https://forge-allura.apache.org/p/allura/tickets/You searched for labels:"scm"enTue, 30 Oct 2018 14:21:20 -0000https://forge-allura.apache.org/p/allura/tickets/8141/Hi, As per scm port for git in development.ini is 8022 and when i created git repo in Allura web interface, it shows me the repo http url to download.. when try to download or clone repo it never work.. Even the port 8022 not listening on host. development.ini config: scm.host.https.git = https://$username@localhost:8022/scm-repo$path scm.host.https_anon.git = https://localhost:8022/scm-repo$path scm.host.https.hg = https://$username@localhost:8022/scm-repo$path scm.host.https_anon.hg = https://localhost:8022/scm-repo$path scm.host.https.svn = https://localhost:8022/scm-repo$path/ scm.host.https_anon.svn = https://localhost:8022/scm-repo$path/ Repo URL: git clone https://root@localhost:8022/scm-repo/p/gittest --init---gittest I tried replacing localhost to public IP address but didnt work. I didnt understand about apache part in https://forge-allura.apache.org/docs/getting_started/scm_host.html. Can you provide me more detailed info how to configure git and svn to work. Thanks, Mohiddin MohiddinTue, 30 Oct 2018 14:21:20 -0000https://forge-allura.apache.org/p/allura/tickets/8141/https://forge-allura.apache.org/p/allura/tickets/8080/When the SCM tool is installed on docker, need to manually give permissions to the allura-data directory before pushing to remote. Command: ```bash docker exec -it alluragit_web_1 chmod -R 777 /allura-data ``` Error: ``` remote: fatal: Unable to create temporary file '/allura-data/scm/git/p/code.git/./objects/pack/tmp_pack_XXXXXX': Permission denied error: failed to push some refs to '/allura-data/scm/git/p/code' ```Rohan VermaThu, 14 Apr 2016 20:08:00 -0000https://forge-allura.apache.org/p/allura/tickets/8080/https://forge-allura.apache.org/p/allura/tickets/6224/Tests using https://sourceforge.net/p/allura/pastebin/518bb63d7929e51a2548196a/ indicate that we should be able to switch the tree browsing for git to get its info directly from GitPython without relying on the fragile index data. This should give us comparable performance without having to wait for the index to complete. Create a GitTreeBrowser subclass to use this implementation (with tests), preferably using AJAX (tree / blob names can be pulled quickly, AJAX to fill in the rest of the last-commit data).Cory JohnsThu, 20 Aug 2015 22:07:55 -0000https://forge-allura.apache.org/p/allura/tickets/6224/https://forge-allura.apache.org/p/allura/tickets/6196/When building the `CommitRunDoc` on `aufs3-linux` on the `aufs` project: https://sourceforge.net/p/allura/pastebin/5182b535f1fd8d0850865493/ There is/was some max recursion failsafe handling added some time ago, so look for that. Also occurs on linux-kernel, de-2440, and chumby-kernelCory JohnsMon, 03 Jun 2013 14:25:48 -0000https://forge-allura.apache.org/p/allura/tickets/6196/https://forge-allura.apache.org/p/allura/tickets/5889/[#5785] fixed import errors when submodules are used in git repos, but they still show up as files in the tree listing (instead of folders) and return a 404 when clicked to view (instead of showing some message indicating that submodules cannot be viewed, similar to viewing binary files).Cory JohnsWed, 15 Apr 2015 06:41:23 -0000https://forge-allura.apache.org/p/allura/tickets/5889/https://forge-allura.apache.org/p/allura/tickets/5781/Super slow requests on viewing files in SVN if missing TreeDocs have to be generated.Cory JohnsThu, 20 Aug 2015 22:07:54 -0000https://forge-allura.apache.org/p/allura/tickets/5781/https://forge-allura.apache.org/p/allura/tickets/5734/Currently, diffs are computed by seeing if a tree / blob node is different in *any* of the parents of the commit. Instead, the node should only be considered changed in the commit if it is different from the corresponding node in *all* of the parent commits. (For a commit with a single parent, this amounts to the same thing. But for merge commits, if the file was changed in one branch but was not touched in the other branch and there are thus