hgext/fastannotate/formatter.py

phases: rework the logic of _pushdiscoveryphase to bound complexity

This rework the various graph traversal in _pushdiscoveryphase to keep the
complexity in check.

This is done though a couple of things:

- first, limiting the space we have to explore, for example, if we are not in
  publishing push, we don't need to consider remote draft roots that are also
  draft locally, as there is nothing to be moved there.

- avoid unbounded descendant computation, and use the faster "rev between"
  computation.

This provide a massive boost to performance when exchanging with repository with
a massive amount of draft, like mozilla-try:

### data-env-vars.name                             = mozilla-try-2023-03-22-zstd-sparse-revlog
  # benchmark.name                                 = hg.command.push
  # bin-env-vars.hg.flavor                         = default
  # bin-env-vars.hg.py-re2-module                  = default
  # benchmark.variants.explicit-rev                = all-out-heads
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.protocol                    = ssh
  # benchmark.variants.reuse-external-delta-parent = default

 ## benchmark.variants.revs                        = any-1-extra-rev
before: 20.346590 seconds
after:  11.232059 seconds (-38.15%, -7.48 seconds)
 ## benchmark.variants.revs                        = any-100-extra-rev
before: 24.752051 seconds
after:  15.367412 seconds (-37.91%, -9.38 seconds)


After this changes, the push operation is still quite too slow. Some of this
can be attributed to general phases slowness (reading all the roots from disk
for example) and other know slowness (not using persistent-nodemap, branchmap,
tags, etc. We are also working on them, but with this series, phase discovery
during push no longer showing up in profile and this is a pretty nice and bit
low-hanging fruit out of the way.

### (same case as the above)
  # benchmark.variants.revs                        = any-1-extra-rev
pre-%ln-change: 44.235070
this-changeset: 11.232059 seconds (-74.61%, -33.00 seconds)
  # benchmark.variants.revs                        = any-100-extra-rev
pre-%ln-change: 49.234697
this-changeset: 15.367412 seconds (-68.79%, -33.87 seconds)


Note that with this change, the `hg push` performance is now much closer to the
`hg pull` performance, even it still lagging behind a bit. (and the overall
performance are still too slow).

### data-env-vars.name                           = mozilla-try-2023-03-22-ds2-pnm
  # benchmark.variants.explicit-rev              = all-out-heads
  # benchmark.variants.issue6528                 = disabled
  # benchmark.variants.protocol                  = ssh
  # benchmark.variants.pulled-delta-reuse-policy = default
  # bin-env-vars.hg.flavor                       = rust
 ## benchmark.variants.revs                      = any-1-extra-rev
hg.command.pull: 6.517450
hg.command.push: 11.219888
 ## benchmark.variants.revs                      = any-100-extra-rev
hg.command.pull: 10.160991
hg.command.push: 14.251107

### data-env-vars.name                           = mozilla-try-2023-03-22-zstd-sparse-revlog
  # bin-env-vars.hg.py-re2-module                = default
  # benchmark.variants.explicit-rev              = all-out-heads
  # benchmark.variants.issue6528                 = disabled
  # benchmark.variants.protocol                  = ssh
  # benchmark.variants.pulled-delta-reuse-policy = default
 ## bin-env-vars.hg.flavor                       = default
 ## benchmark.variants.revs                      = any-1-extra-rev
hg.command.pull: 8.577772
hg.command.push: 11.232059
 ## bin-env-vars.hg.flavor                       = default
 ## benchmark.variants.revs                      = any-100-extra-rev
hg.command.pull: 13.152976
hg.command.push: 15.367412
 ## bin-env-vars.hg.flavor                       = rust
 ## benchmark.variants.revs                      = any-1-extra-rev
hg.command.pull: 8.731982
hg.command.push: 11.178751
 ## bin-env-vars.hg.flavor                       = rust
 ## benchmark.variants.revs                      = any-100-extra-rev
hg.command.pull: 13.184236
hg.command.push: 15.620843

# Copyright 2016-present Facebook. All Rights Reserved.
#
# format: defines the format used to output annotate result
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

from mercurial.node import (
    hex,
    short,
)
from mercurial import (
    encoding,
    pycompat,
    templatefilters,
    util,
)
from mercurial.utils import dateutil

# imitating mercurial.commands.annotate, not using the vanilla formatter since
# the data structures are a bit different, and we have some fast paths.
class defaultformatter:
    """the default formatter that does leftpad and support some common flags"""

    def __init__(self, ui, repo, opts):
        self.ui = ui
        self.opts = opts

        if ui.quiet:
            datefunc = dateutil.shortdate
        else:
            datefunc = dateutil.datestr
        datefunc = util.cachefunc(datefunc)
        getctx = util.cachefunc(lambda x: repo[x[0]])
        hexfunc = self._hexfunc

