20
20
lines by NL. The field delimiters are ommitted in the grammar, line delimiters
21
21
are not - this is done for clarity of reading. All string data is in utf8.
23
MINIKIND = "f" | "d" | "l" | "a" | "r" | "t";
26
WHOLE_NUMBER = {digit}, digit;
28
REVISION_ID = a non-empty utf8 string;
30
dirstate format = header line, full checksum, row count, parent details,
31
ghost_details, entries;
32
header line = "#bazaar dirstate flat format 3", NL;
33
full checksum = "crc32: ", ["-"], WHOLE_NUMBER, NL;
34
row count = "num_entries: ", WHOLE_NUMBER, NL;
35
parent_details = WHOLE NUMBER, {REVISION_ID}* NL;
36
ghost_details = WHOLE NUMBER, {REVISION_ID}*, NL;
38
entry = entry_key, current_entry_details, {parent_entry_details};
39
entry_key = dirname, basename, fileid;
40
current_entry_details = common_entry_details, working_entry_details;
41
parent_entry_details = common_entry_details, history_entry_details;
42
common_entry_details = MINIKIND, fingerprint, size, executable
43
working_entry_details = packed_stat
44
history_entry_details = REVISION_ID;
47
fingerprint = a nonempty utf8 sequence with meaning defined by minikind.
49
Given this definition, the following is useful to know:
50
entry (aka row) - all the data for a given key.
51
entry[0]: The key (dirname, basename, fileid)
55
entry[1]: The tree(s) data for this path and id combination.
56
entry[1][0]: The current tree
57
entry[1][1]: The second tree
59
For an entry for a tree, we have (using tree 0 - current tree) to demonstrate:
60
entry[1][0][0]: minikind
61
entry[1][0][1]: fingerprint
63
entry[1][0][3]: executable
64
entry[1][0][4]: packed_stat
66
entry[1][1][4]: revision_id
25
MINIKIND = "f" | "d" | "l" | "a" | "r" | "t";
28
WHOLE_NUMBER = {digit}, digit;
30
REVISION_ID = a non-empty utf8 string;
32
dirstate format = header line, full checksum, row count, parent details,
33
ghost_details, entries;
34
header line = "#bazaar dirstate flat format 3", NL;
35
full checksum = "crc32: ", ["-"], WHOLE_NUMBER, NL;
36
row count = "num_entries: ", WHOLE_NUMBER, NL;
37
parent_details = WHOLE NUMBER, {REVISION_ID}* NL;
38
ghost_details = WHOLE NUMBER, {REVISION_ID}*, NL;
40
entry = entry_key, current_entry_details, {parent_entry_details};
41
entry_key = dirname, basename, fileid;
42
current_entry_details = common_entry_details, working_entry_details;
43
parent_entry_details = common_entry_details, history_entry_details;
44
common_entry_details = MINIKIND, fingerprint, size, executable
45
working_entry_details = packed_stat
46
history_entry_details = REVISION_ID;
49
fingerprint = a nonempty utf8 sequence with meaning defined by minikind.
51
Given this definition, the following is useful to know::
53
entry (aka row) - all the data for a given key.
54
entry[0]: The key (dirname, basename, fileid)
58
entry[1]: The tree(s) data for this path and id combination.
59
entry[1][0]: The current tree
60
entry[1][1]: The second tree
62
For an entry for a tree, we have (using tree 0 - current tree) to demonstrate::
64
entry[1][0][0]: minikind
65
entry[1][0][1]: fingerprint
67
entry[1][0][3]: executable
68
entry[1][0][4]: packed_stat
72
entry[1][1][4]: revision_id
68
74
There may be multiple rows at the root, one per id present in the root, so the
69
in memory root row is now:
70
self._dirblocks[0] -> ('', [entry ...]),
71
and the entries in there are
74
entries[0][2]: file_id
75
entries[1][0]: The tree data for the current tree for this fileid at /
79
'r' is a relocated entry: This path is not present in this tree with this id,
80
but the id can be found at another location. The fingerprint is used to
81
point to the target location.
82
'a' is an absent entry: In that tree the id is not present at this path.
83
'd' is a directory entry: This path in this tree is a directory with the
84
current file id. There is no fingerprint for directories.
85
'f' is a file entry: As for directory, but it's a file. The fingerprint is the
86
sha1 value of the file's canonical form, i.e. after any read filters have
87
been applied to the convenience form stored in the working tree.
