context: set repo property in basectx It seems like a good practice to call the super constructor. Let's start by passing the repo along to basectx so it can assign it to a private attribute. We should perhaps pass the rev and node along as well, but that requires more work before it can be done. Differential Revision: https://phab.mercurial-scm.org/D2970

context: move reuse of context object to repo.__getitem__ (API) As an example of how weird the basectx.__new__ is: whenever you create a workingctx, basectx.__new__ gets called first. Since our __new__ has a "changeid" argument as second parameter, when create the workingctx(repo, text="blah"), the text gets bound to "changeid". Since a string isn't a basectx, our __new__ ends up not doing anything funny, but that's still very confusing code. Another case is metadataonlyctx.__new__(), which I think exists in order to prevent metadataonlyctx.__init__'s third argument (originalctx) from being interpreted as a changeid in basectx.__new__(), thereby getting reused. Let's move this to repo.__getitem__ instead, where it will be pretty obvious what the code does. After this patch, changectx(ctx) will be an error (it will fail when trying to see if it's a 20-byte string). Differential Revision: https://phab.mercurial-scm.org/D2969

memctx: create parent contexts using "repo[p]" syntax I want to reduce dependence on basectx.__new__() and move that code over to repo.__getitem__(). Differential Revision: https://phab.mercurial-scm.org/D2968

context: avoid using a context object as a changeid I find it misleading to pass changeid=changectx. It currently works to do that because there's weird (IMO) handling of it in basectx.__new__. I'm planning on removing that code. Passing changeid as "changeid" and context as "context" makes it more readable. Note that the documentation of filectx.__init__ doesn't even seem to be aware that a changeid can be a context ("changeset revision, node, or tag"). Differential Revision: https://phab.mercurial-scm.org/D2967

context: remove unwanted assignments in basectx.__new__() (API) The two subclasses in core apparently didn't trust __new__() to do the job anyway (they both reassigned all the properties after). Differential Revision: https://phab.mercurial-scm.org/D2966

streamclonebundle: make sure we accept new stream clone bundle spec When asked specifically, the code do a sanity check on the clone bundle to ensure it's a stream bundle. As we introduced a new stream bundle spec, update the logic to support it. With this final changeset, we can now announce safely a stream v2 clone bundle and old clients would not crash trying to process it. This changeset address the last comment from Gregory Szorc on the stream v2 bundle series. Differential Revision: https://phab.mercurial-scm.org/D1957

streamclonebundle: add a test for stream clone bundle v2 Differential Revision: https://phab.mercurial-scm.org/D1956

bundlespec: add support for some variants This way the stream v2 bundle spec can disable the changegroup part while enabling the stream v2 part. Differential Revision: https://phab.mercurial-scm.org/D1955

bundle: add the possibility to bundle a stream v2 part Differential Revision: https://phab.mercurial-scm.org/D1954

streambundlev2: add a new test-file Add the new test file in a separate changeset before supporting the new format so we better see the differences. Differential Revision: https://phab.mercurial-scm.org/D1953

bundlespec: move computing the bundle contentops in parsebundlespec We will introduce a new bundlespec for stream bundle which will influence the contentops. Differential Revision: https://phab.mercurial-scm.org/D1952

bundlespec: introduce an attr-based class for bundlespec We will add support of contentops in the next patch, introduce a class instead of returning a 4-items tuple. Differential Revision: https://phab.mercurial-scm.org/D2971

templater: factor out unwrapastype() from evalastype() So ParseError of unwrapastype() can be caught reliably.

templater: extract unwrapinteger() function from evalinteger()

templater: extract type conversion from evalfuncarg() Needed by the subsequent patches.

templater: drop bool support from evalastype() Future patches will split evalastype() into two functions, evalrawexp() and unwrapastype(), so we can catch the exception of type conversion. # evaluating function may bubble up inner ParseError thing = evalrawexp(context, mapping, arg) try: return unwrapastype(context, thing) except ParseError: # add hint and reraise However, evalboolean() can't be factored out in this way since it has to process boolean-like symbols as non keyword. Fortunately, it's unlikely that we'll need a general type converter supporting bool, so this patch drops it from the table.

templater: do not use stringify() to concatenate flattened template output

templateutil: reimplement stringify() using flatten()

templateutil: move flatten() from templater It's the same kind of utility as stringify().

