2801: Subversion reviews being treated as Perforce reviews?
- Fixed
- Review Board
| eyu****@gmai***** (Google Code) (Is this you? Claim this profile.) | |
 |
|
| Jan. 17, 2014 |
What version are you running? 1.6.13 What's the URL of the page containing the problem? no particular URL What steps will reproduce the problem? 1. make some change inside a svn repo 2. post-review 3. output: post-review path/to/file Path 'path/to/file' does not match a valid Perforce path. What is the expected output? What do you see instead? it was working fine with RB 1.6.9/RBTool 0.3.4 by just doing post-review inside a svn repo. After upgrading to 1.6.13/RBTool 0.4.2, I'm seeing post-review tries to match the affected file against a Perforce repo.
Can you provide a --debug log for RBTools? This sounds like an RBTools-specific problem.
-
+ NeedInfo
>>> RBTools 0.4.2 >>> Python 2.6.1 (r261:67515, Jul 22 2010, 20:06:52) [GCC 4.1.1] >>> Running on Linux-2.6.18-274.7.1.el5-x86_64-with-redhat-5.7-Final >>> Home = /home/user >>> Current Directory = /home/user/svn/app >>> Checking the repository type. Errors shown below are mostly harmless. [2012-11-15 12:31:12] DEBUG: Checking for a CVS repository... [2012-11-15 12:31:12] DEBUG: Checking for a ClearCase repository... [2012-11-15 12:31:12] DEBUG: Checking for a Git repository... [2012-11-15 12:31:12] DEBUG: Running: git rev-parse --git-dir [2012-11-15 12:31:12] DEBUG: Command exited with rc 128: ['git', 'rev-parse', '--git-dir'] fatal: Not a git repository (or any parent up to mount parent ) Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set). --- [2012-11-15 12:31:12] DEBUG: Checking for a Mercurial repository... [2012-11-15 12:31:12] DEBUG: Checking for a Perforce repository... [2012-11-15 12:31:12] DEBUG: Running: p4 info [2012-11-15 12:31:14] DEBUG: Running: diff --version [2012-11-15 12:31:14] DEBUG: repository info: Path: perforce....com:1666, Base path: None, Supports changesets: True >>> Finished checking the repository type. [2012-11-15 12:31:14] DEBUG: Running: p4 counters >>> HTTP GETting api/ >>> HTTP GETting https://reviewboard....com/api/info/ >>> Using the new web API [2012-11-15 12:31:15] INFO: Generating diff for changenum default [2012-11-15 12:31:15] DEBUG: Running: p4 opened -c default Couldn't find any affected files for this change. Looks like it should continue with SVN after it failed to try p4 rather than stop
The problem is that it's seeing a valid Perforce environment. It's running p4 info and getting results back saying where the repository is. So that's winning over SVN. What needs to happen is that, when not in a Perforce environment, 'p4 info' shouldn't be showing a valid server. Do you know why it is?
I was able to trick post-review by giving a wrong p4 location. but is this the only solution? My company uses p4, svn, git... altogether. It seems not-so-clever to short circuit once a valid p4 environment is found to assume it is a perforce change, doesn't it?
-
- NeedInfo + Confirmed -
+ Component-RBTools
-
- Confirmed + PendingReview