# Copyright 2009-2010 Gregory P. Ward

# Copyright 2009-2010 Intelerad Medical Systems Incorporated

# Copyright 2010-2011 Fog Creek Software

# Copyright 2010-2011 Unity Technologies

#

# This software may be used and distributed according to the terms of the

# GNU General Public License version 2 or any later version.

'''High-level command function for lfconvert, plus the cmdtable.'''

from __future__ import absolute_import

import errno

import hashlib

import os

import shutil

from mercurial.i18n import _

from mercurial import (

    cmdutil,

    context,

    error,

    exthelper,

    hg,

    lock,

    match as matchmod,

    node,

    pycompat,

    scmutil,

    util,

)

from ..convert import (

    convcmd,

    filemap,

)

from . import (

    lfutil,

    storefactory

)

release = lock.release

# -- Commands ----------------------------------------------------------

eh = exthelper.exthelper()

@eh.command('lfconvert',

    [('s', 'size', '',

      _('minimum size (MB) for files to be converted as largefiles'), 'SIZE'),

    ('', 'to-normal', False,

     _('convert from a largefiles repo to a normal repo')),

    ],

    _('hg lfconvert SOURCE DEST [FILE ...]'),

    norepo=True,

    inferrepo=True)

def lfconvert(ui, src, dest, *pats, **opts):

    '''convert a normal repository to a largefiles repository

    Convert repository SOURCE to a new repository DEST, identical to

    SOURCE except that certain files will be converted as largefiles:

    specifically, any file that matches any PATTERN *or* whose size is

    above the minimum size threshold is converted as a largefile. The

    size used to determine whether or not to track a file as a

    largefile is the size of the first version of the file. The

    minimum size can be specified either with --size or in

    configuration as ``largefiles.size``.

    After running this command you will need to make sure that

    largefiles is enabled anywhere you intend to push the new

    repository.

    Use --to-normal to convert largefiles back to normal files; after

    this, the DEST repository can be used without largefiles at all.'''

    opts = pycompat.byteskwargs(opts)

    if opts['to_normal']:

        tolfile = False

    else:

        tolfile = True

        size = lfutil.getminsize(ui, True, opts.get('size'), default=None)

    if not hg.islocal(src):

        raise error.Abort(_('%s is not a local Mercurial repo') % src)

    if not hg.islocal(dest):

        raise error.Abort(_('%s is not a local Mercurial repo') % dest)

    rsrc = hg.repository(ui, src)

    ui.status(_('initializing destination %s\n') % dest)

    rdst = hg.repository(ui, dest, create=True)

    success = False

    dstwlock = dstlock = None

    try:

        # Get a list of all changesets in the source.  The easy way to do this

        # is to simply walk the changelog, using changelog.nodesbetween().

        # Take a look at mercurial/revlog.py:639 for more details.

        # Use a generator instead of a list to decrease memory usage

        ctxs = (rsrc[ctx] for ctx in rsrc.changelog.nodesbetween(None,

            rsrc.heads())[0])

        revmap = {node.nullid: node.nullid}

        if tolfile:

            # Lock destination to prevent modification while it is converted to.

            # Don't need to lock src because we are just reading from its

            # history which can't change.

            dstwlock = rdst.wlock()

            dstlock = rdst.lock()

            lfiles = set()

            normalfiles = set()

            if not pats:

                pats = ui.configlist(lfutil.longname, 'patterns')

            if pats:

                matcher = matchmod.match(rsrc.root, '', list(pats))

            else:

                matcher = None

            lfiletohash = {}

            with ui.makeprogress(_('converting revisions'),

                                 unit=_('revisions'),

                                 total=rsrc['tip'].rev()) as progress:

                for ctx in ctxs:

                    progress.update(ctx.rev())

