# manifest.py - manifest revision class for mercurial
#
# Copyright 2005-2007 Matt Mackall <[email protected]>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2, incorporated herein by reference.

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

class manifestdict(dict):
   def __init__(self, mapping=None, flags=None):
       if mapping is None: mapping = {}
       if flags is None: flags = {}
       dict.__init__(self, mapping)
       self._flags = flags
   def flags(self, f):
       return self._flags.get(f, "")
   def set(self, f, flags):
       self._flags[f] = flags
   def copy(self):
       return manifestdict(dict.copy(self), dict.copy(self._flags))

class manifest(revlog.revlog):
   def __init__(self, opener):
       self.mapcache = None
       self.listcache = None
       revlog.revlog.__init__(self, opener, "00manifest.i")

   def parse(self, lines):
       mfdict = manifestdict()
       parsers.parse_manifest(mfdict, mfdict._flags, lines)
       return mfdict

   def readdelta(self, node):
       r = self.rev(node)
       return self.parse(mdiff.patchtext(self.revdiff(r - 1, r)))

   def read(self, node):
       if node == revlog.nullid:
           return manifestdict() # don't upset local cache
       if self.mapcache and self.mapcache[0] == node:
           return self.mapcache[1]
       text = self.revision(node)
       self.listcache = array.array('c', text)
       mapping = self.parse(text)
       self.mapcache = (node, mapping)
       return mapping

   def _search(self, 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.  This was
       taken from bisect_left, and modified to find line start/end as
       it goes along.

       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 cmp(s, found) == 0:
           # 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 find(self, node, f):
       '''look up entry for a single file efficiently.
       return (node, flags) pair if found, (None, None) if not.'''
       if self.mapcache and node == self.mapcache[0]:
           return self.mapcache[1].get(f), self.mapcache[1].flags(f)
       text = self.revision(node)
       start, end = self._search(text, f)
       if start == end:
           return None, None
       l = text[start:end]
       f, n = l.split('\0')
       return revlog.bin(n[:40]), n[40:-1]

   def add(self, map, transaction, link, p1=None, p2=None,
           changed=None):
       # apply the changes collected during the bisect loop to our addlist
       # return a delta suitable for addrevision
       def addlistdelta(addlist, x):
           # start from the bottom up
           # so changes to the offsets don't mess things up.
           i = len(x)
           while i > 0:
               i -= 1
               start = x[i][0]
               end = x[i][1]
               if x[i][2]:
                   addlist[start:end] = array.array('c', x[i][2])
               else:
                   del addlist[start:end]
           return "".join([struct.pack(">lll", d[0], d[1], len(d[2])) + d[2]
                           for d in x ])

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

       # if we're using the listcache, make sure it is valid and
       # parented by the same node we're diffing against
       if not (changed and self.listcache and p1 and self.mapcache[0] == p1):
           files = sorted(map)
           checkforbidden(files)

           # if this is changed to support newlines in filenames,
           # be sure to check the templates/ dir again (especially *-raw.tmpl)
           hex, flags = revlog.hex, map.flags
           text = ["%s\000%s%s\n" % (f, hex(map[f]), flags(f))
                   for f in files]
           self.listcache = array.array('c', "".join(text))
           cachedelta = None
       else:
           addlist = self.listcache

           checkforbidden(changed[0])
           # combine the changed lists into one list for sorting
           work = [[x, 0] for x in changed[0]]
           work[len(work):] = [[x, 1] for x in changed[1]]
           work.sort()

           delta = []
           dstart = None
           dend = None
           dline = [""]
           start = 0
           # zero copy representation of addlist as a buffer
           addbuf = buffer(addlist)

           # start with a readonly loop that finds the offset of
           # each line and creates the deltas
           for w in work:
               f = w[0]
               # bs will either be the index of the item or the insert point
               start, end = self._search(addbuf, f, start)
               if w[1] == 0:
                   l = "%s\000%s%s\n" % (f, revlog.hex(map[f]), map.flags(f))
               else:
                   l = ""
               if start == end and w[1] == 1:
                   # item we want to delete was not found, error out
                   raise AssertionError(
                           _("failed to remove %s from manifest") % f)
               if dstart != None and dstart <= start and dend >= start:
                   if dend < end:
                       dend = end
                   if l:
                       dline.append(l)
               else:
                   if dstart != None:
                       delta.append([dstart, dend, "".join(dline)])
                   dstart = start
                   dend = end
                   dline = [l]

           if dstart != None:
               delta.append([dstart, dend, "".join(dline)])
           # apply the delta to the addlist, and get a delta for addrevision
           cachedelta = addlistdelta(addlist, delta)

           # the delta is only valid if we've been processing the tip revision
           if self.mapcache[0] != self.tip():
               cachedelta = None
           self.listcache = addlist

       n = self.addrevision(buffer(self.listcache), transaction, link,
                            p1, p2, cachedelta)
       self.mapcache = (n, map)

       return n