view mercurial/manifest.py @ 24545:9e0c67e84896

json: implement {tags} template

Tags is pretty easy to implement. Let's start there.

The output is slightly different from `hg tags -Tjson`. For reference,
the CLI has the following output:

  [
   {
    "node": "e2049974f9a23176c2addb61d8f5b86e0d620490",
    "rev": 29880,
    "tag": "tip",
    "type": ""
   },
   ...
  ]

Our output has the format:

  {
    "node": "0aeb19ea57a6d223bacddda3871cb78f24b06510",
    "tags": [
      {
        "node": "e2049974f9a23176c2addb61d8f5b86e0d620490",
        "tag": "tag1",
        "date": [1427775457.0, 25200]
      },
      ...
    ]
  }

"rev" is omitted because it isn't a reliable identifier. We shouldn't
be exposing them in web APIs and giving the impression it remotely
resembles a stable identifier. Perhaps we could one day hide this behind
a config option (it might be useful to expose when running servers
locally).

The "type" of the tag isn't defined because this information isn't yet
exposed to the hgweb templater (it could be in a follow-up) and because
it is questionable whether different types should be exposed at all.
(Should the web interface really be exposing "local" tags?)

We use an object for the outer type instead of Array for a few reasons.
First, it is extensible. If we ever need to throw more global properties
into the output, we can do that without breaking backwards compatibility
(property additions should be backwards compatible). Second, uniformity
in web APIs is nice. Having everything return objects seems much saner than
a mix of array and object. Third, there are security issues with arrays
in older browsers. The JSON web services world almost never uses arrays
as the main type for this reason.

Another possibly controversial part about this patch is how dates are
defined. While JSON has a Date type, it is based on the JavaScript Date
type, which is widely considered a pile of garbage. It is a non-starter
for this reason.

Many of Mercurial's built-in date filters drop seconds resolution. So
that's a non-starter as well, since we want the API to be lossless where
possible. rfc3339date, rfc822date, isodatesec, and date are all lossless.
However, they each require the client to perform string parsing on top of
JSON decoding. While date parsing libraries are pretty ubiquitous, some
languages don't have them out of the box. However, pretty much every
programming language can deal with UNIX timestamps (which are just
integers or floats). So, we choose to use Mercurial's internal date
representation, which in JSON is modeled as float seconds since UNIX
epoch and an integer timezone offset from UTC (keep in mind
JavaScript/JSON models all "Numbers" as double prevision floating point
numbers, so there isn't a difference between ints and floats in JSON).

# manifest.py - manifest revision class for mercurial
#
# Copyright 2005-2007 Matt Mackall <[e-mail]>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.

from i18n import _
import mdiff, parsers, error, revlog, util, scmutil
import array, struct

propertycache = util.propertycache

def _parse(data):
    """Generates (path, node, flags) tuples from a manifest text"""
    # This method does a little bit of excessive-looking
    # precondition checking. This is so that the behavior of this
    # class exactly matches its C counterpart to try and help
    # prevent surprise breakage for anyone that develops against
    # the pure version.
    if data and data[-1] != '\n':
        raise ValueError('Manifest did not end in a newline.')
    prev = None
    for l in data.splitlines():
        if prev is not None and prev > l:
            raise ValueError('Manifest lines not in sorted order.')
        prev = l
        f, n = l.split('\0')
        if len(n) > 40:
            yield f, revlog.bin(n[:40]), n[40:]
        else:
            yield f, revlog.bin(n), ''

def _text(it):
    """Given an iterator over (path, node, flags) tuples, returns a manifest
    text"""
    files = []
    lines = []
    _hex = revlog.hex
    for f, n, fl in it:
        files.append(f)
        # if this is changed to support newlines in filenames,
        # be sure to check the templates/ dir again (especially *-raw.tmpl)
        lines.append("%s\0%s%s\n" % (f, _hex(n), fl))

    _checkforbidden(files)
    return ''.join(lines)

class _lazymanifest(dict):
    """This is the pure implementation of lazymanifest.

