Summary:
Fixes T4276. This adds "Change is enormous" to pre-commit content rules so we can, e.g., just reject these and not worry about them elsewhere.
Also, use the same numeric limits across the mechanisms so there's a consistent definition of an "enormous" changeset.
Test Plan:
- Set enormous limit to 15 bytes, pushed some changes, got blocked by a rule.
- Set it back, pushed OK.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4276
Differential Revision: https://secure.phabricator.com/D7887
Summary:
Ref T4276. When a change is larger than 2GB, PHP can not read the entire change into a string, so Herald can not process it.
Additionally, we already have a time limit for practical reasons, but it's huge (probably incorrectly). To deal with these things:
- Add an optional byte limit to `diffusion.rawdiffquery`.
- Make the query with a 1GB limit.
- Reduce the diff timeout from 15 hours to 15 minutes.
- Add a "Changeset is enormous" field. This field is true for changes which are too large to process.
This generally makes behaviors more sane:
- We'll always make progress in Herald in a reasonable amount of time.
- Installs can write global rules to handle (or reject) these types of changes.
Test Plan: Set limit to 25 bytes instead of 1GB and ran test console on various changes.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4276
Differential Revision: https://secure.phabricator.com/D7885
Summary:
Fixes T4189. Ref T4151. Allows repositories to have additional custom hooks for operations which can't be expressed with Herald (one such operation is lint).
This adds only local hook directories, since they're easier to use with existing hooks than global directories. I might add global directories eventually.
This doesn't support Mercurial since we have no demand for it and it's more complicated (we lose compatibility and power by just dropping a `hooks.d/` somewhere).
Test Plan:
- Pulled hosted SVN and Git repos to verify the hook directories generate correctly.
- Added a variety of hooks to the hook directories (echo + pass, fail).
- Pushed commits and verified the hooks fired (output expected info, or failed).
- Verified push log reflected the correct error code ("3", external) and detail ("nope.sh") when rejecting.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4151, T4189
Differential Revision: https://secure.phabricator.com/D7884
Summary:
Fixes T4264. Adds:
- New "Repository's projects" field to Herald pre-commit rules, so you can write global rules which act based on projects.
- Allows pre-ref/pre-content rules to bind to projects, and fire for all repositories in that project, so users with limited power can write rules which apply to many repositories.
- The pre-ref and pre-content classes were starting to share a fair amount of code, so I made them both extend an abstract base class.
Test Plan: Wrote new pre-ref and pre-content rules bound to projects, then pushed commits into repositories in those projects and not in those projects. The "repository projects" field populated, and the rules fired for repositories in the relevant projects.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4264
Differential Revision: https://secure.phabricator.com/D7883
Summary: Ref T4264. Ref T2628. Ref T3102. Allows you to associate repositories with projects. In the future, you'll be able to write Herald object rules against projects, use Herald fields like "Repository's projects", and search by project.
Test Plan: See screenshots.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3102, T4264, T2628
Differential Revision: https://secure.phabricator.com/D7881
Summary:
Repositories currently have a no-UI "shortcut" feature which is only used by Facebook (and I'm not sure it's even used). As implemented, this feature is policy-oblivious and kind of nonsensical. Throw it away.
I'm open to reimplementing this, but I want to see some level of interest in it before I do. The new implementation would add shortcuts to each repository, similar to how mirrors work. My original plan was to follow this up with such an implementation (it's half-implemented in my sandbox), but as I worked through it I'm not sure it's really valuable.
Test Plan: Browsed repository list, grep.
Reviewers: btrahan
Reviewed By: btrahan
CC: FacebookPOC, aran
Differential Revision: https://secure.phabricator.com/D7862
Summary: Currently we markup `rXabcd`, but not `rX` on its own. Mark these up as repository object names.
Test Plan: Typed `rPOEMS`, `rPOEMS1`, `rPOEMS139893189`, etc.
Reviewers: btrahan, dctrwatson
Reviewed By: btrahan
CC: aran, poop
Differential Revision: https://secure.phabricator.com/D7859
Summary:
Fixes T4242. It's currently possible to set nonsense defaults and create repositories with unintended policies, because policy configuration isn't part of creation. Instead:
- put a policy page into the creation workflow;
- require the selection of valid policies (i.e., prevent creating a repository you can't view / edit).
Test Plan:
- Created imported and hosted repositories, hit policy selection.
- Edited policies of existing repositories.
- Tried to set nonsense policies.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4242
Differential Revision: https://secure.phabricator.com/D7856
Summary:
Ref T4264. Allows you to create "Object" rules, in addition to Global and Personal rules. If you choose to create an Object rule, you'll be prompted to select an object on a new screen. You must be able to edit and object in order to create rules for it.
Ref T3506. This makes "All" the default filter for the transcript view, which should reduce confusion on smaller installs.
Test Plan:
- Created non-object rules.
- Created object rules.
