Docker container with Cross-platform Command-line Client for Team Foundation Server (Team Explorer Everywhere)

Licensing

• The Dockerfile is issued by the MIT license.
• The image contains (at build time it downloads) Microsoft/team-explorer-everywhere release 14.111.1
  Team Explorer Everywhere Plug-in for Eclipse from Microsoft/team-explorer-everywhere is issued by the MIT license.
• For using tf - Cross-platform Command-line Client for Team Foundation Server you MUST accept MICROSOFT SOFTWARE LICENSE TERMS - MICROSOFT TEAM EXPLORER EVERYWHERE license agreement.
  The image contais a tweak to have this EULA accepted. If you do not accept MICROSOFT TEAM EXPLORER EVERYWHERE license agreement then you MUST NOT use this image containers.

Using tf

Usage of the Cross-platform Command-line Client for TFS is described on the Internet, for eg.:

• at msdn.microsoft.com; commands
• at blogs.msdn.microsoft.com - Install the Cross-platform Command-line Client for Team Foundation Server
• Team Foundation Version Control and the Linux Command Line

There are quite a few differences when using the container. Important notes:

• If your collection (or URL itself) contains spaces, - enclose the URL in doble quotes. Do not use URL encoding like %20. The URL is encoded in the application after accepting parameters.
• If your password contains !, escape it with \ . For eg.: My\!password

TEE License Agreement

Please read and agree with TEE License Agreement (EULA) before using this container.

The image contains the tweak to have this EULA accepted. Else you would get the following output from the console client:

victor@unclev:/srv/storage/Programming/tfs$ tf workspace -new VICTOR -collection:https://mytfshost.net/tfs/My%20Collection
Error: You must accept the End User License Agreement for this product
Run 'tf eula' to accept the End User License Agreement.

The image (or container) with EULA accepted returns:

victor@unclev:~$ tf eula
You have already accepted the terms of this End User License Agreement.

Volumes

• /home/tf/projects
  This folder is added specifically as a placeholder of the workfold in thems of the TF client. Any folder mapped to it in the tee-clc environment is considered as the /home/tf/projects work folder. That is tee-clc-docker is designed to work with the only work folder.
  Leaving this directory not mapped does not make any sense.
• /home/tf/.microsoft/Team Foundation/4.0/Cache The folder where tee-clc stores relevant data, - so called Cache. For eg., workspace. You MUST map this directory to a persistent location for the tf client working correctly.
• /home/tf/.microsoft/Team Foundation/4.0/Configuration The folder where tee-clc stores data.
  You may want having the Configuration folder mapped to persistent storage (for storing settings and data between tf executions).
• /home/tf/.microsoft/Team Foundation/4.0/Logs The folder where tee-clc stores execution logs.
  You may want having the Logs folder mapped to persistent storage for debugging, inspecting tee-clc executions.

Bash alias

Lets assume the container is started and not re-used. The minimum required command would be as follows:

alias tf='docker run --rm -v "$(pwd):/home/tf/projects"  -v "/srv/microsoft/tf/cache:/home/tf/.microsoft/Team Foundation/4.0/Cache" unclev/tf tf'

Hostname

It is not strictly necessary, but it is highly recommended that you spsefify the same hostname of the container on each start. Normal set up includes specifying workspace on the tf client host. The TF saves workspaces against the tf client hostname.

Specifying credentials

The above approaches often cause the error like follows

victor@unclev:/srv/storage/Programming/tfs$ tf workspace -new VICTOR -collection:"https://mytfshost.net/tfs/My Collection"
Default credentials are unavailable because no Kerberos ticket or other authentication token is available.
Username: A client error occurred: Authentication credentials were not explicitly provided and could not be determined from any workspace, or argument paths provided.  Use the workspace or login option (if the noprompt option is specified, the login option value must contain the username, domain, and password).

Usually the Bash alias has to be modified so that it define credentials. With hostnames specified it would be something like:

alias tf='docker run --rm --hostname LINUX -v "$(pwd):/home/tf/projects" -v "/srv/microsoft/tf/cache:/home/tf/.microsoft/Team Foundation/4.0/Cache" unclev/tf tf -login:MYDOMAIN\\myusername,my\!pass -collection:"https://mytfshost.net/tfs/My Collection"'

The above you work with the only collection "https://mytfshost.net/tfs/My Collection". If this is not suitable to you, exclude the -collection option form the alias.

Commands

• Available commands and their options:

tf | more

Examples

With the alias:

alias tf='docker run --rm --hostname LINUX -v "$(pwd):/home/tf/projects"  -v "/srv/microsoft/tf/cache:/home/tf/.microsoft/Team Foundation/4.0/Cache" unclev/tf tf -login:MYDOMAIN\\myusername,my\!pass -collection:"https://mytfshost.net/tfs/My Collection"'

List all directories in "MY.Project":

$ tf dir $/MY.Project
$/MY.Project:
$BuildProcessTemplates
$PRJ40DEV
$CURRENT
$FUTURE
$N02Fix
current_change_log.txt
future_change_log.txt

7 item(s).