    It has not been optimized *at all* and is not lazy.
    """

    def __init__(self, data):
        dict.__init__(self)
        for f, n, fl in _parse(data):
            self[f] = n, fl

    def __setitem__(self, k, v):
        node, flag = v
        assert node is not None
        if len(node) > 21:
            node = node[:21] # match c implementation behavior
        dict.__setitem__(self, k, (node, flag))

    def __iter__(self):
        return iter(sorted(dict.keys(self)))

    def iterkeys(self):
        return iter(sorted(dict.keys(self)))

    def iterentries(self):
        return ((f, e[0], e[1]) for f, e in sorted(self.iteritems()))

    def copy(self):
        c = _lazymanifest('')
        c.update(self)
        return c

    def diff(self, m2, clean=False):
        '''Finds changes between the current manifest and m2.'''
        diff = {}

        for fn, e1 in self.iteritems():
            if fn not in m2:
                diff[fn] = e1, (None, '')
            else:
                e2 = m2[fn]
                if e1 != e2:
                    diff[fn] = e1, e2
                elif clean:
                    diff[fn] = None

        for fn, e2 in m2.iteritems():
            if fn not in self:
                diff[fn] = (None, ''), e2

        return diff

    def filtercopy(self, filterfn):
        c = _lazymanifest('')
        for f, n, fl in self.iterentries():
            if filterfn(f):
                c[f] = n, fl
        return c

    def text(self):
        """Get the full data of this manifest as a bytestring."""
        return _text(self.iterentries())

try:
    _lazymanifest = parsers.lazymanifest
except AttributeError:
    pass

class manifestdict(object):
    def __init__(self, data=''):
        self._lm = _lazymanifest(data)

    def __getitem__(self, key):
        return self._lm[key][0]

    def find(self, key):
        return self._lm[key]

    def __len__(self):
        return len(self._lm)

    def __setitem__(self, key, node):
        self._lm[key] = node, self.flags(key, '')

    def __contains__(self, key):
        return key in self._lm

    def __delitem__(self, key):
        del self._lm[key]

    def __iter__(self):
        return self._lm.__iter__()

    def iterkeys(self):
        return self._lm.iterkeys()

    def keys(self):
        return list(self.iterkeys())

    def _intersectfiles(self, files):
        '''make a new lazymanifest with the intersection of self with files

        The algorithm assumes that files is much smaller than self.'''
        ret = manifestdict()
        lm = self._lm
        for fn in files:
            if fn in lm:
                ret._lm[fn] = self._lm[fn]
        return ret

    def filesnotin(self, m2):
        '''Set of files in this manifest that are not in the other'''
        files = set(self)
        files.difference_update(m2)
        return files

    @propertycache
    def _dirs(self):
        return scmutil.dirs(self)

    def dirs(self):
        return self._dirs

    def hasdir(self, dir):
        return dir in self._dirs

    def matches(self, match):
        '''generate a new manifest filtered by the match argument'''
        if match.always():
            return self.copy()

        files = match.files()
        if (len(files) < 100 and (match.isexact() or
            (not match.anypats() and util.all(fn in self for fn in files)))):
            return self._intersectfiles(files)

        lm = manifestdict('')
        lm._lm = self._lm.filtercopy(match)
        return lm

    def diff(self, m2, clean=False):
        '''Finds changes between the current manifest and m2.

        Args:
          m2: the manifest to which this manifest should be compared.
          clean: if true, include files unchanged between these manifests
                 with a None value in the returned dictionary.

        The result is returned as a dict with filename as key and
        values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the
        nodeid in the current/other manifest and fl1/fl2 is the flag
        in the current/other manifest. Where the file does not exist,
        the nodeid will be None and the flags will be the empty
        string.
        '''
        return self._lm.diff(m2._lm, clean)

    def setflag(self, key, flag):
        self._lm[key] = self[key], flag

    def get(self, key, default=None):
        try:
            return self._lm[key][0]
        except KeyError:
            return default

    def flags(self, key, default=''):
        try:
            return self._lm[key][1]
        except KeyError:
            return default

    def copy(self):
        c = manifestdict('')
        c._lm = self._lm.copy()
        return c

    def iteritems(self):
        return (x[:2] for x in self._lm.iterentries())

    def text(self):
        return self._lm.text()

    def fastdelta(self, base, changes):
        """Given a base manifest text as an array.array and a list of changes
        relative to that text, compute a delta that can be used by revlog.
        """
        delta = []
        dstart = None
        dend = None
        dline = [""]
        start = 0
        # zero copy representation of base as a buffer
        addbuf = util.buffer(base)

