I have more than one workspace mapped for the same server url. the ... could not be found in your workspace, or you do not have permission to access it.)
So, could you please help to understand why rename a branch is not working?
Thank you,
To Rename or move files and folders in TFVC, you must be one of the Contributors for your team project. See Team Foundation Server default groups, permissions, and roles.
I tested that and renamed the branch correctly with the "PendRename" method.
Below code sample for your reference:
using System;
using Microsoft.TeamFoundation.Client;
using Microsoft.TeamFoundation.VersionControl.Client;
namespace RenameBranch
{
class Program
{
static void Main(string[] args)
{
string oldPath = #"E:\andy\0418Scrum\web0418-0823";
string newPath = #"E:\andy\0418Scrum\web0418-1020";
string collection = #"http://server:8080/tfs/DefaultCollection";
var tfsServer = new Uri(collection);
var tpc = new TfsTeamProjectCollection(tfsServer);
var vcs = tpc.GetService<VersionControlServer>();
Workspace workspace = vcs.GetWorkspace("YourWorkspaceName", vcs.AuthorizedUser);
workspace.PendRename(oldPath, newPath);
}
}
}
Then you need CheckIn it of course. Use a "workspace.GetPendingChanges()" and "workspace.CheckIn()" methods to do it.
Thank you,
The problem was, there are different processes that are generating workspaces to the directory, and then, he can't remove it.
So, I solve this issue doing this steps:
1. reserve a workspace local location for my app´
2. see if this workspace is mapped. If yes, I remove it
3. create a workspace to the previous folder in path and create the map
4. get latest
5. rename
Thank you Andy-MSFT for your support.
So we have created an updated version of a WSP for SharePoint 2010 due to our migration/update from 2007 to 2010.
The WSP is a event handler/reciever for ItemAdded() and we have it working as intended. Issue is that the operation seems to only work for one computer/machine and no others.
When the Item is Added to a list the WSP creates a Folder in Shared Documents library, creates a wiki page, then updates the new List Item with links to the Shared Doc and Wiki.
When triggered by Machine #1 and User #1 all operations work, when Machine #2(M2) and user #2(U2) or M3 and U3 non of the tasks take place when a new Item is created.
User #2 can log in on M1 and create a new item and all operations work. But if U1 uses M2 or M3 to create an item the events don't trigger. Machine #1 is able to trigger the event as many times as they want but no other computer is able to.
If you were able to follow is it something with the code or some sort of cache setting on the local machine or the SP server, or something else? Any help is appreciated.
Update: All machines are on the same network. Non of the machines are the server but various personal laptops. Development was done on a separate machine. All are accessing via the same URL. All users have same access. This is on our test site currently which would be switched to being production once migration/upgrade takes place.
Before current .WSP deployment we noticed the same issue but it was reverse, Machine #2 did all the updates but Machine #1 and #3 couldn't. Only thing we can think of was that those machines were the first to trigger the event after deployment.
I'm Not doing the .WSP install but our IT guy is(won't let us have access :/ but I understand) but below is the install commands he is running.
Add-SPSolution -LiteralPath "OurPath/ourFile.wsp"
Install-SPSolution -Identity ourIdentity -WebApplication http://myhost.com/ -GACDeployment
Below is the main part of the code
public class CreateWikiAndFolder : Microsoft.SharePoint.SPItemEventReceiver
{
public override void ItemAdded(SPItemEventProperties properties)
{
try
{
//this.DisableEventFiring();
base.EventFiringEnabled = false;
string sUrlOfWikiPage = string.Empty;
string sUrlOfNewFolder = string.Empty;
string sSubsiteRUL = string.Empty;
string sCurrentItemTitle = properties.ListItem["Title"].ToString();
string sWikiListName = "TR Wikis";
string sDocLibName = "Shared Documents";
string sTRListID = "TR Status";
if (sTRListID.ToUpper().Equals(properties.ListTitle.ToString().ToUpper()))
{
//Create the Folder
sUrlOfNewFolder = CreateFolder(properties.ListItem.Web, sDocLibName, sCurrentItemTitle);
//Create the Wiki
string ItemDispFormUrl = String.Concat(properties.ListItem.Web.Url, "/", properties.ListItem.ParentList.Forms[PAGETYPE.PAGE_DISPLAYFORM].Url, "?ID=", properties.ListItem.ID.ToString());
sUrlOfWikiPage = CreateWiki(properties.ListItem.Web, sWikiListName, sCurrentItemTitle, ItemDispFormUrl, sUrlOfNewFolder);
//Update the current TR Item
SPWeb myWeb = properties.ListItem.Web;
myWeb.AllowUnsafeUpdates = true;
SPListItem myListItem = properties.ListItem;
SPFieldUrlValue shareFolderURLValue = new SPFieldUrlValue();
shareFolderURLValue.Description = "Shared Folder";
shareFolderURLValue.Url = sUrlOfNewFolder ;
myListItem["SharedFolder"] = shareFolderURLValue;
myListItem.Update();
myWeb.AllowUnsafeUpdates = false;
}
base.EventFiringEnabled = true;
}
catch (Exception e)
{
//Currently throwing nothing
}
}
}
It could be a hardcoded path/url, however there is not enough information to identify the problem, I would be glad to update my answer with a more detailed theory if you provide more details or if you share some of your code.