88
'l' is a symlink entry: As for directory, but a symlink. The fingerprint is the
90
't' is a reference to a nested subtree; the fingerprint is the referenced
75
in memory root row is now::
77
self._dirblocks[0] -> ('', [entry ...]),
79
and the entries in there are::
83
entries[0][2]: file_id
84
entries[1][0]: The tree data for the current tree for this fileid at /
89
'r' is a relocated entry: This path is not present in this tree with this
90
id, but the id can be found at another location. The fingerprint is
91
used to point to the target location.
92
'a' is an absent entry: In that tree the id is not present at this path.
93
'd' is a directory entry: This path in this tree is a directory with the
94
current file id. There is no fingerprint for directories.
95
'f' is a file entry: As for directory, but it's a file. The fingerprint is
96
the sha1 value of the file's canonical form, i.e. after any read
97
filters have been applied to the convenience form stored in the working
99
'l' is a symlink entry: As for directory, but a symlink. The fingerprint is
101
't' is a reference to a nested subtree; the fingerprint is the referenced
95
The entries on disk and in memory are ordered according to the following keys:
106
The entries on disk and in memory are ordered according to the following keys::
97
108
directory, as a list of components
101
112
--- Format 1 had the following different definition: ---
102
rows = dirname, NULL, basename, NULL, MINIKIND, NULL, fileid_utf8, NULL,
103
WHOLE NUMBER (* size *), NULL, packed stat, NULL, sha1|symlink target,
105
PARENT ROW = NULL, revision_utf8, NULL, MINIKIND, NULL, dirname, NULL,
106
basename, NULL, WHOLE NUMBER (* size *), NULL, "y" | "n", NULL,
116
rows = dirname, NULL, basename, NULL, MINIKIND, NULL, fileid_utf8, NULL,
117
WHOLE NUMBER (* size *), NULL, packed stat, NULL, sha1|symlink target,
119
PARENT ROW = NULL, revision_utf8, NULL, MINIKIND, NULL, dirname, NULL,
120
basename, NULL, WHOLE NUMBER (* size *), NULL, "y" | "n", NULL,
109
123
PARENT ROW's are emitted for every parent that is not in the ghosts details
110
124
line. That is, if the parents are foo, bar, baz, and the ghosts are bar, then
411
445
self._last_block_index = None
412
446
self._last_entry_index = None
447
# The set of known hash changes
448
self._known_hash_changes = set()
449
# How many hash changed entries can we have without saving
450
self._worth_saving_limit = worth_saving_limit
414
452
def __repr__(self):
415
453
return "%s(%r)" % \
416
454
(self.__class__.__name__, self._filename)
456
def _mark_modified(self, hash_changed_entries=None, header_modified=False):
457
"""Mark this dirstate as modified.
459
:param hash_changed_entries: if non-None, mark just these entries as
460
having their hash modified.
461
:param header_modified: mark the header modified as well, not just the
464
#trace.mutter_callsite(3, "modified hash entries: %s", hash_changed_entries)
465
if hash_changed_entries:
466
self._known_hash_changes.update([e[0] for e in hash_changed_entries])
467
if self._dirblock_state in (DirState.NOT_IN_MEMORY,
468
DirState.IN_MEMORY_UNMODIFIED):
469
# If the dirstate is already marked a IN_MEMORY_MODIFIED, then
470
# that takes precedence.
471
self._dirblock_state = DirState.IN_MEMORY_HASH_MODIFIED
473
# TODO: Since we now have a IN_MEMORY_HASH_MODIFIED state, we
474
# should fail noisily if someone tries to set
475
# IN_MEMORY_MODIFIED but we don't have a write-lock!
476
# We don't know exactly what changed so disable smart saving
477
self._dirblock_state = DirState.IN_MEMORY_MODIFIED
479
self._header_state = DirState.IN_MEMORY_MODIFIED
481
def _mark_unmodified(self):
482
"""Mark this dirstate as unmodified."""
483
self._header_state = DirState.IN_MEMORY_UNMODIFIED
484
self._dirblock_state = DirState.IN_MEMORY_UNMODIFIED
485
self._known_hash_changes = set()
418
487
def add(self, path, file_id, kind, stat, fingerprint):
419
488
"""Add a path to be tracked.