no conflicts, then it wasn't really changed in the merge commit.) This doesn't affect SVN, and probably doesn't affect Mercurial, due to the way they do branching, but Mercurial should be confirmed that it will behave well with this new logic.Cory JohnsTue, 06 Jan 2015 14:27:12 -0000https://forge-allura.apache.org/p/allura/tickets/5734/https://forge-allura.apache.org/p/allura/tickets/5733/`Commit._diffs_copied()` is used to determine if a removed blob was actually moved or renamed, possibly with some changes. However, it is called every time a commit is viewed and hits every file removed from a commit, and it is slow enough to be a problem. Some ideas for optimizing it: * Short-circuit identical blob comparisons by comparing the blob hash first, as is done w/ trees * Use `SequenceMatcher.real_quick_ratio()` to get the upper-bound on the ratio to exclude obvious non-matches quickly, probably followed up with `quick_ratio()` and/or `ratio()` to confirm a match * Raise the `DIFF_SIMILARITY_THRESHOLD` and break after a single match instead of continuing to test all files (though this could give false matches, so maybe not do this one) * Exclude binary or particularly large blobs Finally, we should almost certainly move this computation to `compute_diffs()` instead of doing it every time the commit's diffs are used. Also, currently, children of removed (or the removed side of moved/renamed) trees are not included in the diff to avoid hitting this performance issue too often, which causes the added portion of moved/renamed trees to look like brand new files. Once the performance of `_diffs_copied()` is more reasonable and/or pre-computed, the removed trees short-circuit in `compute_diffs()` needs to be removed.Cory JohnsTue, 11 Aug 2015 20:33:03 -0000https://forge-allura.apache.org/p/allura/tickets/5733/https://forge-allura.apache.org/p/allura/tickets/4624/*Originally created by:* sechastain Hey folks. We're running a hacked up Allura on RHEL 6.2, and trying to use Python 2.6. I know there is the stated compatibility of 2.7, but we haven't run into any significant problems yet... until today. The problem we were encountering was in `repo_refresh.py` and `model/repo.py`, with this error: ValueError: zero length field name in format One of the changes introduced for newest repo stuff uses the new, index-less regex field notation (as described on [stackoverflow](http://stackoverflow.com/a/8498327/328451)) Here's a patch that fixes this little shortcoming of 2.6. diff --git a/Allura/allura/model/repo.py b/Allura/allura/model/repo.py index 442be0a..7cdb9b8 100644 --- a/Allura/allura/model/repo.py +++ b/Allura/allura/model/repo.py @@ -326,7 +326,7 @@ def ls(self): lc_index = dict( (lc.object_id, lc.commit_info) for lc in LastCommitDoc.m.find(dict( - _id=re.compile("^{}:".format(self.repo._id)), + _id=re.compile("^{0}:".format(self.repo._id)), object_id={'$in': oids}))) results = [] def _get_last_commit(oid): diff --git a/Allura/allura/model/repo_refresh.py b/Allura/allura/model/repo_refresh.py index 51c02e2..607451e 100644 --- a/Allura/allura/model/repo_refresh.py +++ b/Allura/allura/model/repo_refresh.py @@ -316,7 +316,7 @@ def _walk_tree(tree, tree_index): d['object_id'] for d in last_commit_collection.find( dict(object_id={'$in': rhs_tree_ids}, - _id=re.compile("^{}:".format(repo_id))), + _id=re.compile("^{0}:".format(repo_id))), { 'object_id': 1, '_id': 0 })) for tree_id in rhs_tree_ids: if tree_id not in last_commits: Since we'll be continuing to work with 2.6, we're happy to help in watching the compatibility... until some really new and novel feature is required, we hope this compatibility can be maintained!AnonymousThu, 20 Aug 2015 22:07:55 -0000https://forge-allura.apache.org/p/allura/tickets/4624/https://forge-allura.apache.org/p/allura/tickets/780/*Originally created by:* n_oostendorp When I view a file that has only been "added" but hasn't been changed, i should not see a "diff" link (currently the target of this link causes 500 ISE)AnonymousThu, 20 Aug 2015 22:07:57 -0000https://forge-allura.apache.org/p/allura/tickets/780/