ui: add support for a tweakdefaults knob We've been talking for years about a one-stop config knob to opt in to better behavior. There have been a lot of ideas thrown around, but they all seem to be too complicated to get anyone to actually do the work.. As such, this patch is the stupidest thing that can possibly work in the name of getting a good feature to users. Right now it's just three config settings that I think are generally uncontroversial, but I expect to add more soon. That will likely include adding new config knobs for the express purpose of adding them to tweakdefaults.

localrepo: remove unused addchangegroup() (API) This completes the cleanup started in d3775db748a0 (localrepo: move the addchangegroup method in changegroup module, 2014-04-01).

changegroup: rename "dh" to the clearer "deltaheads" We have a lot of frequently used abbreviations, but this is not one of them.

changegroup: rename "srccontent" to "cgnodes" It's the list of nodes in the incoming changegroup, so "cgnodes" made more sense to me.

revlog: rename list of nodes from "content" to "nodes" It seems like the reason for "content" is that the variable contains the nodes that the changegroup "contains", see e234eda20984 (revlog: make addgroup returns a list of node contained in the added source, 2012-01-13), but "nodes" seems much clearer.

revlog: delete obsolete comment The comment seems to refer to code that was deleted in 557988c691d1 (revlog.addgroup(): always use _addrevision() to add new revlog entries, 2010-10-08).

revlog: delete dead assignment in addgroup()

pycompat: move the queue related definitions below queue import This helps in understanding why empty and queue are there.

pycompat: move multiline comments above a function to function doc pycompat.py is unorganized and looks ugly. Next few patches will try to make it look more cleaner so that adding more code is easy and reading code also. This patch moves the multiline comments above functions to function docs. While moving, I improved the comments and make them better suitable for func doc. While I was here I drop a unrequired and misplaced comment.

py3: replace dict.iterkeys() with iter(dict) dict.iterkeys() does not exists on Python 3.

py3: explicitly convert dict.values() to a list on py3 dict.values() returns a dict_values() object, so we need to pass it into list() explicitly to get one.

py3: alias long to int and xrange to range in test-ancestor.py on Python 3

py3: add a new bytesurl() to convert a str url into bytes

py3: add a new strurl() which will convert a bytes url to str

killdaemons: fix WaitForSingleObject() error handling logic on Windows The error return is not 0 for this method, so _check() was doing nothing when an error occurred. This forces the error path, much like the check for OpenProcess(). The only unhandled return is now WAIT_ABANDONED, but I don't see how that could happen in this case.

killdaemons: explicitly set the ctypes signatures When I tried importing util.posixfile to work around removing a file opened by another process on Windows, it brought along the declarations in win32.py, which broke the error handling[1]. It doesn't seem worth hacking killdaemons[2] just to isolate these declarations in win32.py, so just declare them here to prevent any future issues. (win32.py mentions the declarations are required by pypy.) [1] https://www.mercurial-scm.org/pipermail/mercurial-devel/2017-May/097905.html [2] https://www.mercurial-scm.org/pipermail/mercurial-devel/2017-May/097907.html

hghave: enable 'serve' on Windows I've been using a local hghaveaddon.py to enable this for a couple of months with reasonable success, and 'killdaemons' is already enabled on Windows. There's one failure[1] in test-http-proxy.t that this adds, which I can't figure out. [1] https://www.mercurial-scm.org/pipermail/mercurial-devel/2017-April/096987.html

contrib: add a ratchet for tests in Python 3 This gives us an easy way to automatically update passing tests in Python 3.

contrib: check in a whitelist of passing tests in Python 3

tests: try and fail more gracefully with broken unicode escapes

md5sum: adapt for python 3 support

profiling: allow configuring minimum display threshold for hotpath statprof.display_hotpath() accepts a "limit" function to choose the minimum threshold for samples to display. The default is 0.05, which means you don't need individual items contributing less than 5%. I had a need to adjust this threshold. We already have a config option for it. So let's reuse it. check-config.py doesn't like having multiple defaults for the ui.configwith() calls. The behavior is obviously correct. I'm not sure if it is worth teaching check-config.py how to ignore this. So I've just accepted the new output.

config: document profiling.show{min,max}

check-config: look for ui.configwith We previously weren't looking for this config helper. And, surprise, profiling.py references config options without docs. If I tried hard enough, I could have combined the regexps using a positive lookbehind assertion or something. But I didn't want to make my brain explode. At some point, we should probably do this linting at the tokenizer or ast layer. I'm not willing to open that can of worms right now.

check-config: use named groups in regexp In preparation for making this regexp a bit more complicated.

check-config: use compiled regexp And split the regexp across multiple lines to make it easier to read.

tests: add test coverage for check-config We didn't have explicit test coverage before. I broke check-config.py as part of writing patches and was lucky I realized it.

exchange: use context manager for bundle1 unbundling The lazy locking is not used for bundle1, so using a regular context manager is clearer.

