Use the pager given by the environment to display long output Unix systems usually have a PAGER environment variable set. If it is set, mercurial will use the pager application to display output. Two configuration variables are available to influence the behaviour of the pager. ui.pager sets the pager application. The pager is only used if ui.usepager is true. By default ui.usepager is disabled.

Fix thinko in test-inherit-mode If we don't change any rwx bit in the last test, hg will skip the calls to chmod since it'll assume they're not needed. This might fix things on BSD systems.

test-serve: replace copy/paste with shell function

revert: update state of files in the "checkout" list This can make a difference when there are filters involved and decode(encode(working-dir-data)) != working-dir-data even though encode(decode(repo-data)) == repo-data An example is a working dir file that uses only \n when you're using the win32text extension.

normallookup: during merges, restore the state saved by remove

dirstate.remove: during merges, remember the previous file state We encode the previous state as a negative file size (AFAICS, previous versions of hg always have size == 0 when state == 'r'). We save the state of 'm'erged and dirty files, because they're the two states that indicate that a file has to be committed on a merge to correctly record per-file history.

debugstate: add --nodates This can be useful for tests.

patch: fix corner case with update + copy patch handling (issue 937) The self patching of files when diffed with a backup is a bit peculiar to me. It makes sense in mpatch, that's less clear in mercurial patching code. Let's document and preserve it for now.

Merge

Backed out changeset 4b81eecc8aa2

Merge

Backed out changeset 91ac1726730a

Automated merge with http://hg.intevation.org/mercurial/crew

issue 1003: send all data properly

Remove unneeded "import os.path" again.

Spacing cleanup

filectx.ancestor: use fctx._repopath to cache filelogs (issue1035) _repopath may be different from _path for workingfilectx when there are renames involved.

Change User-agent field in client to have a lot more info.

Allow hgwebdir collections to follow symlinks.

Revert cfebb175704f, fixing the output of 2 tests

test-mv-cp-st-diff: avoid linkrev collisions The copy detection code relies heavily on the linkrevs, so it's better to avoid collisions in this test.

convert: avoid querying log of foreign svn branches Server configuration may prevent access to other branches, so don't even try to get them, instead of ignoring them afterwards.

patch: check filename is /dev/null for creation or deletion (issue 1033) Otherwise, file creation or deletion is ambiguous with file being filled or emptied.

update tests for copy detection changes

copies: add tests for status -C/diff --git This goes through 42 combinations of copy/rename branch/parent/working dir.

copies: deal with symmetric difference for linear relationships choose the minimum of symmetric difference and the two revs in question

status: find copies and renames beyond the working directory

diff: use copy smarts from copies.py

copies: move findcopies code to its own module - pass in contexts - fold symmetricdifference check into copies.copies

merge: move symmetricdifferences to ancestor.py

merge: simplify some helpers

merge: more simplifications to checkcopies

merge: simplify checkcopies

merge: privatize some functions, unnest some others

merge: notice uncommited copies (issue1000) As mentioned in msg5349 in issue988, "hg update" doesn't take into account uncommitted copies. To reproduce: ---------------- hg init repo cd repo touch foo hg ci -Am 'add foo' echo >> foo hg ci -m 'change foo' hg up -C 0 hg mv foo bar HGMERGE=false thg --debug -y update -------------------------- A similar problem happens with hg merge --force. I'm attaching a possible patch.

Check for patches repo with os.path.isdir not os.path.exists

hook.py: fix redirections introduced by 323b9c55b328 The only non-obvious part is the use of sys.{__stderr__,__stdout__}, which is needed because sshserver overrides sys.stdout. This makes a test that I added back in revision 7939c71f3132 ineffective.