Figured out the issue. I didn't include them with the above file code. But we were StreamWriting to a text file on the server to help us with debugging. Issue was with that, When user 1 was logged on their machine and the log files didn't exist, they would get generated. Now no other users then had read/write access to those files and so it errored out at our debug files for anyone else. But that Windows user could run it as much as they wanted as they were the owner of the file :/
I try to copy the EventLog "Application" to another folder on my PC but always get the error that
"....Could not find a part of the path
'C:\Windows\System32\winevt\Logs\Application.evtx..."
I use the code:
public void collectEventLogsFromSystem(string RequestedlogName,string newFolderPath)
{
string combinedLogToFind = "HKEY_LOCAL_MACHINE\\SYSTEM\\CurrentControlSet\\Services\\Eventlog\\" + RequestedlogName;
string LogEventsPath = (string)Registry.GetValue(combinedLogToFind, "File", null);
if (LogEventsPath != null)
{
System.IO.File.Copy(LogEventsPath, newFolderPath +"\\"+ RequestedlogName, true);
}
}//collectEventLogsFromSystem method
Even if i use explicit folder path it won't work:
System.IO.File.Copy(#"C:\Windows\System32\winevt\Logs\Application.evtx", "c:\\ttt\\Application.evtx", true);
any idea?
I found this answer on StackOverflow which will probably solve your problem. I have a 64 bit machine which exhibited the same behavior. This post by John Rasch solved the issue and explains why it failed.
If you are trying to get the log file while the program is running, the way above will not work. This post Export Event Log (.evtx) without "run as administrator" will allow you to backup the event log, even if you are currently using the event log in your application.
We have an application that uses the Windows 7 performance counters to track the total CPU usage. Every so often the registry at HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Perflib\009 will become corrupted and just be empty.
Manually following the steps found here http://support.microsoft.com/kb/300956 works perfectly. But when I make a C# program to programatically replace the 2 files mentioned no errors, exceptions, etc. But the files are not saved to the proper directory.
Here's what I've been using to test:
static void Main(string[] args)
{
string fileToReadPath1 = #"perfc009.dat";
string fileToReadPath2 = #"perfh009.dat";
FileInfo fileToRead1 = new FileInfo(fileToReadPath1);
FileInfo fileToRead2 = new FileInfo(fileToReadPath2);
FileInfo fileToReplaceInfo1 = new FileInfo(#"C:\Windows\System32\perfc009.dat");
FileInfo fileToReplaceInfo2 = new FileInfo(#"C:\Windows\System32\perfh009.dat");
File.Copy(fileToRead1.FullName, fileToReplaceInfo1.FullName, true);
File.Copy(fileToRead2.FullName, fileToReplaceInfo2.FullName, true);
}
I do make sure to run it with Administrative privileges.
Anyone know why the program would seem to run fine, but not copy the files to that directory? Is there some Windows security thing stopping this?
Thanks to Mike Z and the link shared in the comments above. Turns out because I was running in a 32-bit process, the OS did not allow it to edit those files. Compiling to 64-bit did work.
Reposting the link:
http://msdn.microsoft.com/en-us/library/windows/desktop/aa384187(v=vs.85).aspx
This one seems trivial but the answer has eluded me for a few days now.
I have a Windows batch file, that calls a C# program to do an extra verification that cannot be done in a batch file. After the verification is complete I need to return a status and a string back to the calling shell.
