If resolving a merge conflict result in an empty changesets, we now properly
skip the changeset instead of crashing.
Original patch from Robert Collins <robertc@robertcollins.net>.
This will be used to exclude largefile candidates from the normal file matcher,
which will allow add and addremove dryruns to not print a file as both a normal
and a large file.
Core addremove prints the file relative to cwd only if patterns are provided to
the command. Core add always prints relative to cwd. Also, both methods print
the subrepo prefix when needed. The 'already a largefile' doesn't have an
analog in core, but follows the same rules for consistency.
Both cmdutil.remove() and scmutil.addremove() require verbose mode or an inexact
match to print the filename. Core addremove also prints the file relative to
cwd only if patterns are provided to the command. And finally, both methods
print the subrepo prefix when needed.
This is consistent with addlargefiles(), and will make it easier to get the
paths that are printed correct when recursing into subrepos or invoking from
outside the repository. It also now restricts the path that the addremove is
performed on if a path is given, as is done with normal files.
The repo.status() call needs to exclude clean files when performing an
addremove, because the addremove override method calling this used to pass the
list of files to delete, which caused the matcher to only consider those files
in building the status list. Now the matcher is restricted only to the extent
that the caller requested- usually directories if at all. There's no reason for
addremove to care about clean files anyway- we don't want them deleted.
The code for picking a merge ancestor when rebasing merges had a long and
incorrect comment.
The comment would perhaps have been fine as commit message but does
not make the code more readable or maintainable and is a bad
substitute for correct and readable code.
The correct essense of the comment is quite trivial: a merge of an ancestor of
the rebase destination and an 'outside' revision can be rebased as if it was a
linear change, using 'destination ancestor parent' as base and pretty much
ignoring the 'outside' revision.
The code path where the comment is placed is however also used for other kinds
of merge rebases. The comment is thus not really correct and not helpful. I
think it would be better to drop the comment and rewrite the code.
The more aggressive synchronization of lfdirstate that was backed out in
4fe80f20ab06 masked the problem where lfdirstate would hold an 'R' for a
largefile that was added and then removed without a commit between. We could
just conditionally call lfdirstate.drop() or lfdirstate.remove() here, but this
also properly updates lfdirstate if the standin doesn't exist for the file
somehow (i.e. call drop instead of remove).
Without this change, the precommit status in the commit command immediately
after the test change lists the removed (and never committed) largefile as 'R'.
It can also lead to situations where the status command reports the same, long
after the commit [1].
[1] http://www.selenic.com/pipermail/mercurial-devel/2015-January/065153.html
d20af8be6a14 introduced a significant performance regression: All largefiles
were marked 'normallookup' by linear (or noop) updates and had to be rehashed
by the next command.
The previous change introduced a different solution to the problem d20af8be6a14
solved and we can thus back it out again.
This is an alternative solution to the problem addressed by d20af8be6a14. This
implementation has the advantage that it doesn't mark clean largefiles as
normallookup. We can thus avoid repeated rehashing of all largefiles when
d20af8be6a14 is backed out.
This implementation use the existing 'lfmr' actions that 4f7f7352d9d0
introduced for handling another part of the same cases.
If an uncommitted and deleted file was forgotten, a warning would be emitted,
even though the operation was successful. See the previous patch for
'remove -A' for the exact circumstances, and details about the cause.
The bug report doesn't mention largefiles, but the given recipe doesn't fail
unless the largefiles extension is loaded. The problem only affected normal
files, whether or not any largefiles are committed, and only files that have
not been committed yet. (Files with an 'a' state are dropped from dirstate,
not marked removed.) Further, if the named normal file never existed, the
warning would be printed out twice.
The problem is that the core implementation of remove() calls repo.status(),
which eventually triggers a dirstate.walk(). When the file isn't seen in the
filesystem during the walk, the exception handling finds the file in
dirstate, so it doesn't complain. However, the largefiles implementation
called status() again with all of the original files (including the normal
ones, just dropped). This time, the exception handler doesn't find the file
in dirstate and does complain. This simply excludes the normal files from
the second repo.status() call, which the largefiles extension has no interest
is processing anyway.