Issue937: error messages from hooks not sent over HTTP. Turns out that stderr - where ui.warn would send messages - was not being proxied over the HTTP connection. stdout was, and it seems you need both. (The streams are interleaved for readability.) Tested on Ubuntu 7.10 with lighttpd on hgweb.cgi with HTTP Basic auth, no SSL, using a changeset failing win32text.forbidcrlf.

revlog.py: remove extra close() atomictempfile.rename() already calls close().

adjust test-http output to non-default HGPORT, e.g. with run-tests.py -j

hgweb: clarify which address and port can/cannot be bound at startup (bug 769) The error message at startup when the address/port could not be bound was confusing: hg serve abort: cannot start server: Address already in use Be more explicit: $ hg serve -a localhost abort: cannot start server at 'localhost:8000': Address already in use Also be more explicit on success, showing hostname and ip address/port: $ hg -v serve -a localhost -p 80 listening at http://localhost/ (127.0.0.1:80) We are careful to handle a missconfigured machine whose hostname does not resolve, falling back to the address given at the command line. Remove a dead-code error message.

revlog: make sure the files are closed after an exception happens This prevents a corruption when the writes happen after the truncate initiated by the transaction rollback.

merge with crew

make sure not to reuse an existing append-file from a previous failed pull

Fix issue995 (copy --after and symlinks pointing to a directory) I haven't looked at other places that call os.path.isdir to make sure they're OK.

dirstate: ignore mode changes if the fs does not supports the exec bit This can make a difference when e.g. the repo is exported through NFS (which support exec bits) and CIFS (which does not).

merge: require --force when there are deleted files

add a test for a375ffc2aa1b

localrepo.commit: normalize commit message even for rawcommit. This normalization consists of: - stripping trailing whitespace - always using "\n" as the line separator I think the main reason rawcommit was skipping this normalization was an attempt to preserve hashes during an hg->hg conversion. While this is a nice goal, it's not particularly interesting in practice. Since SHA-1 is so strong, the only safe way to do it is to have absolutely identical revisions. But: - if the original revision was created with a recent version of hg, the commit message will be the same, with or without that normalization - if it was created with an ancient version of hg that didn't do any normalization, even if the commit message is identical, the file list in the changelog is likely to be different (e.g. no removed files), and there were some old issues with e.g. extra file merging, which will end up changing the hash anyway - in any case, if one *really* has to preserve hashes, it's easier (and faster) to fake a partial conversion using something like: hg clone -U -r rev orig-repo new-repo hg -R new-repo log --template '#node# #node#\n' > new-repo/.hg/shamap Additionally, we've had some reports of problems arising from this lack of normalization - e.g. issue871, and a user that was wondering why hg export/hg import was not preserving hashes when there was nothing unusual going on (it was just import doing the normalization that had been skipped). This also means that it's even more unlikely to get identical revisions when going $VCS->hg->$VCS.

debugancestor: use repo.lookup when no revlog was specified

update output of test-hgweb

setup.py: use a simplified custom version of CCompiler.has_function The original one doesn't remove its temporary files and even creates a temporary file in the CWD.

gitweb: show branches in most of the templates

hgweb_mod: add branch helper functions to use in templates

Automated merge with http://hg.intevation.org/mercurial/crew

win32text: use util.binary to detect \0

adjust test-fetch output to non-default HGPORT, e.g. with run-tests.py -j

setup.py: skip inotify if there's no inotify_add_watch Debian Etch doesn't include a sys/inotify.h header, which makes it impossible to compile _inotify.c, making hg uninstallable. The cc.has_function() method is implemented by trying to compile a simple C program. Since there's no redirection involved all error messages are sent to the terminal. This is not particularly pretty but at least it allows the installation to complete.

run-tests.py: add a summary of failed tests at the end

fix incorrect date when committing a tag regression introduced by 989467e8e3a9

merge: fix handling of deleted files

setup.py: os.uname is not available on Windows

Automated merge with http://hg.intevation.org/mercurial/crew

Add inotify extension

convert: rename MAPFILE into REVMAP to disambiguate with filemap

Make mercurial easy installable

Add tests for the fixes to issue1014 (fractional timezones)

test-fetch failed due to non-zero exit code

Hide URL passwords in hg paths output.

