tests: more fixes for f

tests: teach f not to report symlink mode bits They're not meaningful or portable

tests: teach f not to report directory size It's not meaningful or portable

test-histedit-commute: call helper script with sh Buildbot pointed out that this test wasn't passing on Linux+vfat because there's no chmod for shell scripts.

test-tools: fix portability issues

progress: add a lock to prepare for introducing a thread

progress: move update check into helper method

progress: move current topic to member variable

progress: add try/finally to make the diffs for the next commit more readable No change in behavior.

transaction: include backup file in the "undo" transaction Once the transaction is closed, we now write transaction related data for possible future undo. For now, we only do it for full file "backup" because their were not handle at all in that case. In the future, we could move all the current logic to set undo up (that currently exists in localrepository) inside transaction itself, but it is not strictly requires to solve the current situation.

transaction: pass the name of the "undo" journal to the transaction It is time for the transaction to be responsible for setting up the undo data. It is necessary to move this logic into the transaction because many more files are handled now, and the transaction is the object tracking them all. The value can be set to None if no undo should be set.

rollback: have an empty entry for the vfsmap in rollback This empty string key is used for the store. This will be needed to properly rollback backup in a future changesets.

transaction: clarify the name of 'journal' argument for transaction The argument is a string containing the journal name (used as prefix for all other transaction file). This is not the transaction file itself. So we clarify this.

transaction: use 'util.copyfile' for creating backup Using 'copyfile' (single file) instead of 'copyfiles' (tree) will ensures destination file will be overwritten. This will prevent some abort if backup file are left in place for random reason. It also seems more correct.

copyfile: allow optional hardlinking Some code paths use 'copyfiles' (full tree) for a single file to take advantage of the best-effort-hard-linking parameter. We add similar parameter and logic to 'copyfile' (single file) for this purpose. The single file version have the advantage to overwrite the destination file if it exists.

repair: add experimental option to write bundle2 files This adds an experimental option 'strip-bundle2-version' which causes backup bundles to use bundle2 formatting. Especially for generaldelta repositories, this should provide significant performance gains for any operation that needs to write a backup.

changegroup.getsubset: support multiple versions Allow a version parameter to specify which version of the packer should be used

changegroup.writebundle: HG2Y support This diff adds support to writebundle to generate a bundle2 wrapper; upcoming diffs will add an option to write a v2 changegroup part instead of v1 in these bundles.

changegroup.writebundle: provide ui The next diff will add support for writing bundle2 files to writebundle, but the bundle2 generator wants access to a ui object. This changes the signature and callsites to pass one in.

test-module-imports: use test-repo requirement

largefiles: don't rehash largefiles in updatelfiles if standin hash changed Standins are read before and after an update/merge, and all the standins that changes are handed to updatelfiles for getting their corresponding largefiles updated. updatelfiles would then hash the largefile and see if it already matched the new expected hash. If so, it would skip the update. If different, the largefile would be updated. It would happen very rarely that the largefile happened to match the new hash (and thus not the old one) and the hashing would thus be pointless ... and hashing is not cheap. Instead, when it is known that the standin hash changed (from an update), just update the standin unconditionally. If the largefile was "unsure" before the update, it was hashed at that point, so we know there is nothing to preserve. (Also, the hashing in updatelfiles was not used to preserve changes, but only to be lazy about updating the largefile, so nothing is lost by not doing this extra hashing.) There might be rare situations where we now will update largefiles that didn't have to be updated, but in all relevant cases (?) this will improve performance. Updates on a repo with some big largefiles has been seen to go from 9.19 s to 6.8 s - that is 26% less painful.

largefiles: show progress when checking standin hashes in outgoing changesets This checking can take a huge amount of time and we should give user a hint that something is going on.

unbundle: support bundle2 files This adds support for bundle2 files to the unbundle command.

pullbundle2: extract addchangegroup result combining into its own function This will also be used for 'hg unbundle'

histedit: add a test to show that issue4251 is fixed (issue4251) This will help us not regress this case in the future.

commands.debugbundle: bundle2 support This enables debugbundle to print supporting info for bundle2 files.

largefiles: cleanup overrideadd() This was a remnant of the code prior to overridding cmdutil.add().

largefiles: enable subrepo support for add The --large, --normal and --lfsize args couldn't be passed to a subrepo before, and files in the subrepos would be added silently (if -v wasn't specified) as normal files. As an added bonus, 'hg add --dry-run' no longer prints that largefiles would also be added as normal files as well.