1483
1550
if basename_utf8:
1484
1551
parents.add((dirname_utf8, inv_entry.parent_id))
1485
1552
if old_path is None:
1486
adds.append((None, encode(new_path), file_id,
1553
old_path_utf8 = None
1555
old_path_utf8 = encode(old_path)
1556
if old_path is None:
1557
adds.append((None, new_path_utf8, file_id,
1487
1558
inv_to_entry(inv_entry), True))
1488
1559
new_ids.add(file_id)
1489
1560
elif new_path is None:
1490
deletes.append((encode(old_path), None, file_id, None, True))
1491
elif (old_path, new_path) != root_only:
1561
deletes.append((old_path_utf8, None, file_id, None, True))
1562
elif (old_path, new_path) == root_only:
1563
# change things in-place
1564
# Note: the case of a parent directory changing its file_id
1565
# tends to break optimizations here, because officially
1566
# the file has actually been moved, it just happens to
1567
# end up at the same path. If we can figure out how to
1568
# handle that case, we can avoid a lot of add+delete
1569
# pairs for objects that stay put.
1570
# elif old_path == new_path:
1571
changes.append((old_path_utf8, new_path_utf8, file_id,
1572
inv_to_entry(inv_entry)))
1493
1575
# Because renames must preserve their children we must have
1494
1576
# processed all relocations and removes before hand. The sort
1504
1586
self._update_basis_apply_deletes(deletes)
1506
1588
# Split into an add/delete pair recursively.
1507
adds.append((None, new_path_utf8, file_id,
1508
inv_to_entry(inv_entry), False))
1589
adds.append((old_path_utf8, new_path_utf8, file_id,
1590
inv_to_entry(inv_entry), False))
1509
1591
# Expunge deletes that we've seen so that deleted/renamed
1510
1592
# children of a rename directory are handled correctly.
1511
new_deletes = reversed(list(self._iter_child_entries(1,
1593
new_deletes = reversed(list(
1594
self._iter_child_entries(1, old_path_utf8)))
1513
1595
# Remove the current contents of the tree at orig_path, and
1514
1596
# reinsert at the correct new path.
1515
1597
for entry in new_deletes:
1517
source_path = entry[0][0] + '/' + entry[0][1]
1598
child_dirname, child_basename, child_file_id = entry[0]
1600
source_path = child_dirname + '/' + child_basename
1519
source_path = entry[0][1]
1602
source_path = child_basename
1520
1603
if new_path_utf8:
1521
1604
target_path = new_path_utf8 + source_path[len(old_path):]
1523
1606
if old_path == '':
1524
1607
raise AssertionError("cannot rename directory to"
1526
1609
target_path = source_path[len(old_path) + 1:]
1527
1610
adds.append((None, target_path, entry[0][2], entry[1][1], False))
1528
1611
deletes.append(
1529
1612
(source_path, target_path, entry[0][2], None, False))
1531
(encode(old_path), new_path, file_id, None, False))
1533
# changes to just the root should not require remove/insertion
1535
changes.append((encode(old_path), encode(new_path), file_id,
1536
inv_to_entry(inv_entry)))
1613
deletes.append((old_path_utf8, new_path, file_id, None, False))
1537
1614
self._check_delta_ids_absent(new_ids, delta, 1)
1539
1616
# Finish expunging deletes/first half of renames.
1597
1673
# Adds are accumulated partly from renames, so can be in any input
1598
1674
# order - sort it.
1675
# TODO: we may want to sort in dirblocks order. That way each entry
1676
# will end up in the same directory, allowing the _get_entry
1677
# fast-path for looking up 2 items in the same dir work.
1678
adds.sort(key=lambda x: x[1])
1600
1679
# adds is now in lexographic order, which places all parents before
1601
1680
# their children, so we can process it linearly.
1682
st = static_tuple.StaticTuple
1603
1683
for old_path, new_path, file_id, new_details, real_add in adds:
1604
# the entry for this file_id must be in tree 0.
1605
entry = self._get_entry(0, file_id, new_path)
1606
if entry[0] is None or entry[0][2] != file_id:
1607
self._changes_aborted = True
1608
raise errors.InconsistentDelta(new_path, file_id,
1609
'working tree does not contain new entry')
1610
if real_add and entry[1][1][0] not in absent:
1611
self._changes_aborted = True
1612
raise errors.InconsistentDelta(new_path, file_id,
1613
'The entry was considered to be a genuinely new record,'
1614
' but there was already an old record for it.')
1615
# We don't need to update the target of an 'r' because the handling
1616
# of renames turns all 'r' situations into a delete at the original
1618
entry[1][1] = new_details
1684
dirname, basename = osutils.split(new_path)
1685
entry_key = st(dirname, basename, file_id)
1686
block_index, present = self._find_block_index_from_key(entry_key)
1688
self._raise_invalid(new_path, file_id,
1689
"Unable to find block for this record."
1690
" Was the parent added?")