- Triggered object rules against matching and unmatching objects.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T3506, T4264
Differential Revision: https://secure.phabricator.com/D7853
Summary: Ref T4264. Lays the groundwork for new "Object" rule types. Prevents personal "Hook" rules, which don't make any sense.
Test Plan: Created new Maniphest (global/personal available) and Ref Hook (global only) rules.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4264
Differential Revision: https://secure.phabricator.com/D7852
Summary: Some discussion on IRC. This is more consistent with other disabled items, which are click-to-explain.
Test Plan: Viewed UI, clicked link.
Reviewers: btrahan, dctrwatson, asherkin
Reviewed By: asherkin
CC: aran
Differential Revision: https://secure.phabricator.com/D7857
Summary:
Fixes T4270. When you download raw file content, diffs, and patches we currently give them default (all users) visibility.
Instead, bind them to the repository or revision in question.
(This code could use a bit of cleanup at some point.)
Test Plan: Hit the patch and content download links in Diffusion and the patch download link in Differential, got restricted files with accurate policy bindings.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4270
Differential Revision: https://secure.phabricator.com/D7849
Summary: Ref T4264. Instead of a dropdown, make this step more informative.
Test Plan: {F93928}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4264
Differential Revision: https://secure.phabricator.com/D7846
Summary: If `0` isn't an ancestor of the current branch, the `0::x` construction fails. This is uncommon, but not wildly unreasonable. The `ancestors()` construction is simpler anyway.
Test Plan: Viewed some `hg` repos locally (change history, file history) without anything suspicious cropping up.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D7844
Summary:
Ref T4195. A legitimate rule which needs this field is "do not allow commits as root". Interestingly, we have exactly one commit as root in each Phabricator, Arcanist and libphutil.
Since the committer and author don't need to be Phabricator accounts (just the Pusher), the existing "Committer" and "Author" fields can't express this rule (they'll be empty).
Test Plan: {F93406}
Reviewers: btrahan
Reviewed By: btrahan
CC: SEJeff, aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7841
Summary:
Fixes T4195. Allows you to write a rule against a commit's branches.
This completes outstanding work on T4195.
Test Plan: Pushed to Git and Mercurial repositories and verified branches were selected correctly by examining transcripts.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7820
Summary:
Fixes T4257. The `hg heads` command exits with an error code and no output in an empty repository.
Just ignore the error code: we don't have a great way to distinguish between errors, and we ran another `hg` command moments before, so we have at least some confidence it isn't a PATH sort of thing.
Test Plan: Created a new Mercurial repository and pushed to hit the error in T4257. Applied this fix and got a clean push with an accurate push log.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4257
Differential Revision: https://secure.phabricator.com/D7817
Summary:
Ref T4195. This allows you to write rules which disallow merge commits.
Also make the reject message a little more useful.
Test Plan:
remote: This push was rejected by Herald push rule H27.
remote: Change: commit/daed0d448404
remote: Rule: No Merges
remote: Reason: No merge commits allowed. If you must push a merge, include "@force-merge" in the commit message.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7809
Summary: Ref T4195. Ref T2783. We have an old-school implementation of this; move it into a LowLevel query and make callers all run through Conduit. I need the LowLevel query for hooks, to implement an "is merge commit" Herald rule.
Test Plan:
- Ran query via Conduit for SVN, Mercurial, Git.
- Parsed a commit which closed a revision, attach/closed worked correctly.
- Browsed Diffusion.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195, T2783
Differential Revision: https://secure.phabricator.com/D7808
Summary: Refs T4195. Fixes T3936. You can't currently write rules like "block commits unless they're attached to an **accepted** revision"; allow that.
Test Plan: Pushed commits into a rule with this field, saw it work / not crash.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran, mbishopim3
Maniphest Tasks: T3936, T4195
Differential Revision: https://secure.phabricator.com/D7807
Summary: Ref T4195. Allows you to write revision-based commit hooks, e.g. block all commits with no corresponding revision.
Test Plan:
Here's are the fields populating:
{F90989}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7806
Summary: Ref T4195. I need to query commit metadata to figure out which revision a commit is associated with. Move this out of the MessageParser so the code can be called from the HookEngine.
Test Plan: Used `reparse.php` to reparse a variety of SVN, Mercurial and Git commits. Used `var_dump()` to verify sensible fields were returned.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7805
Summary: Ref T4195. I need this for the Herald pre-commit rules, and it generally simplifies things.
Test Plan: Used `reparse.php` plus `var_dump()` to inspect refs in Git, Mercurial and SVN repos. They all looked correct and reparsed correctly.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7804
Summary:
There's no particular reason to allow the user to edit the clone URI field in Diffusion; editing it has no meaning and if you fat finger the keyboard, it's quite possible that the user will either accidentally clear and/or modify the URI before copying (bit me this morning).
Adding a readonly attribute to the input field allows the same benefit (URI is easily selectable) while preventing such accidental input. Fixes T4246.
Test Plan: Verified that the desired behavior is present in both Chrome, Safari, and Firefox. Field remains selectable with one click, but field is not editable.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Maniphest Tasks: T4246
Differential Revision: https://secure.phabricator.com/D7810
Summary: Ref T4195. Adds "Author" and "Committer" fields.