Merge with main

Fixed bash completion for filenames containing spaces.

merge with main

dates: Fix bare times to be relative to "today"

dates: improve timezone handling datestr: - add format specifiers %1 and %2 for timezone hours and minutes - remove timezone and timezone format options - correctly find timezone hours and minutes for fractional and negative timezones - update users strdate: - correctly find timezone hours and minutes for fractional and negative timezones

revlog: report node and file when lookup fails

Merge with crew

fetch: don't proceed if working directory is missing files (issue988)

fetch: rename --force-editor option to --edit, for consistency

dates: fix fractional timezone display rounding bug

churn: show comitter email addresses unclipped (bug 1023) Email addresses were being clipped at a hard-coded 20 chars. Measure the length of all email addresses and accommodate the longest.

gitweb: use {url} as entrypoint in notfound.tmpl instead of /

Support web.baseurl in hgwebdir, overriding SCRIPT_NAME

Fixed typo in tag help, found by John Coomes

hgweb: fix test results missed by 6218:345a23eca8f6

gitweb: remove some line breaks in archive and rss columns. Clean up archive formatting a little in the index page.

Avoid importing mercurial.node/mercurial.repo stuff from mercurial.hg

Add missing import of sys

convert: Remove unused clone method

convert: Fix unbound name error in the subversion backend

imerge: Fix unbound name error and add a test case

Remove unused imports

Expand import * to allow Pyflakes to find problems

Removed trailing spaces from everything except test output

Replaced tab in run-tests.py

tests: add --keep-tmp to run-tests.py to debug test environment When set, you can find the working dir of the test run as a uniquely named subdirectory of --tmpdir so you can inspect it.

fetch: make test reproducible, tiny code cleanup

fetch: switch the default parent used for a merge This treats newly pulled changes as authoritative, and local changes as the "satellite" changes. The prior default behaviour is still available, via the --switch-parent option.

mq: test for qclone of http repo+queue tests f857eac30cd5 and 0c2b443fb3c3

When failing to load an extension, show where Hg tried to load it from.

hgweb/annotate: handle binary files like hgweb/file

test-status: Make assert function definition sh compatible.

Extend/correct acc40572da5b regarding -qA and ignored files. hg status -qA will now hide untracked files as described in the doc string.

'hg status -q' output skips non-tracked files. The '-q' flag was ignored in status command. But this flag can be used to hide non-tracked files in hg status output. This small correction makes status command more general, similar to 'svn status', where '-q' flag has the same effect. The '-u' and '-A' flags have priority over '-q'. A testcase and doc-string for status was extended to cover '-q' flag.

highlight: forgot import

highlight: guess by text when path name is ambiguous Although the docs claim that guess_lexer_for_filename will fall back to using file data, my ubuntu 7.10 pygments library seems to ignore the text argument. So call guess_lexer explicitly on failure.

highlight: make div trimmer work when lexer fails to identify text. In this case, there is no initial <span> tag.

highlight: use iter() instead of generator comprehension

mq: remove unimplemented option -f for strip

highlight: bail out if file is binary

highlight: support annotate, and reduce layering violations.

Introduce templateopts and logopts to reduce duplicate option definitions.

Add option -l/--limit to hg incoming and hg outgoing.

Move finding/checking the log limit to cmdutil

Make hg debugancestor accept -R by making it an optionalrepo command.

debugancestor: use *args instead of *opts, to not confuse with option dicts.

mq: Cleanup: update outdated file header.

convert: Clean up authormap key=value splitting. Introduces a subtle parsing difference: dstauthor can now contain '=' characters.

convert: Avoid redundant newline on authormap errors. The extra rstrip is necessary in case the file has uses different line terminators (i.e. CRLF).

convert: Ignore empty lines in authormap file.

Add config option to disable putting .hg_archival.txt inside archives.

add a blame alias for annotate The original cvs command was called blame. SVN has both, and other VC systems typically also have this as an alias. Makes things easier for converts.

warning on push should mention heads, not branches This makes the concept easier to relate to similar messages given on pull, and allows for a distinction between heads and (named) branches.

commands: lock() the repo while unbundling (issue1004)

patch: fix iterhunks() with trailing binary file removal Like some renames or copy operations, binary file removal does not generate any "file" or "hunk" action, but was not tagged as such and let iterhunk() assume no hunk was applied for the deleted file.

debugancestor: make the index argument optional, defaulting to 00changelog.i

Renamed "manifest" to "files" in web templates. manifest.tmpl is still used, so people having their own templates don't have to change them. "cmd=manifest" still works, new style URLs are not affected, because they already used "/file/".