Now the return value is trivial and my C# console app simply sets a return value (exit code if you will). And I thought the string will also be a piece of cake. I attempted to define a new shell variable using the:
Environment.SetEnvironmentVariable("ERR", "Some text");
This call should (and does) define a shell variable within the current process - that is the very C# process that created the variable. The value is lost as soon as the C# app terminates and the shell that created the C# app knows nothing about the variable. So... A call with no particular use... At all... Unless perhaps if I created a child process from the C3 app, perhaps it would inherit my variables.
The EnvironmentVariableTarget.Machine and EnvironmentVariableTarget.User targets for the SetEnvironmentVariable call don't solve the problem either, as only a newly created process will get these new values from the registry.
So the only working solution I can think of is:
write to stdout
write to a file
encode extra meaning into the return value
The first two are a bit ugly and the last one has its limitations and problems.
Any other ideas (how to set a shell variable in the parent process)? Maybe such shell variable modifications are a security concern (think PATH)...
Thank-you for your time.
I had the same problem as Ryan and the only thing that came to my mind as a work-around was to write a batch in error out to set the variable and to call it from the batch.
ConsoleApplication1.exe:
'put some sensible code here
'put result in variable myResult
Dim myResult As String = Guid.NewGuid().ToString("D").ToUpperInvariant()
Console.WriteLine("Normal output from the consonle app")
Console.Error.WriteLine("#ECHO OFF")
Console.Error.WriteLine("SET zzzResult={0}", myResult)
Test.cmd (the calling batch):
#ECHO OFF
:Jump to folder of batch file
PUSHD %~d0%~p0
:Define a temp file
SET zzzTempFile=%TEMP%\TMP%Random%.CMD
:Call .NET console app
ConsoleApplication1.exe 2>%zzzTempFile%
:Call the generated batch file
CALL %zzzTempFile%
:Clean up temp file
DEL %zzzTempFile%
:Clean up variable
SET zzzTempFile=
:Do something with the result
ECHO Yeah, we finally got it!
ECHO:
ECHO The value is "%zzzResult%".
ECHO:
:Clean up result variable
SET zzzResult=
:Go back to original folder
POPD
That should do the trick. And yes, I do know this is an old post and Ryan is solving other issues by now, but there might be still somebody else out there having the same problem...
What you are asking is to be able to arbitrarily write to the memory space of a running process. For good reason, this is not possible without SeDebugPrivilege.
Any of the three solutions you list will work. Stdout is the standard way to communicate with a batch script.
By the way, you're writing a Windows batch file. I'm pretty sure the ship has already sailed on "a bit ugly".
If you want to put a value of some output into a variable in the batch you can use the following construct:
FOR /F "usebackq tokens=4 delims=\[\] " %i IN (`ver`) DO set VERSION=%i
ECHO %VERSION%
Output on my OS:
6.1.7601
'usebackq' means we are using back quotes which gives the ability to use a fileset in the command quoted with double quotes. You may not need this. 'tokens' means the index in the resulting string array to select (it can be a range M-N). If you need to skip lines use 'skip=X'). 'delims' are the string separators to use (like string-Split() in .Net).
You will put your console app instead of 'ver' and adapt the delimiters and tokens to match your specific output. If you have more variables to fill you will need to make the if a bit more complex but that should make a good start.
My BAT is a bit rusty, but I think it's possible to retrieve the 'exit' code from processes you've run externally, perhaps via %ERRORLEVEL%. If that's the case, make sure to exit your program via
Environment.Exit(123); // where 123 = error code
You can't add any messages, so you'll have to do that in the .bat file.
If this isn't the case, stdout is probably the best way.
After stumbling on this myself as well recently, I came up with this approach. What I did is run the bat file using the Process class, i.e.
// Spawn your process as you normally would... but also have it dump the environment varaibles
Process process = new Process();
process.StartInfo.FileName = mybatfile.bat;
process.StartInfo.Arguments = #"&&set>>envirodump.txt";
process.StartInfo.UseShellExecute = false;
process.StartInfo.RedirectStandardOutput = true;
process.StartInfo.RedirectStandardError = false;
process.Start();
string output = process.StandardOutput.ReadToEnd();
process.WaitForExit();
// Read the environment variable lines into a string array
string[] envirolines = File.ReadAllLines("envirodump.txt");
File.Delete("envirodump.txt");
// Now simply set the environment variables in the parent process
foreach(string line in a)
{
string var = line.Split('=')[0];
string val = line.Split('=')[1];
Environment.SetEnvironmentVariable(var, val);
}
This seems to have worked for me. It's not the cleanest approach, but will work in a bind. :)