Test Plan:
Created a rule using these fields:
{F90897}
...then pushed git, mercurial and svn commits and verified the correct values populated in the transcript:
{F90898}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7802
Summary:
Ref T4195. To implement the "Author" and "Committer" rules, I need to resolve author/committer strings into Phabricator users.
The code to do this is currently buried in the daemons. Extract it into a standalone query.
I also added `bin/repository lookup-users <commit>` to test this query, both to improve confidence I'm getting this right and to provide a diagnostic command for users, since there's occasionally some confusion over how author/committer strings resolve into valid users.
Test Plan:
I tested this using `bin/repository lookup-users` and `reparse.php --message` on Git, Mercurial and SVN commits. Here's the `lookup-users` output:
>>> orbital ~/devtools/phabricator $ ./bin/repository lookup-users rINIS3
Examining commit rINIS3...
Raw author string: epriestley
Phabricator user: epriestley (Evan Priestley )
Raw committer string: null
>>> orbital ~/devtools/phabricator $ ./bin/repository lookup-users rPOEMS165b6c54f487c8
Examining commit rPOEMS165b6c54f487...
Raw author string: epriestley <git@epriestley.com>
Phabricator user: epriestley (Evan Priestley )
Raw committer string: epriestley <git@epriestley.com>
Phabricator user: epriestley (Evan Priestley )
>>> orbital ~/devtools/phabricator $ ./bin/repository lookup-users rINIH6d24c1aee7741e
Examining commit rINIH6d24c1aee774...
Raw author string: epriestley <hg@yghe.net>
Phabricator user: epriestley (Evan Priestley )
Raw committer string: null
>>> orbital ~/devtools/phabricator $
The `reparse.php` output was similar, and all VCSes resolved authors correctly.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T1731, T4195
Differential Revision: https://secure.phabricator.com/D7801
Summary: Ref T4195. Even though we use `svnlook` in the hook itself, I need this query elsewhere, so provide it and merge the classes into one which does the right thing.
Test Plan:
- Used `reparse.php` to reparse messages for Git, SVN and Mercurial commits, using `var_dump()` to examine the commit refs for sanity.
- Used `reparse.php` to reparse changes for an SVN commit.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7800
Summary: There were a number of places that were generating nonsense queries for both hosted and non-hosted subversion repositories.
Test Plan: Attempted several activities in Diffusion with both a hosted and non-hosted subversion repository, including viewing various types of diffs and raw files.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Differential Revision: https://secure.phabricator.com/D7799
Summary: If you push a large binary and the data crosses multiple data frames, we can end up in a loop in the parser.
Test Plan:
After this change, I was able to push a 95MB binary in 7s, which seems reasonable:
>>> orbital ~/repos/INIS $ svn st
A large2.bin
>>> orbital ~/repos/INIS $ ls -alh
total 390648
drwxr-xr-x 6 epriestley admin 204B Dec 18 17:14 .
drwxr-xr-x 98 epriestley admin 3.3K Dec 16 11:19 ..
drwxr-xr-x 7 epriestley admin 238B Dec 18 17:14 .svn
-rw-r--r-- 1 epriestley admin 80B Dec 18 15:07 README
-rw-r--r-- 1 epriestley admin 95M Dec 18 16:53 large.bin
-rw-r--r-- 1 epriestley admin 95M Dec 18 17:14 large2.bin
>>> orbital ~/repos/INIS $ time svn commit -m 'another large binary'
Adding (bin) large2.bin
Transmitting file data .
Committed revision 25.
real 0m7.215s
user 0m5.327s
sys 0m0.407s
>>> orbital ~/repos/INIS $
There may be room to improve this by using `PhutilRope`.
Reviewers: wrotte, btrahan, wotte
Reviewed By: wotte
CC: aran
Differential Revision: https://secure.phabricator.com/D7798
Summary: Ref T4195. Same as D7793, but for mercurial. (As usual, SVN needs some goofy nonsense instead, so the next diff will just make this field work.)
Test Plan: Ran `reparse.php` on Git and Mercurial commits, var_dump'd the output and it looked correct.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7795
Summary: Ref T4195. I need to issue this command from the pre-commit hook to get commit bodies for hooks.
Test Plan: Ran `reparse.php --message --trace` and dumped the $ref, which looked correct.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7793
Summary: We currently have a lot of calls to `addCrumb(id(new PhabricatorCrumbView())->...)` which can be expressed much more simply with a convenience method. Nearly all crumbs are only textual.
Test Plan:
- This was mostly automated, then I cleaned up a few unusual sites manually.
- Bunch of grep / randomly clicking around.
Reviewers: btrahan, chad
Reviewed By: btrahan
CC: hach-que, aran
Differential Revision: https://secure.phabricator.com/D7787
Summary: Ref T4195. Add Mercurial support to the content hook phase.