Removed old-style hgweb templates

add explicit extension disabling syntax to hgrc.5

Permit XML entities to be escaped in template output. Useful for creating XML documents directly from Hg logging. Can also be used for HTML. For use in content, will escape '&', '<', and for completeness '>' (although it is not strictly necessary). For use in attributes, will also escape ' and ". Will also replace nonprinting (ASCII) control characters with spaces, since these are illegal in XML.

convert: add shallow, single branch svn conversions via svn.startrev

convert: allow svn trunk/branches/tags detection to be skipped - eluding convert.svn.branches defaults to "branches" - convert.svn.branches= disables branches detection - convert.svn.branches=/ is equivalent to former convert.svn.branches=

Add --all option to bundle command

Fix test-convert.out after 55455556f921

convert: improve subversion source documentation

hgweb: fix parameter mixup (issue1001)

tests: easier hg serve error diagnosis

util: make walkrepos() return .hg/patches if present

Allow "file://localhost/" in addition to "file:///" (issue728) file://localhost/ URLs are generated by Mac OS X API calls like [NSURL fileURLWithPath:]. Imported patch from Paul Bx.

mq: make qclone ask remote source repo for qbase using lookup protocol

Document log date ranges and mention 'hg help dates' for all commands (issue998)

Add test case for backout on named branches (issue665)

Reverse the way backout is doing the merge Currently, backout is creating a backout revision as a child node of the backed out node and will leave you at this new head. This has several drawbacks: * this changes the current head * when there is a long history between the backed out node and the current head, this will generate a huge number of diffs that are scary at first sight, and not very natural to review before commit. The change consists to switch back to the original node as soon as the backout node (which becomes the new tip) has been created. Then the --merge option can just merge this new tip in the current node. * the current head/node is not changed from the user's point of view * even without using the --merge option, the backout revision is still easy to locate, as this is the tip * the merge is much more intuitive as diffs of the merge is right you are looking to backout

tests: hide the name of the branch cache file

test-mq-pull-from-bundle: don't print the contents of $HGRCPATH

repo.status: also compare flags for files in the lookup list. We might be able to do something smarter about this in dirstate.status for files in normallookup state, but that would require some extra care to keep backwards compatibility.

keyword: remove "identify" and "remove" from nokwcommands These two commands care about the list of modified files returned by repo.status and we may need to do a full content comparison to populate that list.

test-glog: pass the list of files to commit test-glog uses debugsetparents instead of update+merge to create some funky DAGs, and so the dirstate contents won't be consistent with the checked out revision. Passing an explicit list of files to commit reduces a bit the dependency on the dirstate. Using a non-deprecated rawcommit might be better here.

hgweb: no i18n in protocol responses

improve changegroup.readbundle(), use it in hgweb

Also search for .hgrc if mercurial.ini not found on windows

hgweb: use bundletypes from mercurial.changegroup

Merge with crew

Infer a --repository argument from command arguments when reasonable. In particular: if invoked without -R from a CWD not inside a repo, having been passed one or more file paths as command arguments, where the nearest enclosing repo of all of those paths is the same, quietly infer a -R option for that repo. Otherwise abort with an error message as before.

hgweb: separate protocol calls from interface calls (issue996) The protocol functions are already pretty careful about not raising exceptions to the caller, and have their own error handling. We can formalize this a little bit to make it clearer (before, the exception handlers for a limited number of exceptions coming from the interface bits would blow up because some variables aren't instantiated for the protocol calls).

update output of test-convert

repair.py: rewrite a loop, making it cleaner and faster

Speed up hg grep by avoiding useless manifest parsing In the kernel repo (tip = 2b89f7111b96), a "hg grep mpm MAINTAINERS" goes from ~165s to 0.7s. This could get even a bit faster if we broke out of the loop after the first match, but I'm not sure how that would interact with the --follow code. This is obviously an extreme example, but other cases should also benefit from this patch.

cmdutil.walkchangerevs: use '-1:0' instead ot 'tip:0' This avoids a possibly expensive tag lookup.

revlog.revision: avoid opening the datafile without need. If there's no inline data, revlog.revision opens the data file every time it's called. This is useful if we're going to call chunk many times, but, if we're going to call it only once, it's better to let chunk open the file - if we're lucky, all the data we're going to need is already cached and we won't need to even look at the file.

convert: document splicemap, allow setting of multiple parents

merge backout

Backed out changeset b913d3aacddc (see issue971/msg5317)

Simplify utils.walkrepos().