        # start with a readonly loop that finds the offset of
        # each line and creates the deltas
        for f, todelete in changes:
            # bs will either be the index of the item or the insert point
            start, end = _msearch(addbuf, f, start)
            if not todelete:
                h, fl = self._lm[f]
                l = "%s\0%s%s\n" % (f, revlog.hex(h), fl)
            else:
                if start == end:
                    # item we want to delete was not found, error out
                    raise AssertionError(
                            _("failed to remove %s from manifest") % f)
                l = ""
            if dstart is not None and dstart <= start and dend >= start:
                if dend < end:
                    dend = end
                if l:
                    dline.append(l)
            else:
                if dstart is not None:
                    delta.append([dstart, dend, "".join(dline)])
                dstart = start
                dend = end
                dline = [l]

        if dstart is not None:
            delta.append([dstart, dend, "".join(dline)])
        # apply the delta to the base, and get a delta for addrevision
        deltatext, arraytext = _addlistdelta(base, delta)
        return arraytext, deltatext

def _msearch(m, s, lo=0, hi=None):
    '''return a tuple (start, end) that says where to find s within m.

    If the string is found m[start:end] are the line containing
    that string.  If start == end the string was not found and
    they indicate the proper sorted insertion point.

    m should be a buffer or a string
    s is a string'''
    def advance(i, c):
        while i < lenm and m[i] != c:
            i += 1
        return i
    if not s:
        return (lo, lo)
    lenm = len(m)
    if not hi:
        hi = lenm
    while lo < hi:
        mid = (lo + hi) // 2
        start = mid
        while start > 0 and m[start - 1] != '\n':
            start -= 1
        end = advance(start, '\0')
        if m[start:end] < s:
            # we know that after the null there are 40 bytes of sha1
            # this translates to the bisect lo = mid + 1
            lo = advance(end + 40, '\n') + 1
        else:
            # this translates to the bisect hi = mid
            hi = start
    end = advance(lo, '\0')
    found = m[lo:end]
    if s == found:
        # we know that after the null there are 40 bytes of sha1
        end = advance(end + 40, '\n')
        return (lo, end + 1)
    else:
        return (lo, lo)

def _checkforbidden(l):
    """Check filenames for illegal characters."""
    for f in l:
        if '\n' in f or '\r' in f:
            raise error.RevlogError(
                _("'\\n' and '\\r' disallowed in filenames: %r") % f)


# apply the changes collected during the bisect loop to our addlist
# return a delta suitable for addrevision
def _addlistdelta(addlist, x):
    # for large addlist arrays, building a new array is cheaper
    # than repeatedly modifying the existing one
    currentposition = 0
    newaddlist = array.array('c')

    for start, end, content in x:
        newaddlist += addlist[currentposition:start]
        if content:
            newaddlist += array.array('c', content)

        currentposition = end

    newaddlist += addlist[currentposition:]

    deltatext = "".join(struct.pack(">lll", start, end, len(content))
                   + content for start, end, content in x)
    return deltatext, newaddlist

def _splittopdir(f):
    if '/' in f:
        dir, subpath = f.split('/', 1)
        return dir + '/', subpath
    else:
        return '', f

class treemanifest(object):
    def __init__(self, dir='', text=''):
        self._dir = dir
        self._dirs = {}
        # Using _lazymanifest here is a little slower than plain old dicts
        self._files = {}
        self._flags = {}
        for f, n, fl in _parse(text):
            self[f] = n
            if fl:
                self.setflag(f, fl)

    def _subpath(self, path):
        return self._dir + path

    def __len__(self):
        size = len(self._files)
        for m in self._dirs.values():
            size += m.__len__()
        return size

    def __str__(self):
        return '<treemanifest dir=%s>' % self._dir

    def iteritems(self):
        for p, n in sorted(self._dirs.items() + self._files.items()):
            if p in self._files:
                yield self._subpath(p), n
            else:
                for f, sn in n.iteritems():
                    yield f, sn

    def iterkeys(self):
        for p in sorted(self._dirs.keys() + self._files.keys()):
            if p in self._files:
                yield self._subpath(p)
            else:
                for f in self._dirs[p].iterkeys():
                    yield f

    def keys(self):
        return list(self.iterkeys())

    def __iter__(self):
        return self.iterkeys()

    def __contains__(self, f):
        if f is None:
            return False
        dir, subpath = _splittopdir(f)
        if dir:
            if dir not in self._dirs:
                return False
            return self._dirs[dir].__contains__(subpath)
        else:
            return f in self._files

    def get(self, f, default=None):
        dir, subpath = _splittopdir(f)
        if dir:
            if dir not in self._dirs:
                return default
            return self._dirs[dir].get(subpath, default)
        else:
            return self._files.get(f, default)