                    _lfconvert_addchangeset(rsrc, rdst, ctx, revmap,

                        lfiles, normalfiles, matcher, size, lfiletohash)

            if rdst.wvfs.exists(lfutil.shortname):

                rdst.wvfs.rmtree(lfutil.shortname)

            for f in lfiletohash.keys():

                if rdst.wvfs.isfile(f):

                    rdst.wvfs.unlink(f)

                try:

                    rdst.wvfs.removedirs(rdst.wvfs.dirname(f))

                except OSError:

                    pass

            # If there were any files converted to largefiles, add largefiles

            # to the destination repository's requirements.

            if lfiles:

                rdst.requirements.add('largefiles')

                rdst._writerequirements()

        else:

            class lfsource(filemap.filemap_source):

                def __init__(self, ui, source):

                    super(lfsource, self).__init__(ui, source, None)

                    self.filemapper.rename[lfutil.shortname] = '.'

                def getfile(self, name, rev):

                    realname, realrev = rev

                    f = super(lfsource, self).getfile(name, rev)

                    if (not realname.startswith(lfutil.shortnameslash)

                            or f[0] is None):

                        return f

                    # Substitute in the largefile data for the hash

                    hash = f[0].strip()

                    path = lfutil.findfile(rsrc, hash)

                    if path is None:

                        raise error.Abort(_("missing largefile for '%s' in %s")

                                          % (realname, realrev))

                    return util.readfile(path), f[1]

            class converter(convcmd.converter):

                def __init__(self, ui, source, dest, revmapfile, opts):

                    src = lfsource(ui, source)

                    super(converter, self).__init__(ui, src, dest, revmapfile,

                                                    opts)

            found, missing = downloadlfiles(ui, rsrc)

            if missing != 0:

                raise error.Abort(_("all largefiles must be present locally"))

            orig = convcmd.converter

            convcmd.converter = converter

            try:

                convcmd.convert(ui, src, dest, source_type='hg', dest_type='hg')

            finally:

                convcmd.converter = orig

        success = True

    finally:

        if tolfile:

            rdst.dirstate.clear()

            release(dstlock, dstwlock)

        if not success:

            # we failed, remove the new directory

            shutil.rmtree(rdst.root)

def _lfconvert_addchangeset(rsrc, rdst, ctx, revmap, lfiles, normalfiles,

        matcher, size, lfiletohash):

    # Convert src parents to dst parents

    parents = _convertparents(ctx, revmap)

    # Generate list of changed files

    files = _getchangedfiles(ctx, parents)

    dstfiles = []

    for f in files:

        if f not in lfiles and f not in normalfiles:

            islfile = _islfile(f, ctx, matcher, size)

            # If this file was renamed or copied then copy

            # the largefile-ness of its predecessor

            if f in ctx.manifest():

                fctx = ctx.filectx(f)

                renamed = fctx.renamed()

                if renamed is None:

                    # the code below assumes renamed to be a boolean or a list

                    # and won't quite work with the value None

                    renamed = False

                renamedlfile = renamed and renamed[0] in lfiles

                islfile |= renamedlfile

                if 'l' in fctx.flags():

                    if renamedlfile:

                        raise error.Abort(

                            _('renamed/copied largefile %s becomes symlink')

                            % f)

                    islfile = False

            if islfile:

                lfiles.add(f)

            else:

                normalfiles.add(f)

        if f in lfiles:

            fstandin = lfutil.standin(f)

            dstfiles.append(fstandin)

            # largefile in manifest if it has not been removed/renamed

            if f in ctx.manifest():

                fctx = ctx.filectx(f)

                if 'l' in fctx.flags():

                    renamed = fctx.renamed()

                    if renamed and renamed[0] in lfiles:

                        raise error.Abort(_('largefile %s becomes symlink') % f)

                # largefile was modified, update standins

                m = hashlib.sha1('')

                m.update(ctx[f].data())

                hash = node.hex(m.digest())

                if f not in lfiletohash or lfiletohash[f] != hash:

                    rdst.wwrite(f, ctx[f].data(), ctx[f].flags())

                    executable = 'x' in ctx[f].flags()

                    lfutil.writestandin(rdst, fstandin, hash,

                        executable)

                    lfiletohash[f] = hash

        else:

            # normal file

            dstfiles.append(f)

    def getfilectx(repo, memctx, f):

        srcfname = lfutil.splitstandin(f)

        if srcfname is not None:

            # if the file isn't in the manifest then it was removed

            # or renamed, return None to indicate this

            try:

                fctx = ctx.filectx(srcfname)

            except error.LookupError:

                return None

            renamed = fctx.renamed()

            if renamed:

                # standin is always a largefile because largefile-ness

                # doesn't change after rename or copy

                renamed = lfutil.standin(renamed[0])

            return context.memfilectx(repo, memctx, f,

                                      lfiletohash[srcfname] + '\n',

                                      'l' in fctx.flags(), 'x' in fctx.flags(),

                                      renamed)

        else:

            return _getnormalcontext(repo, ctx, f, revmap)

    # Commit

    _commitcontext(rdst, parents, ctx, dstfiles, getfilectx, revmap)

def _commitcontext(rdst, parents, ctx, dstfiles, getfilectx, revmap):

    mctx = context.memctx(rdst, parents, ctx.description(), dstfiles,

                          getfilectx, ctx.user(), ctx.date(), ctx.extra())

    ret = rdst.commitctx(mctx)

    lfutil.copyalltostore(rdst, ret)

    rdst.setparents(ret)

    revmap[ctx.node()] = rdst.changelog.tip()

# Generate list of changed files

def _getchangedfiles(ctx, parents):

    files = set(ctx.files())

    if node.nullid not in parents:

        mc = ctx.manifest()

        for pctx in ctx.parents():

            for fn in pctx.manifest().diff(mc):

                files.add(fn)

    return files

# Convert src parents to dst parents

def _convertparents(ctx, revmap):

    parents = []

    for p in ctx.parents():

        parents.append(revmap[p.node()])

    while len(parents) < 2:

        parents.append(node.nullid)

    return parents

# Get memfilectx for a normal file

def _getnormalcontext(repo, ctx, f, revmap):

    try:

        fctx = ctx.filectx(f)

    except error.LookupError:

        return None

    renamed = fctx.renamed()

    if renamed:

        renamed = renamed[0]

    data = fctx.data()

    if f == '.hgtags':

        data = _converttags (repo.ui, revmap, data)

    return context.memfilectx(repo, ctx, f, data, 'l' in fctx.flags(),

                              'x' in fctx.flags(), renamed)

# Remap tag data using a revision map

def _converttags(ui, revmap, data):

    newdata = []

    for line in data.splitlines():

        try:

            id, name = line.split(' ', 1)

        except ValueError:

            ui.warn(_('skipping incorrectly formatted tag %s\n')

                % line)

            continue

        try:

            newid = node.bin(id)

        except TypeError:

            ui.warn(_('skipping incorrectly formatted id %s\n')

                % id)

            continue

        try:

            newdata.append('%s %s\n' % (node.hex(revmap[newid]),

                name))

        except KeyError:

            ui.warn(_('no mapping for id %s\n') % id)

            continue

    return ''.join(newdata)

def _islfile(file, ctx, matcher, size):

    '''Return true if file should be considered a largefile, i.e.

    matcher matches it or it is larger than size.'''

    # never store special .hg* files as largefiles

    if file == '.hgtags' or file == '.hgignore' or file == '.hgsigs':

        return False

    if matcher and matcher(file):

        return True

    try:

        return ctx.filectx(file).size() >= size * 1024 * 1024

    except error.LookupError:

        return False

def uploadlfiles(ui, rsrc, rdst, files):

    '''upload largefiles to the central store'''

    if not files:

        return

    store = storefactory.openstore(rsrc, rdst, put=True)

    at = 0

    ui.debug("sending statlfile command for %d largefiles\n" % len(files))

    retval = store.exists(files)

    files = [h for h in files if not retval[h]]

    ui.debug("%d largefiles need to be uploaded\n" % len(files))

    with ui.makeprogress(_('uploading largefiles'), unit=_('files'),

                         total=len(files)) as progress:

        for hash in files:

            progress.update(at)

            source = lfutil.findfile(rsrc, hash)