add: pass options via keyword args The largefiles extensions needs to be able to pass --large, --normal and --lfsize to subrepos via cmdutil.add() and hgsubrepo.add(). Rather than add additional special purpose arguments, stop extracting the existing args from the **opts passed to commands.add() and just pass them along.

largefiles: don't pop largefile-specific arguments to the add command The arguments will need to stay present when making add work with subrepos.

share: replace the bookmarks.shared file with an entry on a new "shared" file cd79fb4d75fd introduced a way to share bookmarks. When a repository share that shares bookmarks was created, a .hg/bookmarks.shared file was created to mark the repository share as one that shares its bookmarks. We have plans to introduce other levels of sharing, including a "full share" mode. Rather than creating a new ".shared" file for each new thing that we may want to share It seems better to create a single "shared" file that will list what is shared for a given shared repository. This should make it much easier to get a list of everything that is shared by a given shared repository. The shared file contains a list of shared "items" (such as bookmarks). Each shared "item" is added as a new line in the file. For now the only possible entry in the file is "bookmarks".

docker: support Fedora 21

rpm: make Python 2.7.9 the default Python to include in rpms for EL 5 Use the new and more TLS support in Python 2.7.9.

contrib: make Python 2.7.9 the default in Makefile.python We should utilize (and test) the big API changes and new TLS functionality in Python 2.7.9 whenever possible.

localrepo: remove all external users of localrepo.wopener This change touches every module in which repository.wopener was being used, and changes it for the equivalent repository.wvfs. It should now be possible to remove localrepo.wopener.

localrepo: remove all external users of localrepo.sopener This change touches every module in which repository.sopener was being used, and changes it for the equivalent repository.svfs. It should now be possible to remove localrepo.sopener.

localrepo: remove all external users of localrepo.opener This change touches every module in which repository.opener was being used, and changes it for the equivalent repository.vfs. This is meant to make it easier to split the repository.vfs into several separate vfs. It should now be possible to remove localrepo.opener.

log: use namespace logname and colorname Now that we have the machinary to change the log name and the color label used, let's use that. Tests have been updated accordingly.

namespaces: add colorname member to namespace object Previously, there was no way to change the color label used for 'hg log' output. This patch just adds the member to the object, a future patch will change 'hg log' to use this.

namespaces: add logname member to namespace object Previously, there was no way to change the name used for 'hg log' output. This was inconvenient for extensions that want a template name longer than 12 characters (e.g. remotebookmarks) but a different name for 'hg log'. This patch only adds the member to the object, a future patch will update the 'hg log' code.

namespaces: use named args for namespace api This is just a style change but makes adding new arguments more robust for callers.

namespaces: make the constructor into named args None of the arguments are truly optional but this makes adding future arguments more robust and perhaps optional.

dispatch: only check compatibility against major and minor versions (BC) Extensions can declare compatibility with Mercurial versions. If an error occurs, Mercurial will attempt to pin blame on an extension that isn't marked as compatible. While all bets are off when it comes to the internal API, my experience has shown that a monthly/patch release of Mercurial has never broken any of the extensions I've written. I think that expecting extensions to declare compatibility with every patch release of Mercurial is asking a bit much and adds little to no value. This patch changes the blame logic from exact version matching to only match on the major and minor Mercurial versions. This means that extensions only need to mark themselves as compatible with the major, quarterly releases and not the monthly ones in order to stay current and avoid what is almost certainly unfair blame. This will mean less work for extension authors and almost certainly fewer false positives in the blame attribution.

test-extension: use a realistic Mercurial version Mercurial doesn't define the 3rd "patch" version field for major releases. Don't use it in tests.

test-extension: improve test readability There was a mountain of text in this file that made reading the tests difficult. Inserting some line breaks greatly improves the situation.

bundle2: enforce parttype as alphanumerical The binary format description has always stated that the parttype should be simple, but it was never really enforced. Recent discussions have convinced me we want to keep the part type simple and easy to debug. There is enough extensibility in the rest of the format.

tests: conditionalize test-tools.t for Windows support I still get the following diff on Windows 7 with NTFS. I'm not sure if it is expected and we should wildcard the link count, or if something needs to be fixed. @@ -58,7 +58,7 @@ $ ln bar baz $ f bar -n baz -l --hexdump -t --sha1 --lines=9 -B 20 - bar: file, links=2, newer than baz, sha1=612ca68d0305c821750a + bar: file, links=0, newer than baz, sha1=612ca68d0305c821750a\r (esc) 0000: 31 0a 32 0a 33 0a 34 0a 35 0a 36 0a 37 0a 38 0a |1.2.3.4.5.6.7.8.| 0010: 39 0a |9.|

dirstate: ignore negative debug.dirstate.delaywrite values - they crashed it Sleep can only travel forward in time, not back.