        # special handling working copy "changeset" and "rev" functions
        if self.opts.get(b'rev') == b'wdir()':
            orig = hexfunc
            hexfunc = lambda x: None if x is None else orig(x)
            wnode = hexfunc(repo[b'.'].node()) + b'+'
            wrev = b'%d' % repo[b'.'].rev()
            wrevpad = b''
            if not opts.get(b'changeset'):  # only show + if changeset is hidden
                wrev += b'+'
                wrevpad = b' '
            revenc = lambda x: wrev if x is None else (b'%d' % x) + wrevpad

            def csetenc(x):
                if x is None:
                    return wnode
                return pycompat.bytestr(x) + b' '

        else:
            revenc = csetenc = pycompat.bytestr

        # opt name, separator, raw value (for json/plain), encoder (for plain)
        opmap = [
            (b'user', b' ', lambda x: getctx(x).user(), ui.shortuser),
            (b'number', b' ', lambda x: getctx(x).rev(), revenc),
            (b'changeset', b' ', lambda x: hexfunc(x[0]), csetenc),
            (b'date', b' ', lambda x: getctx(x).date(), datefunc),
            (b'file', b' ', lambda x: x[2], pycompat.bytestr),
            (b'line_number', b':', lambda x: x[1] + 1, pycompat.bytestr),
        ]
        fieldnamemap = {b'number': b'rev', b'changeset': b'node'}
        funcmap = [
            (get, sep, fieldnamemap.get(op, op), enc)
            for op, sep, get, enc in opmap
            if opts.get(op)
        ]
        # no separator for first column
        funcmap[0] = list(funcmap[0])
        funcmap[0][1] = b''
        self.funcmap = funcmap

    def write(self, annotatedresult, lines=None, existinglines=None):
        """(annotateresult, [str], set([rev, linenum])) -> None. write output.
        annotateresult can be [(node, linenum, path)], or [(node, linenum)]
        """
        pieces = []  # [[str]]
        maxwidths = []  # [int]

        # calculate padding
        for f, sep, name, enc in self.funcmap:
            l = [enc(f(x)) for x in annotatedresult]
            pieces.append(l)
            if name in [b'node', b'date']:  # node and date has fixed size
                l = l[:1]
            widths = pycompat.maplist(encoding.colwidth, set(l))
            maxwidth = max(widths) if widths else 0
            maxwidths.append(maxwidth)

        # buffered output
        result = b''
        for i in range(len(annotatedresult)):
            for j, p in enumerate(pieces):
                sep = self.funcmap[j][1]
                padding = b' ' * (maxwidths[j] - len(p[i]))
                result += sep + padding + p[i]
            if lines:
                if existinglines is None:
                    result += b': ' + lines[i]
                else:  # extra formatting showing whether a line exists
                    key = (annotatedresult[i][0], annotatedresult[i][1])
                    if key in existinglines:
                        result += b':  ' + lines[i]
                    else:
                        result += b': ' + self.ui.label(
                            b'-' + lines[i], b'diff.deleted'
                        )

            if result[-1:] != b'\n':
                result += b'\n'

        self.ui.write(result)

    @util.propertycache
    def _hexfunc(self):
        if self.ui.debugflag or self.opts.get(b'long_hash'):
            return hex
        else:
            return short

    def end(self):
        pass


class jsonformatter(defaultformatter):
    def __init__(self, ui, repo, opts):
        super(jsonformatter, self).__init__(ui, repo, opts)
        self.ui.write(b'[')
        self.needcomma = False

    def write(self, annotatedresult, lines=None, existinglines=None):
        if annotatedresult:
            self._writecomma()

        pieces = [
            (name, pycompat.maplist(f, annotatedresult))
            for f, sep, name, enc in self.funcmap
        ]
        if lines is not None:
            pieces.append((b'line', lines))
        pieces.sort()

        seps = [b','] * len(pieces[:-1]) + [b'']

        result = b''
        lasti = len(annotatedresult) - 1
        for i in range(len(annotatedresult)):
            result += b'\n {\n'
            for j, p in enumerate(pieces):
                k, vs = p
                result += b'  "%s": %s%s\n' % (
                    k,
                    templatefilters.json(vs[i], paranoid=False),
                    seps[j],
                )
            result += b' }%s' % (b'' if i == lasti else b',')
        if lasti >= 0:
            self.needcomma = True

        self.ui.write(result)

    def _writecomma(self):
        if self.needcomma:
            self.ui.write(b',')
            self.needcomma = False

    @util.propertycache
    def _hexfunc(self):
        return hex

    def end(self):
        self.ui.write(b'\n]\n')