    def __getitem__(self, f):
        dir, subpath = _splittopdir(f)
        if dir:
            return self._dirs[dir].__getitem__(subpath)
        else:
            return self._files[f]

    def flags(self, f):
        dir, subpath = _splittopdir(f)
        if dir:
            if dir not in self._dirs:
                return ''
            return self._dirs[dir].flags(subpath)
        else:
            if f in self._dirs:
                return ''
            return self._flags.get(f, '')

    def find(self, f):
        dir, subpath = _splittopdir(f)
        if dir:
            return self._dirs[dir].find(subpath)
        else:
            return self._files[f], self._flags.get(f, '')

    def __delitem__(self, f):
        dir, subpath = _splittopdir(f)
        if dir:
            self._dirs[dir].__delitem__(subpath)
            # If the directory is now empty, remove it
            if not self._dirs[dir]._dirs and not self._dirs[dir]._files:
                del self._dirs[dir]
        else:
            del self._files[f]
            if f in self._flags:
                del self._flags[f]

    def __setitem__(self, f, n):
        assert n is not None
        dir, subpath = _splittopdir(f)
        if dir:
            if dir not in self._dirs:
                self._dirs[dir] = treemanifest(self._subpath(dir))
            self._dirs[dir].__setitem__(subpath, n)
        else:
            self._files[f] = n[:21] # to match manifestdict's behavior

    def setflag(self, f, flags):
        """Set the flags (symlink, executable) for path f."""
        dir, subpath = _splittopdir(f)
        if dir:
            if dir not in self._dirs:
                self._dirs[dir] = treemanifest(self._subpath(dir))
            self._dirs[dir].setflag(subpath, flags)
        else:
            self._flags[f] = flags

    def copy(self):
        copy = treemanifest(self._dir)
        for d in self._dirs:
            copy._dirs[d] = self._dirs[d].copy()
        copy._files = dict.copy(self._files)
        copy._flags = dict.copy(self._flags)
        return copy

    def _intersectfiles(self, files):
        '''make a new treemanifest with the intersection of self with files

        The algorithm assumes that files is much smaller than self.'''
        ret = treemanifest()
        for fn in files:
            if fn in self:
                ret[fn] = self[fn]
                flags = self.flags(fn)
                if flags:
                    ret.setflag(fn, flags)
        return ret

    def filesnotin(self, m2):
        '''Set of files in this manifest that are not in the other'''
        files = set()
        def _filesnotin(t1, t2):
            for d, m1 in t1._dirs.iteritems():
                if d in t2._dirs:
                    m2 = t2._dirs[d]
                    _filesnotin(m1, m2)
                else:
                    files.update(m1.iterkeys())

            for fn in t1._files.iterkeys():
                if fn not in t2._files:
                    files.add(t1._subpath(fn))

        _filesnotin(self, m2)
        return files

    @propertycache
    def _alldirs(self):
        return scmutil.dirs(self)

    def dirs(self):
        return self._alldirs

    def hasdir(self, dir):
        topdir, subdir = _splittopdir(dir)
        if topdir:
            if topdir in self._dirs:
                return self._dirs[topdir].hasdir(subdir)
            return False
        return (dir + '/') in self._dirs

    def matches(self, match):
        '''generate a new manifest filtered by the match argument'''
        if match.always():
            return self.copy()

        files = match.files()
        if (match.isexact() or
            (not match.anypats() and util.all(fn in self for fn in files))):
            return self._intersectfiles(files)

        m = self.copy()
        for fn in m.keys():
            if not match(fn):
                del m[fn]
        return m

    def diff(self, m2, clean=False):
        '''Finds changes between the current manifest and m2.

        Args:
          m2: the manifest to which this manifest should be compared.
          clean: if true, include files unchanged between these manifests
                 with a None value in the returned dictionary.

