Mercurial > hg > mercurial-crew-with-dirclash
annotate mercurial/manifest.py @ 1677:11d12bd6e1dc
cleanup of revlog.group when repository is local
revlog.group cached every chunk from the revlog, the behaviour was
needed to minimize the roundtrip with old-http.
The patch export the information that the repository is local or not
from the repository object down to the revlog.
Then it uses the workaround for old-http only if the repository is non-local.
The memory used server side when pulling goes down to less than 30Mo maximum
whereas without the patch more than 160Mo was used when cloning the linux kernel
repository.
The time used by cloning is roughly the same (although some caching could be
implemented if needed):
before
110.25user 20.90system 2:52.00elapsed 76%CPU (0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (0major+708707minor)pagefaults 0swaps
after
112.85user 22.98system 2:50.66elapsed 79%CPU (0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (0major+862862minor)pagefaults 0swaps
| author | Benoit Boissinot <benoit.boissinot@ens-lyon.org> |
|---|---|
| date | Thu, 08 Dec 2005 15:12:02 +0100 |
| parents | bf4e7ef08741 |
| children | b345cc4c22c0 |
| rev | line source |
|---|---|
| 1089 | 1 # manifest.py - manifest revision class for mercurial |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
2 # |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
3 # Copyright 2005 Matt Mackall <mpm@selenic.com> |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
4 # |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
5 # This software may be used and distributed according to the terms |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
6 # of the GNU General Public License, incorporated herein by reference. |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
7 |
|
1541
bf4e7ef08741
fixed some stuff pychecker shows, marked unclear/wrong stuff with XXX
twaldmann@thinkmo.de
parents:
1534
diff
changeset
|
8 import struct |
| 262 | 9 from revlog import * |
|
1400
cf9a1233738a
i18n first part: make '_' available for files who need it
Benoit Boissinot <benoit.boissinot@ens-lyon.org
parents:
1098
diff
changeset
|
10 from i18n import gettext as _ |
| 262 | 11 from demandload import * |
| 1534 | 12 demandload(globals(), "bisect array") |
| 79 | 13 |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
14 class manifest(revlog): |
|
1677
11d12bd6e1dc
cleanup of revlog.group when repository is local
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
1541
diff
changeset
|
15 def __init__(self, opener, local=True): |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
16 self.mapcache = None |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
17 self.listcache = None |
|
1677
11d12bd6e1dc
cleanup of revlog.group when repository is local
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
1541
diff
changeset
|
18 revlog.__init__(self, opener, "00manifest.i", "00manifest.d", |
|
11d12bd6e1dc
cleanup of revlog.group when repository is local
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
1541
diff
changeset
|
19 local=local) |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
20 |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
21 def read(self, node): |
| 313 | 22 if node == nullid: return {} # don't upset local cache |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
23 if self.mapcache and self.mapcache[0] == node: |
| 561 | 24 return self.mapcache[1] |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
25 text = self.revision(node) |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
26 map = {} |
| 276 | 27 flag = {} |
| 1534 | 28 self.listcache = array.array('c', text) |
| 29 lines = text.splitlines(1) | |
| 30 for l in lines: | |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
31 (f, n) = l.split('\0') |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
32 map[f] = bin(n[:40]) |
| 276 | 33 flag[f] = (n[40:-1] == "x") |
| 34 self.mapcache = (node, map, flag) | |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
35 return map |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
36 |
| 276 | 37 def readflags(self, node): |
| 313 | 38 if node == nullid: return {} # don't upset local cache |
|
358
9f4077d7ef6f
[PATCH] manifest.readflags performance buglet
mpm@selenic.com
parents:
350
diff
changeset
|
39 if not self.mapcache or self.mapcache[0] != node: |
| 276 | 40 self.read(node) |
| 41 return self.mapcache[2] | |
| 42 | |
| 1534 | 43 def diff(self, a, b): |
| 44 return mdiff.textdiff(str(a), str(b)) | |
| 45 | |
| 741 | 46 def add(self, map, flags, transaction, link, p1=None, p2=None, |
| 47 changed=None): | |
| 1534 | 48 |
| 49 # returns a tuple (start, end). If the string is found | |
| 50 # m[start:end] are the line containing that string. If start == end | |