Test Plan:
Here are some `commit` push logs for a Mercurial repo:
{F90689}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7792
Summary: Ref T4195. Adds support for diff content rules.
Test Plan: Pushed SVN and Git changes through, saw them generate reasonable transcripts. Mercurial still isn't hooked up to this phase.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7791
Summary: Ref T4195. This doesn't provide any interesting fields yet (content, affected paths, commit message) but fires the hook correctly.
Test Plan: Added a blocking hook and saw it fire.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7789
Summary:
Ref T615. Ref T4237. With `--debug`, Mercurial will echo an "ignoring untrusted configuration option" warning **to stdout** if `.hgrc` has the wrong owner.
However, we need `--debug` to make `{parents}` usable, at least until the patches I got into the upstream are widely deployed. So after getting `--debug` output, strip off any leading warnings.
These warnings should always be in English, at least, since we set `LANG` explicitly.
Test Plan: Unit tests. @asherkin, maybe you can confirm this? I can't actually get the warning, but I think my `hg` in PATH is just a bit out of date.
Reviewers: asherkin, btrahan
Reviewed By: asherkin
CC: asherkin, aran
Maniphest Tasks: T615, T4237
Differential Revision: https://secure.phabricator.com/D7784
Summary: Ref T4195. SVN has no such thing as refs (I was thinking about writing a quasi-ref anyway like `HEAD: r23 -> r24`, but I'm not sure it would actually be useful). And content is very easy to build.
Test Plan: Pushed some stuff to SVN, got logs from it.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7766
Summary: Ref T4195. This doesn't actually work like I thought it did: it only fires locally, when you run `hg tag`. Mercurial tags are also weird and basically don't make any sense and everyone should use bookmarks instead. We could implement some flavor of this eventually, but I'd like to see users request it first. They can implement their own with content-based hooks once those work, anyway.
Test Plan: This code didn't do anything.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7765
Summary:
Ref T4195. This pulls the central logic of HookEngine up one level and makes all the git stuff genrate PushLogs.
In future diffs, everything will generate PushLogs and we can hand those off to Herald.
Test Plan:
Pushed a pile of valid/invalid stuff:
{F89256}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7761
Summary: Fixes T4224. If you `git merge-base A B`, and they have //no// ancestor, the command exits with an error. Assume errors mean "no ancestry" and continue.
Test Plan: Completely rewrite a repository with a `--force` push.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4224
Differential Revision: https://secure.phabricator.com/D7756
Summary: We run `git` on a different port than 22, so would like to reflect this change in the UI.
Test Plan: Set diffusion.ssh-port in settings, then make sure it's reflected on the Diffusion repository Clone URI.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran, dctrwatson
Differential Revision: https://secure.phabricator.com/D7755
Summary: Fixes T4223. The output of `ls-tree` is partially delimited by spaces
and partially delimited by `\t`. The code I added in D7744 to help debug the
issue in T4159 doesn't work properly for files with 7 or more bytes in their
filesize, because the internals use `%7s`.
Auditors: btrahan
Summary: Most checks were actually in place, but `ExecFuture` throws a `CommandException` which wasn't taken into account.
Test Plan: look at the first command and no longer saw an exception. Also, other commits worked as well.
Reviewers: richardvanvelzen
Reviewed By: richardvanvelzen
CC: krisbuist, Korvin, epriestley, aran
Differential Revision: https://secure.phabricator.com/D7730
Summary: This locks push logs down a little bit and makes them slightly more administrative. Primarily, don't show IPs to googlebot, etc.
Test Plan: Viewed push logs as edit and non-edit users.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D7722
Summary:
Ref T4195. Like the previous diffs, these both create a useful log and give us an object to hand off to Herald.
Surface this information in Diffusion, too, and clean things up a little bit.
Test Plan: {F87565}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7718
Summary: Ref T4195. Add UI options to filter push logs by pusher and repository. Add a link from the repository view page to the push logs.
Test Plan: Viewed a hosted repository, clicked logs link, saw logs. Filtered lgos by repo/pusher.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7713
Summary: Ref T4195. Stores remote address and protocol in the logs, where possible.
Test Plan: Pushed some stuff, looked at the log, saw data.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7711
Summary:
Ref T4195. This log serves two purposes:
- It's a log, so you can see what happened. Particularly, in Git/Hg, there is no other way to tell:
- Who //pushed// a change (vs committed / authored)?
- When was a change pushed?
- What was the old value of some tag/branch before someone destroyed it?
- We can hand these objects off to Herald to implement pre-commit rules.
This is a very basic implementation, but gets some data written and has a basic UI for it.
Test Plan: {F87339}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4195
Differential Revision: https://secure.phabricator.com/D7705
Summary:
Ref T4189. This doesn't add any rules yet, but does all the heavy lifting to figure out what's changed and put it in a consuamble (if somewhat ad-hoc) datastructure, which lists all the ref and tag modifications and all the new commits in a consistent way.
From here, it should be fairly straightforward to add top-level rules (e.g., ff pushes only).
