If the CVS repo somehow has a symbolic name that references a revision
consisting of a single number (e.g. BAD_TAG: 1), convert will fail when
attempting to find the branches, preventing the initial import from
working.
This patch skips those symbolic names--without warning.
For servers with branchmap support, the algorithm now works as follows:
1. A list of branches in outgoing changesets is created.
2. Using the remote branchmap, a check for new branches is performed.
3. A map (from branch to head list) of locally known remote heads before
the push is created, and one which, after step 4, will contain the locally
known remote heads after the push.
4. The post-push head map is updated with the outgoing changesets, using the
branch cache update mechanism.
5. A check for new heads is performed, by comparing the length of the head list
before and after push, for each branch. If there are new heads, an error
depending on whether or not there are incoming changes on the branch,
is returned.
6. If the push is allowed, a warning is written if there are incoming changes
on any branches involved in the push.
For old servers, an algorithm similar to step 4-6 above is used to check for
new topological heads only.
Two bugs are also fixed:
1. Sometimes you would be allowed to push new branch heads without --force.
A test for this case was added.
2. You would get the "note: unsynced remote changes!" warning if there were any
incoming changesets, even if they were on unrelated branches.
This bug happens if the filesystem doesn't support exec-bit, during merges,
for example in 0b01431fee25 on the hg repo.
If f is not in p1, but is in p2 and has the x-bit in p2, since the dirstate is
based on p1, and the FS doesn't support the exec-bit, the dirstate can't
"guess" the right bit.
We instead fix it in workingcontext.flags()/manifest.
Previously, it only returned revisions that were in the revlog when it
was originally opened; revisions added since then were invisible.
This broke revlog._partialmatch() and therefore repo.lookup().
(Credit to Benoit Boissinot for simplifying my original test script
and for the actual fix.)
This lets us change the threshold at which a *.d file will be split
out, which should make it much easier to construct test cases that
probe revlogs with a separate data file.
(issue2137)
The options parsing logic replaces the options.blacklist initial value,
a list of filenames, with a dict of test names to filenames. When
runchildren rebuilds the command line to launch child processes the dict
is flattened and provided as a malformed argument when a filename is
expected.
Fix is to remove the blacklist option from child invocations since the
filtering is handled in the parent anyway.
If you have a diff line that matches a header line, the patch splitter
currently breaks your patch at this line. For example a line like:
+key: value
This can lead to "malformed patch" exceptions. Now fixed.
If, on Windows, your repository and working copy are in a reparse point and you
use os.removedirs, you will remove non-empty reparse points, disabling the
reparse point. See @6b1369445b7b.