stringutil: move person function from templatefilters Move the person function from template filters to the stringutil module, so it can be reused in the mailmap template function. Differential Revision: https://phab.mercurial-scm.org/D2960

stringutil: add isauthorwellformed function The regular expression for this function formerly lived at https://hg.mozilla.org/hgcustom/version-control-tools/file/tip/hghooks/mozhghooks/author_format.py#l13 Differential Revision: https://phab.mercurial-scm.org/D2959

test-lfs-test-server: add a testcase for `hg serve` I haven't figured out yet how to make the authentication checks work for a specific list of users, so the 'web.allow-push' list is wildcarded. (It appears that the client doesn't react to a 401 by sending authentication data, which may be caused in part by not having all of the headers in httpbasicauthhandler's http_error_auth_reqed(), compared to a run of test-http.t. But in any case, we should probably have a separate set of tests for various authentication scenarios. As it is, without the wildcard, no push access is granted.) There are several deviations from the `lfs-test-server` case: - `hg serve` emits a Server header. I think Gregory indicated that this isn't easily suppressed. - `hg serve` names the "basic" transfer handler in the Batch API response. Not having to specify it was for backwards compatability, so this seems like the right thing to do. (`lfs-test-server` doesn't name it, whether it was explicitly requested by the client or not.) - PUT status for a newly created file is 201, per RFC-2616 [1]. The Basic Transfer API [2] shows an example upload transcript with a 200 response. It doesn't make much sense to re-upload a file (unless it is corrupt) in an example, but I wouldn't be surprised if some other implementations also expect 200 because of this. But the RFC says MUST use 201 for creation. - The Content-Type for the file transfers is "application/octet-stream", like the sample transcript (though I don't see it explicitly called out in the text elsewhere). Using "text/plain" seems clearly wrong. - `lfs-test-server` isn't removing the action property and sending back an error code like the spec calls out when a file is missing or corrupt. Doing so on the `hg serve` side reveals a bug in our client code when handling the response- it indicates the remote file is missing instead of corrupt around line 452. I'll probably glob over the Content-Length differences once this settles down. Prior to the recent hgweb refactoring, the Batch API response was using chunked encodings instead. Back to the RFC, I have no idea if the python framework handles the "MUST NOT ignore any Content-* (e.g. Content-Range) headers that it does not understand or implement and MUST return a 501" for a PUT request. [1] https://www.w3.org/Protocols/rfc2616/rfc2616-sec9.html#sec9.6 [2] https://github.com/git-lfs/git-lfs/blob/master/docs/api/basic-transfers.md#uploads

push: use "repo['.']" instead of old form "repo['']" Note that this does not conflict with the commit message of my previous patch: I found this after writing the previous patch (besides, I very easily forget things). Differential Revision: https://phab.mercurial-scm.org/D2965

context: change default changeid from old form '' to '.' I don't think I've seen repo[''] in the codebase at least in the last two years. Differential Revision: https://phab.mercurial-scm.org/D2964

remove: add dry-run functionality

lfs: add support for serving blob files

lfs: add server side support for the Batch API

lfs: add basic routing for the server side wire protocol processing The recent hgweb refactoring yielded a clean point to wrap a function that could handle this, so I moved the routing for this out of the core. While not an hg wire protocol, this seems logically close enough. For now, these handlers do nothing other than check permissions. The protocol requires support for PUT requests, so that has been added to the core, and funnels into the same handler as GET and POST. The permission checking code was assuming that anything not checking 'pull' or None ops should be using POST. But that breaks the upload check if it checks 'push'. So I invented a new 'upload' permission, and used it to avoid the mandate to POST. A function wrap point could be added, but security code should probably stay grouped together. Given that anything not 'pull' or None was requiring POST, the comment on hgweb.common.permhooks is probably wrong- there is no 'read'. The rationale for the URIs is that the spec for the Batch API[1] defines the URL as the LFS server url + '/objects/batch'. The default git URLs are: Git remote: https://git-server.com/foo/bar LFS server: https://git-server.com/foo/bar.git/info/lfs Batch API: https://git-server.com/foo/bar.git/info/lfs/objects/batch '.git/' seems like it's not something a user would normally track. If we adhere to how git defines the URLs, then the hg-git extension should be able to talk to a git based server without any additional work. The URI for the transfer requests starts with '.hg/' to ensure that there are no conflicts with tracked files. Since these are handed out by the Batch API, we can change this at any point in the future. (Specifically, it might be a good idea to use something under the proposed /api/ namespace.) In any case, no files are stored at these locations in the repository directory. I started a new module for this because it seems like a good idea to keep all of the security sensitive server side code together. There's also an issue with `hg verify` in that it will want to download *all* blobs in order to run. Sadly, there's no way in the protocol to ask the server to verify the content of a blob it may have. (The verify action is for storing files on a 3rd party server, and then informing the LFS server when that completes.) So we may end up implementing a custom transfer adapter that simply indicates if the blobs are valid, and fall back to basic transfers for non-hg servers. In other words, this code is likely to get bigger before this is made non-experimental. [1] https://github.com/git-lfs/git-lfs/blob/master/docs/api/batch.md

test-lfs: drop trailing ', ' item separators from debug JSON output The trailing space looks weird when conditionalizing the line. The commas shouldn't be necessary because of the indenting. The `lfs-test-server` isn't sending all of the same items (notably, the "transfer" attribute is missing), so having the commas means more lines need to be conditionalized.

lfs: add a blob verification method to the local store A corrupt blob can be signaled through the Batch API response, without actually transferring the file. A true/false indicator is slightly easier than immediately catching an exception.