This is a copy/paste (with the necessary tweaks) of the composenormalfilematcher
method currently on default, which does the inverse- this trims the normal files
out of the matcher. It will be used in the next patch.
This is in the way of passing a matcher to removelargefiles(). This method is
called in exactly two places- first in overrides.addremove() (but only if the
pattern list passed to it is not empty), and second in the commands.remove()
override. But since the latter calls commands.remove() first, which also does
this check, it isn't needed here.
To reduce amount of changes for review-ability, previous patch uses
"args" as argument name of "dodiff()", even though "args" includes
also the name of command to be executed (or full-path of it).
This patch replaces "args" by more appropriate name "cmdline".
Before this patch, all command line arguments for external tools are
quoted by the combination of "shlex.split" and "util.shellquote". But
this causes some problems.
- some problematic commands can't work correctly with quoted arguments
For example, 'WinMerge /r ....' is OK, but 'WinMerge "/r" ....' is
NG. See also below for detail about this problem.
https://bitbucket.org/tortoisehg/thg/issue/3978/
- quoting itself may change semantics of arguments
For example, when the environment variable CONCAT="foo bar baz':
- mydiff $CONCAT => mydiff foo bar baz (taking 3 arguments)
- mydiff "$CONCAT" => mydiff "foo bar baz" (taking only 1 argument)
For another example, single quoting (= "util.shellquote") on POSIX
environment prevents shells from expanding environment variables,
tilde, and so on:
- mydiff "$HOME" => mydiff /home/foobar
- mydiff '$HOME' => mydiff $HOME
- "shlex.split" can't handle some special characters correctly
It just splits specified command line by whitespaces.
For example, "echo foo;echo bar" is split into ["echo",
"foo;echo", "bar"].
On the other hand, if quoting itself is omitted, users can't specify
options including space characters with "--option" at runtime.
The root cause of this issue is that "shlex.split + util.shellquote"
combination loses whether users really want to quote each command line
elements or not, even though these can be quoted arbitrarily in
configurations.
To resolve this problem, this patch does:
- prevent configurations from being processed by "shlex.split" and
"util.shellquote"
only (possibly) "findexe"-ed or "findexternaltool"-ed command path
is "util.shellquote", because it may contain whitespaces.
- quote options specified by "--option" via command line at runtime
This patch also makes "dodiff()" take only one "args" argument instead
of "diffcmd" and "diffopts. It also omits applying "util.shellquote"
on "args", because "args" should be already stringified in "extdiff()"
and "mydiff()".
The last hunk for "test-extdiff.t" replaces two whitespaces by single
whitespace, because change of "' '.join()" logic causes omitting
redundant whitespaces.
Up until now we compared the "path" and "sharedpath" repository properties to
check if a repository is shared. This was relying an implementation detail of
shared repositories. In order to make it possible to change the way shared
repositories are implemented, we encapsulate this check into its own localrepo
method, called shared.
This new method returns None if the repository is shared, and something else
(for now a string describing the short of share) otherwise.
The reason why I did not call this method "isshared" and made it return a
boolean is that I plan to introduce a new type of shared repository soon.
# NOTES:
This is the first patch in a series whose purpose is to add support for
creating "full repository shares", which are repositories that share everything
with the repository source except their current revision, branch and bookmark.
This series is RFC because I am not very sure of some of the solutions I took.
Comments are welcome!
The comment about status being buggy with a repo proxy seems to be that status
wasn't being redirected to testing the largefiles themselves- lfstatus was
always False, and so the original status method was being called instead of
doing the largefiles processing. This is because when the various largefile
command overrides call 'repo.lfstatus = True', __setattr__() in repoview is in
turn setting the value on the unfiltered repo, and the value in the filtered
repo remains False.
Explicitly looking at the attribute on the unfiltered repo keeps all views in
sync.