Fix bad behaviour when specifying an invalid date (issue700) commit (aborts _after_ typing in a commit message) backout (aborted after the initial revert) tag (edited .hgtags and couldn't commit) import (patch applied, then commit fails) qnew (aborts on bad dates, but writes any valid date into the # Date header) qrefresh (like qnew) sign (like tag) fetch (merge, merge, merge, merge, abort)

Report filenames with disallowed characters as suggested by Mika Eloranta (see issue671)

hgweb: Quote filenames when downloading raw files.

hgweb: Pass only filename instead of full path when downloading raw files. Before this patch "filename=foo/bar" was sent and e.g. Firefox offered to save the file as "foo-bar" instead of just "bar".

merge with crew

Make annotae/grep print short dates with -q/--quiet. Move shortdate() from templatefilters to util to avoid code duplication.

Add hg grep -d/--date to list the commit date of matched revisions. Suggested by Greg Lindahl

util: always define a dummy lookup_reg()

convert: improve cycles detection message

Merge with crew-stable

bundlerepo: fix inconsistency of parsed and internal name (issue #821)

bundlerepo: test exposing inconsistency of parsed and internal name (issue #821)

Simplify the output of test-bheads Also remove unneeded -d and -u commit options.

merge with crew

Make hgweb annotate link to target line numbers (issue623)

fix test output after f7f25f58693a

merged Edward Lee's line anchors patch

Add line anchors to annotate, changeset, diff, file views for hgweb

automatically update the branch cache when tip changes

move the reading of branch.cache from _branchtags to branchtags

localrepo: keep the UTF-8 version of branchcache around

localrepo.invalidate: invalidate branchcache

keyword: do not inspect node1 for diff if node2 is given

keyword: add test for patch.diff monkey

keyword: make main class and hg command accessible Switch from global vars to top level dictionary. Goal: make it easier for external tools (like tortoisehg) to hook into keyword extension.

keyword: move expand/shrink decisions into kwtemplater - do not clobber wrapper code - get rid of clumsy kwtemplater attributes

try to fix test-inherit-mode on HFS+ HFS+ doesn't like setgid directories, so avoid them in most of the test and do a quick check that doesn't change the output when everything's OK.

cmdutil.commit: use relative paths in the error messages This is more consistent with other error messages and requires less makeup in test-commit and test-symlink-basic.

util.pathto: return '.' instead of an empty string This could happen with something like pathto(repo.root, 'foo/bar', 'foo/bar')

status: put added files that have disappeared in the deleted list This gives the user an indication that something went wrong with this file: $ hg add foo $ rm foo $ hg status foo ! foo Fixes issue212.

revert: unify forget and remove lists This doesn't make a difference right now, but after the next revision some files in state 'a' may end up in the deleted list, and revert won't be able to just remove all files in that list.

revert: only call dirstate.normal when we know the file is clean

commands.revert: don't call hg.revert commands.revert calculates everything that has to be done and then calls hg.revert to checkout and remove files. Unfortunately, hg.revert has to recalculate everything and that can take a long while, since it always operates on the whole working dir. Changing commands.revert to manually checkout and remove files makes things considerably faster, especially if we're reverting a single file in a repo with a huge number of files. This should be enough to close issue857.

revert: don't let repo.status walk the whole working dir Problem diagnosed by pmezard.

Set bundle.mainreporoot only after checking that it's a local repo Fixes issue977.