| 51 # the string was not found and they indicate the proper sorted | |
| 52 # insertion point. This was taken from bisect_left, and modified | |
| 53 # to find line start/end as it goes along. | |
| 54 # | |
| 55 # m should be a buffer or a string | |
| 56 # s is a string | |
| 57 # | |
| 58 def manifestsearch(m, s, lo=0, hi=None): | |
| 59 def advance(i, c): | |
| 60 while i < lenm and m[i] != c: | |
| 644 | 61 i += 1 |
| 1534 | 62 return i |
| 63 lenm = len(m) | |
| 64 if not hi: | |
| 65 hi = lenm | |
| 66 while lo < hi: | |
| 67 mid = (lo + hi) // 2 | |
| 68 start = mid | |
| 69 while start > 0 and m[start-1] != '\n': | |
| 70 start -= 1 | |
| 71 end = advance(start, '\0') | |
| 72 if m[start:end] < s: | |
| 73 # we know that after the null there are 40 bytes of sha1 | |
| 74 # this translates to the bisect lo = mid + 1 | |
| 75 lo = advance(end + 40, '\n') + 1 | |
| 76 else: | |
| 77 # this translates to the bisect hi = mid | |
| 78 hi = start | |
| 79 end = advance(lo, '\0') | |
| 80 found = m[lo:end] | |
| 81 if cmp(s, found) == 0: | |
| 82 # we know that after the null there are 40 bytes of sha1 | |
| 83 end = advance(end + 40, '\n') | |
| 84 return (lo, end+1) | |
| 85 else: | |
| 86 return (lo, lo) | |
| 644 | 87 |
| 88 # apply the changes collected during the bisect loop to our addlist | |
| 1534 | 89 # return a delta suitable for addrevision |
| 90 def addlistdelta(addlist, x): | |
| 91 # start from the bottom up | |
| 644 | 92 # so changes to the offsets don't mess things up. |
| 1534 | 93 i = len(x) |
| 644 | 94 while i > 0: |
| 95 i -= 1 | |
| 1534 | 96 start = x[i][0] |
| 97 end = x[i][1] | |
| 98 if x[i][2]: | |
| 99 addlist[start:end] = array.array('c', x[i][2]) | |
| 644 | 100 else: |
| 101 del addlist[start:end] | |
| 1534 | 102 return "".join([struct.pack(">lll", d[0], d[1], len(d[2])) + d[2] \ |
| 103 for d in x ]) | |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
104 |
| 644 | 105 # if we're using the listcache, make sure it is valid and |
| 106 # parented by the same node we're diffing against | |
| 741 | 107 if not changed or not self.listcache or not p1 or \ |
| 108 self.mapcache[0] != p1: | |
| 644 | 109 files = map.keys() |
| 110 files.sort() | |
| 111 | |
| 1534 | 112 text = ["%s\000%s%s\n" % |
| 644 | 113 (f, hex(map[f]), flags[f] and "x" or '') |
| 114 for f in files] | |
| 1534 | 115 self.listcache = array.array('c', "".join(text)) |
| 644 | 116 cachedelta = None |
| 117 else: | |
| 1534 | 118 addlist = self.listcache |
| 644 | 119 |
| 120 # combine the changed lists into one list for sorting | |
| 121 work = [[x, 0] for x in changed[0]] | |
| 122 work[len(work):] = [[x, 1] for x in changed[1]] | |
| 123 work.sort() | |
| 124 | |
| 125 delta = [] | |
| 1534 | 126 dstart = None |
| 127 dend = None | |
| 128 dline = [""] | |
| 129 start = 0 | |
| 130 # zero copy representation of addlist as a buffer | |
| 131 addbuf = buffer(addlist) | |
| 644 | 132 |
| 1534 | 133 # start with a readonly loop that finds the offset of |
| 134 # each line and creates the deltas | |
| 644 | 135 for w in work: |
| 136 f = w[0] | |
| 741 | 137 # bs will either be the index of the item or the insert point |
| 1534 | 138 start, end = manifestsearch(addbuf, f, start) |
| 644 | 139 if w[1] == 0: |
| 741 | 140 l = "%s\000%s%s\n" % (f, hex(map[f]), |
| 141 flags[f] and "x" or '') | |
| 644 | 142 else: |
| 1534 | 143 l = "" |
| 144 if start == end and w[1] == 1: | |
| 145 # item we want to delete was not found, error out | |
| 146 raise AssertionError( | |
|
1402
9d2c2e6b32b5
i18n part2: use '_' for all strings who are part of the user interface
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents:
1400
diff
changeset
|
147 _("failed to remove %s from manifest\n") % f) |
| 1534 | 148 if dstart != None and dstart <= start and dend >= start: |
| 149 if dend < end: | |
| 150 dend = end | |
| 151 if l: | |
| 152 dline.append(l) | |
| 644 | 153 else: |
| 1534 | 154 if dstart != None: |
| 155 delta.append([dstart, dend, "".join(dline)]) | |
| 156 dstart = start | |
| 157 dend = end | |
| 158 dline = [l] | |
| 644 | 159 |
| 1534 | 160 if dstart != None: |
| 161 delta.append([dstart, dend, "".join(dline)]) | |
| 162 # apply the delta to the addlist, and get a delta for addrevision | |
| 163 cachedelta = addlistdelta(addlist, delta) | |
| 644 | 164 |
| 1534 | 165 # the delta is only valid if we've been processing the tip revision |
| 166 if self.mapcache[0] != self.tip(): | |
| 167 cachedelta = None | |
| 168 self.listcache = addlist | |
| 169 | |
| 170 n = self.addrevision(buffer(self.listcache), transaction, link, p1, \ | |
| 171 p2, cachedelta) | |
| 302 | 172 self.mapcache = (n, map, flags) |
|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
173 |
|
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
174 return n |