Test Plan: Output is huge, see comments.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4189
Differential Revision: https://secure.phabricator.com/D7687
Summary: Ref T4189. Fixes T2066. Mercurial has a //lot// of hooks so I'm not 100% sure this is all we need to install (we may need separate hooks for tags/bookmarks) but it should cover most of what we're after at least.
Test Plan:
- `bin/repository pull`'d a Mercurial repo and got a hook install.
- Pushed to a Mercurial repository over SSH and HTTP, with good/bad hooks. Saw hooks fire.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2066, T4189
Differential Revision: https://secure.phabricator.com/D7685
Summary:
Ref T4189. This adds SVN support, which was a little more messy than I though. Principally, we can not use `PHABRICATOR_USER` for Subversion, because it strips away the entire environment for "security reasons".
Instead, use `--tunnel-user` plus `svnlook author` to figure out the author.
Also fix "ssh://" clone URIs, which needs to be "svn+ssh://".
Test Plan:
- Made SVN commits through the hook.
- Made Git commits, too, to make sure I didn't break anything.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4189
Differential Revision: https://secure.phabricator.com/D7683
Summary:
Ref T4189. T4189 describes most of the intent here:
- When updating hosted repositories, sync a pre-commit hook into them instead of doing a `git fetch`.
- The hook calls into Phabricator. The acting Phabricator user is sent via PHABRICATOR_USER in the environment. The active repository is sent via CLI.
- The hook doesn't do anything useful yet; it just veifies basic parameters, does a little parsing, and exits 0 to allow the commit.
Test Plan:
- Performed Git pushes and pulls over SSH and HTTP.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4189
Differential Revision: https://secure.phabricator.com/D7682
Summary: This was broken in rP51fb1ca16d7f.
Test Plan: Imported a repository with file:/// location, it worked.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Differential Revision: https://secure.phabricator.com/D7636
Summary: Fixes T2230. This isn't a total walk in the park to configure, but should work for early adopters now.
Test Plan: Read documentation, browsed UI.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7634
Summary:
Ref T4038. This adds everything except the actual pushing part for mirrors.
This isn't the most beautiful or sophisticated UI, but I want get the authoritative repositories self-hosted and get users beta-ing hosting as soon as possible. We can do transactions, etc., later on.
Test Plan: See screenshots.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4038
Differential Revision: https://secure.phabricator.com/D7632
Summary: Fixes T4122. Ref T2230. Instead of storing credentials on each repository, store them in Passphrase. This allows easy creation/management of many repositories which share credentials.
Test Plan:
- Upgraded repositories.
- Created and edited repositories.
- Pulled HTTP and SSH repositories.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230, T4122
Differential Revision: https://secure.phabricator.com/D7629
Summary:
Ref T2230. When fully set up, we have up to three users who all need to write into the repositories:
- The webserver needs to write for HTTP receives.
- The SSH user needs to write for SSH receives.
- The daemons need to write for "git fetch", "git clone", etc.
These three users don't need to be different, but in practice they are often not likely to all be the same user. If for no other reason, making them all the same user requires you to "git clone httpd@host.com", and installs are likely to prefer "git clone git@host.com".
Using three different users also allows better privilege separation. Particularly, the daemon user can be the //only// user with write access to the repositories. The webserver and SSH user can accomplish their writes through `sudo`, with a whitelisted set of commands. This means that even if you compromise the `ssh` user, you need to find a way to escallate from there to the daemon user in order to, e.g., write arbitrary stuff into the repository or bypass commit hooks.
This lays some of the groundwork for a highly-separated configuration where the SSH and HTTP users have the fewest privileges possible and use `sudo` to interact with repositories. Some future work which might make sense:
- Make `bin/phd` respect this (require start as the right user, or as root and drop privileges, if this configuration is set).
- Execute all `git/hg/svn` commands via sudo?
Users aren't expected to configure this yet so I haven't written any documentation.
Test Plan:
Added an SSH user ("dweller") and gave it sudo by adding this to `/etc/sudoers`:
dweller ALL=(epriestley) SETENV: NOPASSWD: /usr/bin/git-upload-pack, /usr/bin/git-receive-pack
Then I ran git pushes and pulls over SSH via "dweller@localhost". They successfully interacted with the repository on disk as the "epriestley" user.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7589
Summary:
Ref T4039. This fixes an issue where a user with the ability to create repositories could view repositories he is otherwise not permitted to see, by following these steps:
- Suppose you want to see repository "A".
- Create a repository with the same VCS, called "B".
- Edit the local path, changing "/var/repo/B" to "/var/repo/A".
- Now it points at a working copy of a repository you can't see.
- Although you won't be able to make it through discovery (the pull will fail with the wrong credentials), you can read some information out of the repository directly through the Diffusion UI, probably?
I'm not sure this was really practical to execute since there are a bunch of sanity checks along most/all of the major pathways, but lock it down since normal users shouldn't be editing it anyway. In the best case, this would make a mess.