unbundle: use context manager for transaction

clonebundle: use context managers for lock and transaction

changeset_templater: render template specified by templatespec tuple

formatter: render template specified by templatespec tuple

formatter: put topic in templatespec tuple This will allow us to change the initial template reference depending on how the template is looked up. For example, -Tdefault => (ref='changeset', tmpl=None, mapfile='map-cmdline.default') -T'{rev}' => (ref='', tmpl='{rev}', mapfile=None) A literal template given by -T option will be stored as an unnamed template, which will free up the template namespace so that we can load named templates from [templates] section of user config.

cmdutil: pass templatespec tuple directly to changeset_templater (API) A fewer number of arguments should be better.

formatter: wrap (tmpl, mapfile) by named tuple I'm going to add more options to the templatespec tuple. cmdutil.logtemplatespec() is just an alias now, but it will be changed to a factory function later.

cmdutil: factor out helper to create changeset_templater with literal template changeset_templater has lots of arguments, but most callers only need to specify a literal template 'tmpl'. "hg debugtemplate" has no diff option, which means 'opts' were effectively {}, so dropped opts.

test-dirstate-race: stabilize for Windows Without quoting, the output after "custom merge tool" in the next test is an abort trying to open "C:UsersMattAppData..."

formatter: document lookuptemplate()

cmdutil: rename gettemplate() to _lookuplogtemplate() This function is only useful when processing log options.

formatter: inline gettemplater() Since it's highly use-case dependent how template should be looked up, gettemplater() function isn't useful. Keeping it would introduce another bug I've made and fixed earlier in this series.

formatter: factor out function to create templater from literal or map file (tmpl, mapfile) will be packed into a named tuple later.

graphlog: do not look up graphnodetemplate in template paths (BC) ui.graphnodetemplate config should be a literal template as ui.logtemplate is. The use of formatter.gettemplater() is valid only for a template string given by -T/--template option.

formatter: open raw template file in posix semantics This should have no effect, but seems good for code consistency.

formatter: open raw template file in binary mode (BC) I believe it was a mistake to open file in text mode. Now '\r' is preserved on Windows, but it should be okay to print '\r\n' as long as users live in such platform.

formatter: close raw template file explicitly

largefiles: avoid a crash when archiving a subrepo with largefiles disabled This path is also used for extdiff, which is how I crossed paths with it. Without this, an AttributeError occurs looking for 'lfstatus' on localrepository. See also d414c28db84d. The other archive method is for the archival.py override, so it doesn't need to be special cased like this. (It looks like it is only called for the top level repo.) Likewise, the transplant override is also for commands.py. The other overrides set lfstatus before examining it.

patchbomb: make getaddrs function easier to work with Prior to this the return value was potentially None, a string, or a list of strings. It now always returns a list of strings where each string is always only one email address

patchbomb: look for non-empty publicurl, not a non-None one Otherwise it's impossible to turn this feature back off, which is making writing of tests awkward.

patchbomb: make variable name for publicurl always be publicurl

bisect: simpler approach for option validation message Yuya Nishihara gave this suggestion on the mailing list after the previous patch was queued, and honestly this seems much simpler and looks more efficient.

pushkey: use False/True for return values from push functions It was particularly unclear in phases.pushphase() whether the 0/1 returned were the 0/1 for public/draft phase or for False/True

smartset: micro optimize spanset.slice() to narrow range accordingly -1ms isn't significant, but seems not bad. revset #0: limit(0:9999, 100, 9000) 6) 0.001145 7) 0.000214 revset #3: last(0:9999, 100) 6) 0.000197 7) 0.000171

smartset: micro optimize baseset.slice() to use slice of list -1ms isn't significant, but seems not bad. revset #1: limit(0::9999, 100, 9000) 5) 0.001681 6) 0.000790

smartset: extract method to slice abstractsmartset Sub classes can provide optimized implementations.

smartset: extract spanset factory to make it constructed without a repo This renames the spanset class to _spanset, and moves its __init__ to new spanset() function. spanset() is now a factory function. This allows us to construct a spanset without keeping a repo instance.

smartset: change repr of spanset to show revisions as half-open range Before, an empty spanset was displayed as '<spanset+ 0:-1>', which seemed confusing.

fsmonitor: don't write out state if identity has changed (issue5581) Inspired by the dirstate fix in dc7efa2826e4, this should fix any race conditions with the fsmonitor state changing from underneath. Since we now grab the wlock for any non-invalidate writes, the only situation this appears to happen in is with a concurrent invalidation. Test that.

fsmonitor: write state with wlock held and dirstate unchanged (issue5581) This means that the state will not be written if: (1) either the wlock can't be obtained (2) something else came along and changed the dirstate while we were in the middle of a status run.

workingctx: add a way for extensions to run code at status fixup time Some extensions like fsmonitor need to run code after dirstate.status is called, but while the wlock is held. The extensions could grab the wlock again, but that has its own peculiar race issues. For example, fsmonitor would not like its state to be written out if the dirstate has changed underneath (see issue5581 for what can go wrong in that sort of case). To protect against these sorts of issues, allow extensions to declare that they would like to run some code to run at fixup time. fsmonitor will switch to using this in the next patch in the series.

workingctx: also pass status tuple into poststatusfixup fsmonitor is going to need this to compute its set of notable files to persist.