view mercurial/narrowspec.py @ 39282:828a45233036

debugcommands: introduce debugrevlogindex (BC)

`hg debugindex` was originally invented for displaying revlog index
data and is highly tailored towards that. e.g. it accepts a --format
option to display index data for a particular revlog version and
displays things like offset and length.

As we support non-revlog storage, it makes sense for `hg debugindex`
to display generic index data and for there to exist storage-specific
or storage-aware debug* commands for dumping non-generic index data.

This commit effectively renames `hg debugindex` to
`hg debugrevlogindex` and replaces `hg debugindex` with a version that
is storage agnostic.

Tests using --format have been updated to use `hg debugrevlogindex`.

Output is backwards compatible. The replacement command uses the
formatter, which means output can be templatized.

At some point, we may want to tweak output (e.g. to add the revision
size). But I don't feel like taking a bigger BC break at the moment.

The renamed command/function had to be moved because check-code
enforces alphabetical ordering of commands in this file.

Differential Revision: https://phab.mercurial-scm.org/D4358

# narrowspec.py - methods for working with a narrow view of a repository
#
# Copyright 2017 Google, Inc.
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

from __future__ import absolute_import

import errno

from .i18n import _
from . import (
    error,
    match as matchmod,
    repository,
    sparse,
    util,
)

FILENAME = 'narrowspec'

def parseserverpatterns(text):
    """Parses the narrowspec format that's returned by the server."""
    includepats = set()
    excludepats = set()

    # We get one entry per line, in the format "<key> <value>".
    # It's OK for value to contain other spaces.
    for kp in (l.split(' ', 1) for l in text.splitlines()):
        if len(kp) != 2:
            raise error.Abort(_('Invalid narrowspec pattern line: "%s"') % kp)
        key = kp[0]
        pat = kp[1]
        if key == 'include':
            includepats.add(pat)
        elif key == 'exclude':
            excludepats.add(pat)
        else:
            raise error.Abort(_('Invalid key "%s" in server response') % key)

    return includepats, excludepats

def normalizesplitpattern(kind, pat):
    """Returns the normalized version of a pattern and kind.

    Returns a tuple with the normalized kind and normalized pattern.
    """
    pat = pat.rstrip('/')
    _validatepattern(pat)
    return kind, pat

def _numlines(s):
    """Returns the number of lines in s, including ending empty lines."""
    # We use splitlines because it is Unicode-friendly and thus Python 3
    # compatible. However, it does not count empty lines at the end, so trick
    # it by adding a character at the end.
    return len((s + 'x').splitlines())

def _validatepattern(pat):
    """Validates the pattern and aborts if it is invalid.

    Patterns are stored in the narrowspec as newline-separated
    POSIX-style bytestring paths. There's no escaping.
    """

    # We use newlines as separators in the narrowspec file, so don't allow them
    # in patterns.
    if _numlines(pat) > 1:
        raise error.Abort(_('newlines are not allowed in narrowspec paths'))

    components = pat.split('/')
    if '.' in components or '..' in components:
        raise error.Abort(_('"." and ".." are not allowed in narrowspec paths'))

def normalizepattern(pattern, defaultkind='path'):
    """Returns the normalized version of a text-format pattern.

    If the pattern has no kind, the default will be added.
    """
    kind, pat = matchmod._patsplit(pattern, defaultkind)
    return '%s:%s' % normalizesplitpattern(kind, pat)

def parsepatterns(pats):
    """Parses a list of patterns into a typed pattern set."""
    return set(normalizepattern(p) for p in pats)

def format(includes, excludes):
    output = '[include]\n'
    for i in sorted(includes - excludes):
        output += i + '\n'
    output += '[exclude]\n'
    for e in sorted(excludes):
        output += e + '\n'
    return output

def match(root, include=None, exclude=None):
    if not include:
        # Passing empty include and empty exclude to matchmod.match()
        # gives a matcher that matches everything, so explicitly use
        # the nevermatcher.
        return matchmod.never(root, '')
    return matchmod.match(root, '', [], include=include or [],
                          exclude=exclude or [])

def needsexpansion(includes):
    return [i for i in includes if i.startswith('include:')]

def load(repo):
    try:
        spec = repo.svfs.read(FILENAME)
    except IOError as e:
        # Treat "narrowspec does not exist" the same as "narrowspec file exists
        # and is empty".
        if e.errno == errno.ENOENT:
            return set(), set()
        raise
    # maybe we should care about the profiles returned too
    includepats, excludepats, profiles = sparse.parseconfig(repo.ui, spec,
                                                            'narrow')
    if profiles:
        raise error.Abort(_("including other spec files using '%include' is not"
                            " suported in narrowspec"))
    return includepats, excludepats

def save(repo, includepats, excludepats):
    spec = format(includepats, excludepats)
    repo.svfs.write(FILENAME, spec)

def savebackup(repo, backupname):
    if repository.NARROW_REQUIREMENT not in repo.requirements:
        return
    vfs = repo.vfs
    vfs.tryunlink(backupname)
    util.copyfile(repo.svfs.join(FILENAME), vfs.join(backupname), hardlink=True)

def restorebackup(repo, backupname):
    if repository.NARROW_REQUIREMENT not in repo.requirements:
        return
    util.rename(repo.vfs.join(backupname), repo.svfs.join(FILENAME))

def clearbackup(repo, backupname):
    if repository.NARROW_REQUIREMENT not in repo.requirements:
        return
    repo.vfs.unlink(backupname)

def restrictpatterns(req_includes, req_excludes, repo_includes, repo_excludes):
    r""" Restricts the patterns according to repo settings,
    results in a logical AND operation

    :param req_includes: requested includes
    :param req_excludes: requested excludes
    :param repo_includes: repo includes
    :param repo_excludes: repo excludes
    :return: include patterns, exclude patterns, and invalid include patterns.

    >>> restrictpatterns({'f1','f2'}, {}, ['f1'], [])
    (set(['f1']), {}, [])
    >>> restrictpatterns({'f1'}, {}, ['f1','f2'], [])
    (set(['f1']), {}, [])
    >>> restrictpatterns({'f1/fc1', 'f3/fc3'}, {}, ['f1','f2'], [])
    (set(['f1/fc1']), {}, [])
    >>> restrictpatterns({'f1_fc1'}, {}, ['f1','f2'], [])
    ([], set(['path:.']), [])
    >>> restrictpatterns({'f1/../f2/fc2'}, {}, ['f1','f2'], [])
    (set(['f2/fc2']), {}, [])
    >>> restrictpatterns({'f1/../f3/fc3'}, {}, ['f1','f2'], [])
    ([], set(['path:.']), [])
    >>> restrictpatterns({'f1/$non_exitent_var'}, {}, ['f1','f2'], [])
    (set(['f1/$non_exitent_var']), {}, [])
    """
    res_excludes = set(req_excludes)
    res_excludes.update(repo_excludes)
    invalid_includes = []
    if not req_includes:
        res_includes = set(repo_includes)
    elif 'path:.' not in repo_includes:
        res_includes = []
        for req_include in req_includes:
            req_include = util.expandpath(util.normpath(req_include))
            if req_include in repo_includes:
                res_includes.append(req_include)
                continue
            valid = False
            for repo_include in repo_includes:
                if req_include.startswith(repo_include + '/'):
                    valid = True
                    res_includes.append(req_include)
                    break
            if not valid:
                invalid_includes.append(req_include)
        if len(res_includes) == 0:
            res_excludes = {'path:.'}
        else:
            res_includes = set(res_includes)
    else:
        res_includes = set(req_includes)
    return res_includes, res_excludes, invalid_includes