Test Plan: {F81391}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4039
Differential Revision: https://secure.phabricator.com/D7580
Summary:
Small step forward which improves existing stuff or lays groudwork for future stuff:
- Currently, to check for email verification, we have to single-query the email address on every page. Instead, denoramlize it into the user object.
- Migrate all the existing users.
- When the user verifies an email, mark them as `isEmailVerified` if the email is their primary email.
- Just make the checks look at the `isEmailVerified` field.
- Add a new check, `isUserActivated()`, to cover email-verified plus disabled. Currently, a non-verified-but-not-disabled user could theoretically use Conduit over SSH, if anyone deployed it. Tighten that up.
- Add an `isApproved` flag, which is always true for now. In a future diff, I want to add a default-on admin approval queue for new accounts, to prevent configuration mistakes. The way it will work is:
- When the queue is enabled, registering users are created with `isApproved = false`.
- Admins are sent an email, "[Phabricator] New User Approval (alincoln)", telling them that a new user is waiting for approval.
- They go to the web UI and approve the user.
- Manually-created accounts are auto-approved.
- The email will have instructions for disabling the queue.
I think this queue will be helpful for new installs and give them peace of mind, and when you go to disable it we have a better opportunity to warn you about exactly what that means.
Generally, I want to improve the default safety of registration, since if you just blindly coast through the path of least resistance right now your install ends up pretty open, and realistically few installs are on VPNs.
Test Plan:
- Ran migration, verified `isEmailVerified` populated correctly.
- Created a new user, checked DB for verified (not verified).
- Verified, checked DB (now verified).
- Used Conduit, People, Diffusion.
Reviewers: btrahan
Reviewed By: btrahan
CC: chad, aran
Differential Revision: https://secure.phabricator.com/D7572
Summary: We don't actually support this yet, so hide the configuration.
Test Plan: Edited branches for an hg repo.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D7563
Summary:
Ref T2230. As far as I can tell, getting SVN working over HTTP is incredibly complicated. It's all DAV-based and doesn't appear to have any kind of binary we can just execute and pass requests through to. Don't support it for now.
- Disable it in the UI.
- Make sure all the error messages are reasonable.
Test Plan: Tried to HTTP an SVN repo. Tried to clone a Git repo with SVN, got a good error message.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7562
Summary:
Ref T2230. Very rarely, even though we've flushed the connection and sent all the data, we'll close the connection before Git is happy with it and it will flip out with an error like this:
fatal: The remote end hung up unexpectedly
fatal: early EOF
fatal: index-pack failed
This is hard to reproduce because it depends on the order of read/write operations we can't directly control. I only saw it about 2% of the time, by just running `git pull` over and over again.
Waiting for Git to close its side of the connection seems to fix it.
Test Plan: Ran `git clone` a ton of times without seeing the error again. Ran `git push` a ton of times with new commits.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7558
Summary:
Ref T2230. The SVN protocol has a sensible protocol format with a good spec here:
http://svn.apache.org/repos/asf/subversion/trunk/subversion/libsvn_ra_svn/protocol
Particularly, compare this statement to the clown show that is the Mercurial wire protocol:
> It is possible to parse an item without knowing its type in advance.
WHAT A REASONABLE STATEMENT TO BE ABLE TO MAKE ABOUT A WIRE PROTOCOL
Although it makes substantially more sense than Mercurial, it's much heavier-weight than the Git or Mercurial protocols, since it isn't distributed.
It's also not possible to figure out if a request is a write request (or even which repository it is against) without proxying some of the protocol frames. Finally, several protocol commands embed repository URLs, and we need to reach into the protocol and translate them.
Test Plan: Ran various SVN commands over SSH (`svn log`, `svn up`, `svn commit`, etc).
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7556
Summary:
Ref T2230. This is substantially more complicated than Git, but mostly because Mercurial's protocol is a like 50 ad-hoc extensions cobbled together. Because we must decode protocol frames in order to determine if a request is read or write, 90% of this is implementing a stream parser for the protocol.
Mercurial's own parser is simpler, but relies on blocking reads. Since we don't even have methods for blocking reads right now and keeping the whole thing non-blocking is conceptually better, I made the parser nonblocking. It ends up being a lot of stuff. I made an effort to cover it reasonably well with unit tests, and to make sure we fail closed (i.e., reject requests) if there are any parts of the protocol I got wrong.
A lot of the complexity is sharable with the HTTP stuff, so it ends up being not-so-bad, just very hard to verify by inspection as clearly correct.
Test Plan:
- Ran `hg clone` over SSH.
- Ran `hg fetch` over SSH.
- Ran `hg push` over SSH, to a read-only repo (error) and a read-write repo (success).
Reviewers: btrahan, asherkin
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7553
Summary: Ref T2230. Fixes T4079. As it turns out, this is Git being weird. See comments for some detials about what's going on here.
Test Plan: Created shallow and deep Git clones.