There is no --after for addremove, so the printing for addremove can be hoisted
out of the 'not after' check. The difference between the two remove messages
reflects the existing difference between core remove and core addremove styles
for printing the file.
There are still some pre-existing issues here. Core addremove only prints on
inexact matches or when verbose. But since the largefiles that are being
removed are passed to removelargefiles() as a pattern list, there is never an
inexact match, which would keep the largefiles from being printed at all unless
verbose is specified. Therefore, the output is a little more aggressive than
core. The addremove print style here is also inconsistent with core- it should
use matcher.uipath(f) instead of f. These can be fixed once a matcher is passed
in.
The function only adds the hash content of the file to the set to upload if the
file in the ctx is a standin. It is called by overrides.summaryremotehook(),
which is called in the summary method. The largefiles extension switches
'lfstatus' on in summary, so the standins shouldn't be visible when obtaining a
context there.
The reason this wasn't noticed before is that the 'lfstatus' attribute is only
being set on the unfiltered repo because of how repoview delegates attribute
assignment. Therefore any filtered view will return a context containing
standins, whether or not 'lfstatus' was set in the various overrides methods.
That will be fixed in the next patch. But without this change, the next patch
would have test failures for 'summary --large' stating there are no files to
upload.
When a directory was renamed and a new untracked file was added in the
new directory and the remote directory added a file by the same name
in the old directory, the local untracked file gets overwritten, as
demonstrated by the broken test case in test-rename-dir-merge.
Fix by checking for unknown files for 'dg' actions too. Since
_checkunknownfile() currently expects the same filename in both
contexts, we need to add a new parameter for the remote filename to
it.
This simplifies largefiles' overridecalculateupdates(), which no
longer has to do the conversion it started doing in 478d610ca1b0
(largefiles: rewrite merge code using dictionary with entry per file,
2014-12-09).
To keep this patch small, we'll leave the name 'actionbyfile' in
overrides.py. It will be renamed in the next patch.
The hack for method monkey patching on repoview has been ruled out as
fragile, so we are rolling it back. We'll expand the explanation in the next
changeset.
The stored path contains platform specific separators, so splitting on '/.hg'
returned the string unmodified on Windows. The source was then looked for in
$source/.hg/.hg, which obviously fails. This caused cascading errors in
test-share.t relating to the recent bookmark support.
Merged files are considered modified at commit time even if only 1 parent
differs. In this case we must use the change context of this parent for
expansion.
The issue went unnoticed for long because it is only apparent until the next
update to the merge revision - except in test-keyword where it was always
staring us in the face.
This patch adds the -B/--bookmarks option to the share command added by the
share extension. All it does for now is create a marker, 'bookmarks.shared',
that will be used by future code to implement the sharing functionality.
Unicode and Python's unicode.splitlines() treat several extra legacy
ASCII codepoints as linebreaks, even though the vast bulk of computing
and Python's own str.splitlines() do not. Rather than introduce line
numbering confusion, we filter them out when highlighting.
The state mapping also contains some magic negative values (detached
parent, ignored revision). Blindly reading the state thus lead to
unfortunate usage of the negative value as an update destination. We
now filter them out.
We do a minor alteration of the test to catch this.
Mq tried to insert headers in the right order. Sometimes it would stop
searching before checking all headers and it could thus duplicate a header
instead of replacing it.
Changeset 5b64e22ecd8e introduced the examination of exec bit of
largefiles in "hg status --rev REV" case, but it doesn't avoid it on
the platform being unaware of exec-bit (e.g. on NTFS of Windows).
Fix inconsistent handling of plain header separation in mq patcheader - and
contrary to 05acc6157816, do it in the direction of having an empty line
between header and description. Plain patches are like mails and should thus
have an empty line between headers and body in compliance with RFC 822 3.1.
The new rebase revset didn't check for the case when there are no common
ancestors. Now it does. The new behavior should be the same as the pre-3.2
behavior. Added a test.