        The result is returned as a dict with filename as key and
        values of the form ((n1,fl1),(n2,fl2)), where n1/n2 is the
        nodeid in the current/other manifest and fl1/fl2 is the flag
        in the current/other manifest. Where the file does not exist,
        the nodeid will be None and the flags will be the empty
        string.
        '''
        result = {}
        emptytree = treemanifest()
        def _diff(t1, t2):
            for d, m1 in t1._dirs.iteritems():
                m2 = t2._dirs.get(d, emptytree)
                _diff(m1, m2)

            for d, m2 in t2._dirs.iteritems():
                if d not in t1._dirs:
                    _diff(emptytree, m2)

            for fn, n1 in t1._files.iteritems():
                fl1 = t1._flags.get(fn, '')
                n2 = t2._files.get(fn, None)
                fl2 = t2._flags.get(fn, '')
                if n1 != n2 or fl1 != fl2:
                    result[t1._subpath(fn)] = ((n1, fl1), (n2, fl2))
                elif clean:
                    result[t1._subpath(fn)] = None

            for fn, n2 in t2._files.iteritems():
                if fn not in t1._files:
                    fl2 = t2._flags.get(fn, '')
                    result[t2._subpath(fn)] = ((None, ''), (n2, fl2))

        _diff(self, m2)
        return result

    def text(self):
        """Get the full data of this manifest as a bytestring."""
        flags = self.flags
        return _text((f, self[f], flags(f)) for f in self.keys())

class manifest(revlog.revlog):
    def __init__(self, opener):
        # During normal operations, we expect to deal with not more than four
        # revs at a time (such as during commit --amend). When rebasing large
        # stacks of commits, the number can go up, hence the config knob below.
        cachesize = 4
        usetreemanifest = False
        usemanifestv2 = False
        opts = getattr(opener, 'options', None)
        if opts is not None:
            cachesize = opts.get('manifestcachesize', cachesize)
            usetreemanifest = opts.get('usetreemanifest', usetreemanifest)
            usemanifestv2 = opts.get('usemanifestv2', usemanifestv2)
        self._mancache = util.lrucachedict(cachesize)
        revlog.revlog.__init__(self, opener, "00manifest.i")
        self._usetreemanifest = usetreemanifest
        self._usemanifestv2 = usemanifestv2

    def _newmanifest(self, data=''):
        if self._usetreemanifest:
            return treemanifest('', data)
        return manifestdict(data)

    def _slowreaddelta(self, node):
        r0 = self.deltaparent(self.rev(node))
        m0 = self.read(self.node(r0))
        m1 = self.read(node)
        md = self._newmanifest()
        for f, ((n0, fl0), (n1, fl1)) in m0.diff(m1).iteritems():
            if n1:
                md[f] = n1
                if fl1:
                    md.setflag(f, fl1)
        return md

    def readdelta(self, node):
        if self._usemanifestv2:
            return self._slowreaddelta(node)
        r = self.rev(node)
        d = mdiff.patchtext(self.revdiff(self.deltaparent(r), r))
        return self._newmanifest(d)

    def readfast(self, node):
        '''use the faster of readdelta or read'''
        r = self.rev(node)
        deltaparent = self.deltaparent(r)
        if deltaparent != revlog.nullrev and deltaparent in self.parentrevs(r):
            return self.readdelta(node)
        return self.read(node)

    def read(self, node):
        if node == revlog.nullid:
            return self._newmanifest() # don't upset local cache
        if node in self._mancache:
            return self._mancache[node][0]
        text = self.revision(node)
        arraytext = array.array('c', text)
        m = self._newmanifest(text)
        self._mancache[node] = (m, arraytext)
        return m

    def find(self, node, f):
        '''look up entry for a single file efficiently.
        return (node, flags) pair if found, (None, None) if not.'''
        m = self.read(node)
        try:
            return m.find(f)
        except KeyError:
            return None, None

    def add(self, m, transaction, link, p1, p2, added, removed):
        if (p1 in self._mancache and not self._usetreemanifest
            and not self._usemanifestv2):
            # If our first parent is in the manifest cache, we can
            # compute a delta here using properties we know about the
            # manifest up-front, which may save time later for the
            # revlog layer.

            _checkforbidden(added)
            # combine the changed lists into one list for sorting
            work = [(x, False) for x in added]
            work.extend((x, True) for x in removed)
            # this could use heapq.merge() (from Python 2.6+) or equivalent
            # since the lists are already sorted
            work.sort()

            arraytext, deltatext = m.fastdelta(self._mancache[p1][1], work)
            cachedelta = self.rev(p1), deltatext
            text = util.buffer(arraytext)
        else:
            # The first parent manifest isn't already loaded, so we'll
            # just encode a fulltext of the manifest and pass that
            # through to the revlog layer, and let it handle the delta
            # process.
            text = m.text()
            arraytext = array.array('c', text)
            cachedelta = None

        n = self.addrevision(text, transaction, link, p1, p2, cachedelta)
        self._mancache[n] = (m, arraytext)

        return n