linkrev: use the right manifest content when adjusting linrev (issue4499) When the manifest revision is stored as a delta against a non-parent revision, '_adjustlinkrev' could miss some file update because it was using the delta only. We now use the 'fastread' method that uses the delta only when it makes sense. A test showcasing on the of possible issue have been added.

unpacker: fix missing arg for py2.4

branchcache: make _rbcrevslen handling more safe self._rbcrevslen is used to keep track of the number of good records on disk. It should thus not be updated before the records actually have been written to disk.

branchcache: add debug output whenever cache files use truncate The cache files are usually append only but will automatically be truncated and recover in exceptional situations. Add a debug notice when such exceptional situations are encountered.

tests: rework revision branch cache tests The tests that were added with the revision branch cache in 678f53865c68 had suffered from bit rot in the development iterations. They were no longer that "good". Now, the tests are rewritten and reworked to be more readable and maintainable and relevant for the actual implementation. This also utilizes the new 'f' helper tool for keeping an eye the cache files.

tests: add 'f' tool for cross platform file operations in the tests This tool is like the collection of tools found in a unix environment but are cross platform and stable and suitable for our needs in the test suite. The main reason it is "needed" now is for hexdump of revision branch cache to keep an eye on how it changes and make sure the format is stable. It is a very generic tool that can end up being used a lot in tests, so I gave it very generic name.

run-tests: include testdir in $PATH so tests easily can use helper tools The testdir is already added to $PYTHONPATH - I think it makes sense and is convenient to add it to $PATH too. The following binaries are invoked from tests using full path with $TESTDIR/ - they can now be used without specifying path: dumbhttp.py dummyssh filterpyflakes.py generate-working-copy-states.py get-with-headers.py hghave histedit-helpers.sh killdaemons.py md5sum.py notcapable printenv.py readlink.py revlog-formatv0.py run-tests.py svn-safe-append.py svnxml.py tinyproxy.py

filelog: use censored revlog flag bit to quickly check if a node is censored

revlog: verify censored flag when hashing added revision fulltext When receiving a delta via exchange, three possible storage outcomes emerge: 1. The delta is added directly to the revlog. ("fast-path") 2. A freshly-computed delta with a different base is stored. 3. The new revision's fulltext is computed and stored outright. Both (2) and (3) require materializing the full text of the new revision by applying the delta to its base. This is typically followed by a hash check. The new flags argument allows callers to _addrevision to signal that they expect that hash check to fail. We can use this opportunity to verify that expectation. If the hash fails, require the flag be set; if the hash passes, require the flag be unset. Rather than simply eliding the hash check, this approach provides some assurance that the censored flag is not applied to valid revisions. Read more at: http://mercurial.selenic.com/wiki/CensorPlan

revlog: add flags argument to _addrevision, update callers use default flags For revlog index flags to be useful to other parts of Mercurial, they need to be settable when writing revisions. The current use case for revlog index flags is the censorship feature: http://mercurial.selenic.com/wiki/CensorPlan While the censor flag could be inferred in _addrevision by interrogating the text/delta being added, that would bury the censorship logic and inappropriately couple it to all revision creation.

revlog: define censored flag for revlogng index This flag bit will be used to cheaply signal censorship presence to upper layers (exchange, verify). It indicates that censorship metadata is present but does not attest to the verifiability of that metadata. For the censorship design, see: http://mercurial.selenic.com/wiki/CensorPlan

localrepo: remove all internal uses of localrepo.wopener It has been replaced with localrepo.wvfs.

localrepo: remove all internal uses of localrepo.sopener It has been replaced with localrepo.svfs.

localrepo: remove all internal uses of localrepo.opener It has been replaced with localrepo.vfs. In the future we may split the vfs into different vfs objects to access different elements of the repository.

sslutil: drop defunct ssl version constants Nobody outside sslutil should be using these constants anyway.

sslutil: use saner TLS settings on Python 2.7.9 Asking for TLSv1 locks us out of TLSv1_2 etc. This is at least less bad. Ideally we'd use ssl.create_default_context(), but that causes more mayhem in the testsuite than I really want to deal with right now.