Create a new workspace:

victor@unclev:/srv/storage/Programming/tfs$ tf workspace -new VICTOR
Workspace 'VICTOR' created.

List cashed workspaces:

victor@unclev:/srv/storage/Programming/tfs$ tf workspaces
Collection: https://mytfshost.net/tfs/my collection/
Workspace Owner               Computer Comment
--------- ------------------- -------- ----------------------------------------
VICTOR    Kulitchenko, Victor LINUX    

Find specific version(s) by label

victor@unclev:/srv/storage/Programming/tfs$ tf labels -format:brief | grep v1.20
v1.20.30                           Kulitchenko, Victor Jan 30, 2013 4:51:23 AM

Create a Mapping Between a Local Directory and Source Control (Note: this maps the current folder v1.20.30 to its internal representation /home/tf/projects)

victor@unclev:/srv/storage/Programming/tfs$ mkdir v1.20.30
victor@unclev:/srv/storage/Programming/tfs$ cd v1.20.30
victor@unclev:/srv/storage/Programming/tfs/v1.20.30$ tf workfold -map $/MY.Project /home/tf/projects -workspace:VICTOR

Get the specified version (by label)

victor@unclev:/srv/storage/Programming/tfs/v1.20.30$ tf get -version:Lv1.20.30
/home/tf/
Getting Fix/fixtimeinsf.prg
[...]
Getting wincrypt.h
victor@unclev:/srv/storage/Programming/tfs/v1.20.30$ ls -l
drwxr-xr-x 2 victor victor    4096 янв  2 13:37 Fix
[...]
-rw-r--r-- 1 victor victor   10667 янв  2 13:38 wincrypt.h

Get the latest version

victor@unclev:/srv/storage/Programming/tfs/v1.20.30$ mkdir ../latest && cd ../latest
victor@unclev:/srv/storage/Programming/tfs/latest$ tf get -version:T
[...]
Deleting /home/tf/projects/Fix
ls -l
drwxr-xr-x  2 victor victor 4096 янв  2 13:55 BuildProcessTemplates
drwxr-xr-x  3 victor victor 4096 янв  2 13:55 prj40dev
drwxr-xr-x 13 victor victor 4096 янв  2 13:56 PRJ40DEV
drwxr-xr-x  4 victor victor 4096 янв  2 13:56 current
drwxr-xr-x 12 victor victor 4096 янв  2 13:57 CURRENT
-rw-r--r--  1 victor victor 4056 янв  2 13:57 current_change_log.txt
drwxr-xr-x  5 victor victor 4096 янв  2 13:57 future
drwxr-xr-x 12 victor victor 4096 янв  2 13:58 FUTURE
-rw-r--r--  1 victor victor  612 янв  2 13:58 future_change_log.txt
drwxr-xr-x  4 victor victor 4096 янв  2 13:58 N02ReclassFix

NB! It gets directories (eg. PRJ40DEV and prj40dev) as they appear in the item paths. As the linux filesystem is case sensitive, it stores them twice even though the Windows version treats them as the same directory.

Create another project mapping.

victor@unclev:/srv/storage/Programming/tfs/latest$ mkdir -p ../DEV.Shared/SomeCoolLib && cd ../DEV.Shared/SomeCoolLib
victor@unclev:/srv/storage/Programming/tfs/DEV.Shared/SomeCoolLib$ DEV.Shared/SomeCoolLib$ tf workfold -map $/DEV.Shared/SomeCoolLib /home/tf/projects -workspace:VICTOR
victor@unclev:/srv/storage/Programming/tfs/DEV.Shared/SomeCoolLib$ tf workfold
===============================================================================
Workspace:  VICTOR
Collection: https://mytfshost.net/tfs/my collection/
 $/DEV.Shared/SomeCoolLib: /home/tf/projects

From the tf point this is the folder of the previous project, doing get it deletes the previously mapped project files from the mapped folder (even though the actual folder is another folder and the previos project has never existed there). This may case unexpected behaviour.

Persistent projects (working with several projects at a time)

If you work with several projects in a parallel then consider mapping your projects folder to /home/tf/projects:

alias tf='docker run --rm --hostname LINUX -v "/full_path_to_your_projects_workdir:/home/tf/projects"  -v "/srv/microsoft/tf/cache:/home/tf/.microsoft/Team Foundation/4.0/Cache" unclev/tf tf -login:MYDOMAIN\\myusername,my\!pass -collection:"https://mytfshost.net/tfs/My Collection"'