Reviewers: hach-que, btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4079, T2230
Differential Revision: https://secure.phabricator.com/D7554
Summary:
Ref T2230. In Git, we can determine if a command is read-only or read/write from the command itself, but this isn't the case in Mercurial or SVN.
For Mercurial and SVN, we need to proxy the protocol that's coming over the wire, look at each request from the client, and then check if it's a read or a write. To support this, provide a more flexible version of `passthruIO`.
The way this will work is:
- The SSH IO channel is wrapped in a `ProtocolChannel` which can parse the the incoming stream into message objects.
- The `willWriteCallback` will look at those messages and determine if they're reads or writes.
- If they're writes, it will check for write permission.
- If we're good to go, the message object is converted back into a byte stream and handed to the underlying command.
Test Plan: Executed `git clone`, `git clone --depth 3`, `git push` (against no-write repo, got error), `git push` (against valid repo).
Reviewers: btrahan
Reviewed By: btrahan
CC: hach-que, asherkin, aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7551
Summary:
Fixes T4067. The way `DiffusionCommitQuery` works prevents it from loading SVN identifiers in some cases without additional constraints, since "12345" might be an SVN revision 12345, or it might be the first 5 characters of a Git commit hash.
Introduce `withRepository()` as a shorthand for `withDefaultRepository()` + `withRepositoryIDs()`. This tells the query to:
- Only look in the given repository; and
- use the more liberal identifier resolution rules while doing so.
The practical impact this has is that blame tooltips in SVN work again. The other queries which are fixed here were never run in SVN (which doesn't have first-class branches or tags); I've cleaned them up only for completeness.
Test Plan:
- Viewed blame in SVN, saw information again instead of empty tooltip.
- Viewed brnaches/tags in Mercurial and Git.
{F79226}
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T4067
Differential Revision: https://secure.phabricator.com/D7523
Summary: Ref T2230. This is easily the worst thing I've had to write in a while. I'll leave some notes inline.
Test Plan: Ran `hg clone http://...` on a hosted repo. Ran `hg push` on the same. Changed sync'd both ways.
Reviewers: asherkin, btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7520
Summary: This is starting to get a bit sizable and it turns out Mercurial is sort of a beast, so split the VCS serve stuff into a separate controller.
Test Plan: Pushed and pulled an authenticated Git repository.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran, hach-que
Differential Revision: https://secure.phabricator.com/D7494
Summary: This is a little funky but fixes an issue with Git repos that are
non-bare needing "origin/" to resolve branches other than "master". Eventually
this should get cleaned up.
Test Plan: Reporting user verified this fixed their issue.
Auditors: btrahan
Summary:
Expands on D7488, which looks way better than the config checks. I'm leaving the config checks for now, but maybe we should just get rid of them? This advice is delivered in a far more timely way.
- Check for normal VCS binaries too.
- Link to `environment.append-paths`.
- Get rid of untranslated names (I think they're probably not too useful?)
Test Plan: See screenshots.
Reviewers: hach-que, btrahan
Reviewed By: hach-que
CC: aran
Differential Revision: https://secure.phabricator.com/D7495
Summary:
Currently if 'git-http-backend' is not on the PATH, there is no visible message to the user other than "info/refs: is this a valid git repository?" when trying to clone. This adds a setup check so that if there are any Git repositories in use, it will check for the existance of the "git-http-backend" binary in the PATH.
I believe this is shipped by default alongside the git package on most distros, but in some (such as OpenSUSE), this binary isn't on the PATH by default.
Test Plan: Removed `/usr/lib/git` from my `environment.append-paths` and saw the message appear. Added it back and the message went away.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Maniphest Tasks: T4050
Differential Revision: https://secure.phabricator.com/D7488
Summary:
Ref T1493. Diffusion has some garbagey behavior for things we can't resolve. Common cases are:
- Looking at a branch that doesn't exist.
- Looking at a repository with no branches.
- Looking at a commit that doesn't exist.
- Looking at an empty repository.
In these cases, we generally fatal unhelpfully. I want to untangle this mess.
This doesn't help much, but does clean things up a bit. We currently have two separate query paths, "stablecommitname" and "expandshortcommit". These are pretty much doing the same thing -- taking some ref like "master" or "default" or a tag name or part of a commit name, and turning it into a full commit name. Merge them into a single "resolverefs" method.
This simplifies the code a fair bit, and gives us better error messages. They still aren't great, but they're like this now:
Ref "7498aec194ecf2d333e0e2baddd9d5cdf922d7f1" is ambiguous or does not exist.
...instead of just:
ERR-INVALID-COMMIT
Test Plan: Looked at Git, Mercurial and Subversion repositories that were empty and non-empty. Looked at branches/heads. Tried to look at invalid commits. Looked at tags. All of this still works, and some behaviors are a bit better than they used to be.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T1493
Differential Revision: https://secure.phabricator.com/D7484
Summary: Fixes the junk I broke in D7484. Before that, tag content was a side effect of resolving the ref name. Now, fetch it explicitly in `diffusion.tagsquery`.