Fix test-hgweb-commands.out after 9b817c0a82f1

Propagating changes back to working dirs when changing files in external

test-convert-baz: set username with baz, not tla

Define tla username locally in baz and tla tests Suggested by Alexis S. L. Carvalho.

convert: fix --datesort ordering Two branches a and b starting at root, with commits interleaved like: root a0 a1 b0 a2 a3 b1 were converted in the following order: root a0 b0 a1 b1 a2 a3 Replace depth based toposort with a more classic traversal method.

[RFC] convert: fix --datesort The problem is that previously commit.date was used for sorting, but it's a string like "1 Jan xxx 2007", so it it wrong to use it for sorting. Another problem is that why we are using depth for sorting -- I have no clear answer -- it seems to be plain wrong. This patch is just an RFC.

push: exit(1) if "push creates new remote branches!"

hgweb: Fix stray quote in logo img tag

hgweb: add new logos

Add the new "droplet" logo

cat --decode: Drop short option, use opts.get() instead of opts[] - hg archive --no-decode has no short option, too, and maybe both could use -d in the future to select revisions by date. - opts.get makes python scripts calling cat() happy, because they don't have to pass the new option.

Option --decode for hg cat to apply decode filters. (Does not trigger expansion of keyword arguments when using keyword extension, though that might be desirable as well.)

keyword: monkeypatch patch.diff for non-interactive diffs Avoid clobbering diffs with expanded keywords eg. for notify extension.

merge with main

merge with crew-stable

clone: make things work when source is a repo object

clone: expand the path before saving it in .hg/hgrc This should fix issue976

better handle errors with date parsing (issue983) Windows cannot always handle a 1970-01-01 parameter to time.mktime().

better handle errors with date parsing (issue983) Windows cannot always handle a 1970-01-01 parameter to time.mktime().

hgweb: disable cgitb by default in hgwebdir.fcgi too

Document and fix tla/baz conversion tests

convert: detect baz before tla The former seems to handle repositories built by the latter, the opposite is not true.

keyword: add glog to nokwcommands

keyword: enable all monkey patches using _kwtemplater at reposetup - no need for try-except block - underscore prefix for private methods - add doc strings

hgweb: disable cgitb by default

convert: added GNU Arch (tla) tests and related fixes

convert: added gnu arch (baz) tests

convert: try to be smarter about CVS branching Better handles this case: The output from cvsps -A -u --cvs-direct -q: --------------------- PatchSet 1 Date: 2008/02/08 20:33:28 Author: fk Branch: HEAD Tag: (none) Log: initial Members: file_one:INITIAL->1.1 --------------------- PatchSet 2 Date: 2008/02/08 20:33:32 Author: fk Branch: branch_name Ancestor branch: HEAD Tag: (none) Log: new file on branch Members: file_two:1.1->1.1.2.1

filemerge: more backwards compatible behavior for ui.merge if ui.merge matches a tool in [merge-tools], it is used as the default tool selection but may be rejected by binary or gui or other checks. If ui.merge does not match a merge tool, it is used unconditionally.

filemerge: add 'checkchanged' merge tool property

ignore: split up huge patterns Some versions of Python silently mishandle large regexes, so split them up at a conservative 20k.

Use chmod 0770 and g+s instead of 02770 which does't work on Solaris 8.

keyword: no expansion in web diffs Expansion in hgweb view of changesets and diffs is not needed and only distracting. Expansion stays enable in file and archive requests where it makes sense.

keyword: make restrict entirely a kwtemplater attribute

keyword: declare globals in a more compact way

keyword: avoid os import by using util.splitpath

Automated merge with http://hg.west.spy.net/hg/hacks/hg

Option to log to only show changesets within a specified branch.

Strip filter name from command before passing to filter function. The new registration of in-process data filters (introduced in f8ad3b76e923 & 11af38a592ae) failed to correctly strip the filter name from its arguments before passing the "command" to the filter function. Thus a registration such as [decode] *.gz = compress: -9 would result in the associated filter function being called with the argument 'compress: -9' rather than just '-9' as expected.

make the journal/undo files from transactions inherit the mode from .hg/store

add test-inherit-mode

hghave: detect unix-style permissions By "unix-style" I mean: - user/group/other permissions - umask determines original permissions