Test Plan: Looked at a tag, saw the annotation/message.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D7485
Summary: Ref T1493. Consolidate these a bit; they might need some more magic once we do `--noupdate` checkouts. Mostly just trying to clean up and centralize this code a bit.
Test Plan: Viewed and `bin/repository discover`'d Mercurial repos with and without any branches.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T1493
Differential Revision: https://secure.phabricator.com/D7480
Summary: Adds summary (description) and test plan icons to make these area's more unique and differentiated over general sections.
Test Plan: Test a diff, a commit, a task
Reviewers: epriestley, btrahan
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Differential Revision: https://secure.phabricator.com/D7493
Summary: This disables CSRF checking around the `$repository->writeStatusMessage` so that pushing changes over HTTP to Git repositories doesn't fail miserably.
Test Plan: Applied this fix and I could `git push` to hosted repositories again.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Maniphest Tasks: T4052
Differential Revision: https://secure.phabricator.com/D7490
Summary: This fixes an issue where Git authentication would always fail on an install with `policy.allow-public` set to false. This is because when public access is allowed, anonymous users can query the user list. However, when public access is not allowed, you have to be authenticated before you can read any of the user objects.
Test Plan:
Prior to this fix, I get:
```
james@james-laptop:~/git/8> git clone http://phabricator.local/diffusion/TEST/
Cloning into 'TEST'...
fatal: unable to access 'http://phabricator.local/diffusion/TEST/': The requested URL returned error: 403
```
when `policy.allow-public` is false. After this fix I get:
```
james@james-laptop:~/git/8> git clone http://phabricator.local/diffusion/TEST/
Cloning into 'TEST'...
remote: Counting objects: 102, done.
remote: Compressing objects: 100% (71/71), done.
remote: Total 102 (delta 6), reused 0 (delta 0)
Receiving objects: 100% (102/102), 9.89 KiB | 0 bytes/s, done.
Resolving deltas: 100% (6/6), done.
Checking connectivity... done
```
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley
CC: Korvin, epriestley, aran
Maniphest Tasks: T4049
Differential Revision: https://secure.phabricator.com/D7489
Summary:
Ref T2230. This will need some more refinement, but basically it adds a "Create" vs "Import" step before we go through the paged workflow.
- If you choose "Create", we skip the remote URI / auth stuff, and then set the "hosted" flag.
- If you choose "Import", we do what we do now.
Test Plan: Created and imported repos.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7475
Summary:
- Warn about "Read/Write" instead of disabling it, to prevent edits which mutate it after changing a hosted repository to an unhosted one.
- Warn about authenticated connections with HTTPS auth disabled, and link to the relevant setting.
- When "Autoclose" is disabled, show that "Autoclose Branches" won't have an effect.
- For hosted repositories, show the HTTP and SSH clone URIs.
- Make them easy to copy/paste.
- Link to credential management.
- Show if they're read-only.
- This could be a bit nicer-looking than it is.
Test Plan: Looked at repositories in a bunch of states and made various edits to them.
Reviewers: btrahan, chad
Reviewed By: btrahan
CC: chad, aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7471
Summary: Depends on D7642. This updates the authentication logic so that HTTP writes can be made to Git repositories hosted by Phabricator.
Test Plan: Set the policy to allow me to push and I was able to. Changed the policy to disallow push and I was no longer able to push.
Reviewers: #blessed_reviewers, hach-que
Reviewed By: hach-que
CC: Korvin, epriestley, aran
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7468
Summary: This allows users to set their HTTP access passwords via Diffusion interface.
Test Plan: Clicked the "Set HTTP Access Password" link, set a password and saw it appear in the DB.
Reviewers: #blessed_reviewers, hach-que, btrahan
Reviewed By: hach-que
CC: Korvin, epriestley, aran, jamesr
Maniphest Tasks: T2230
Differential Revision: https://secure.phabricator.com/D7462
Summary:
Ref T2350. Fixes T2231.
- Adds log flags around discovery.
- Adds message flags for "needs update". This is basically an out-of-band hint to the daemons that a repository should be pulled sooner than normal. We set the flag when users push a revision, and expose a Conduit method that `arc land` will be able to use.
Test Plan: See screenshots.
Reviewers: btrahan, chad
Reviewed By: btrahan
CC: aran
Maniphest Tasks: T2350, T2231
Differential Revision: https://secure.phabricator.com/D7467
Summary:
`RepositoryStatusMessage` is basically a key/value table associated with a repository that I'm using to let the daemons store the most recent event of a given type, so we can easily show it on the status dashboard. I think this will be a lot easier for users to figure out than digging through logfiles.
I'm also going to write the "this needs a pull" status here eventually, for reducing the time lapse between pushes and discovery.
- Add storage for these messages.
- Have the pull engine populate the INIT phase. I'll do the FETCH phase next.
- Update the status readout to show all the various states.
Test Plan: See screenshots.
Reviewers: btrahan
Reviewed By: btrahan
CC: aran
Differential Revision: https://secure.phabricator.com/D7461