1691
block = self._dirblocks[block_index][1]
1692
entry_index, present = self._find_entry_index(entry_key, block)
1694
if old_path is not None:
1695
self._raise_invalid(new_path, file_id,
1696
'considered a real add but still had old_path at %s'
1699
entry = block[entry_index]
1700
basis_kind = entry[1][1][0]
1701
if basis_kind == 'a':
1702
entry[1][1] = new_details
1703
elif basis_kind == 'r':
1704
raise NotImplementedError()
1706
self._raise_invalid(new_path, file_id,
1707
"An entry was marked as a new add"
1708
" but the basis target already existed")
1710
# The exact key was not found in the block. However, we need to
1711
# check if there is a key next to us that would have matched.
1712
# We only need to check 2 locations, because there are only 2
1714
for maybe_index in range(entry_index-1, entry_index+1):
1715
if maybe_index < 0 or maybe_index >= len(block):
1717
maybe_entry = block[maybe_index]
1718
if maybe_entry[0][:2] != (dirname, basename):
1719
# Just a random neighbor
1721
if maybe_entry[0][2] == file_id:
1722
raise AssertionError(
1723
'_find_entry_index didnt find a key match'
1724
' but walking the data did, for %s'
1726
basis_kind = maybe_entry[1][1][0]
1727
if basis_kind not in 'ar':
1728
self._raise_invalid(new_path, file_id,
1729
"we have an add record for path, but the path"
1730
" is already present with another file_id %s"
1731
% (maybe_entry[0][2],))
1733
entry = (entry_key, [DirState.NULL_PARENT_DETAILS,
1735
block.insert(entry_index, entry)
1737
active_kind = entry[1][0][0]
1738
if active_kind == 'a':
1739
# The active record shows up as absent, this could be genuine,
1740
# or it could be present at some other location. We need to
1742
id_index = self._get_id_index()
1743
# The id_index may not be perfectly accurate for tree1, because
1744
# we haven't been keeping it updated. However, it should be
1745
# fine for tree0, and that gives us enough info for what we
1747
keys = id_index.get(file_id, ())
1749
block_i, entry_i, d_present, f_present = \
1750
self._get_block_entry_index(key[0], key[1], 0)
1753
active_entry = self._dirblocks[block_i][1][entry_i]
1754
if (active_entry[0][2] != file_id):
1755
# Some other file is at this path, we don't need to
1758
real_active_kind = active_entry[1][0][0]
1759
if real_active_kind in 'ar':
1760
# We found a record, which was not *this* record,
1761
# which matches the file_id, but is not actually
1762
# present. Something seems *really* wrong.
1763
self._raise_invalid(new_path, file_id,
1764
"We found a tree0 entry that doesnt make sense")
1765
# Now, we've found a tree0 entry which matches the file_id
1766
# but is at a different location. So update them to be
1768
active_dir, active_name = active_entry[0][:2]
1770
active_path = active_dir + '/' + active_name
1772
active_path = active_name
1773
active_entry[1][1] = st('r', new_path, 0, False, '')
1774
entry[1][0] = st('r', active_path, 0, False, '')
1775
elif active_kind == 'r':
1776
raise NotImplementedError()
1778
new_kind = new_details[0]
1780
self._ensure_block(block_index, entry_index, new_path)
1620
1782
def _update_basis_apply_changes(self, changes):
1621
1783
"""Apply a sequence of changes to tree 1 during update_basis_by_delta.
1653
1809
null = DirState.NULL_PARENT_DETAILS
1654
1810
for old_path, new_path, file_id, _, real_delete in deletes:
1655
1811
if real_delete != (new_path is None):
1656
self._changes_aborted = True
1657
raise AssertionError("bad delete delta")
1812
self._raise_invalid(old_path, file_id, "bad delete delta")
1658
1813
# the entry for this file_id must be in tree 1.
1659
1814
dirname, basename = osutils.split(old_path)
1660
1815
block_index, entry_index, dir_present, file_present = \
1661
1816
self._get_block_entry_index(dirname, basename, 1)
1662
1817
if not file_present:
1663
self._changes_aborted = True
1664
raise errors.InconsistentDelta(old_path, file_id,
1818
self._raise_invalid(old_path, file_id,
1665
1819
'basis tree does not contain removed entry')
1666
1820
entry = self._dirblocks[block_index][1][entry_index]
1821
# The state of the entry in the 'active' WT
1822
active_kind = entry[1][0][0]
1667
1823
if entry[0][2] != file_id:
1668
self._changes_aborted = True
1669
raise errors.InconsistentDelta(old_path, file_id,
1824
self._raise_invalid(old_path, file_id,
1670
1825
'mismatched file_id in tree 1')
1672
if entry[1][0][0] != 'a':
1673
self._changes_aborted = True
1674
raise errors.InconsistentDelta(old_path, file_id,
1675
'This was marked as a real delete, but the WT state'
1676
' claims that it still exists and is versioned.')
1827
old_kind = entry[1][1][0]
1828
if active_kind in 'ar':
1829
# The active tree doesn't have this file_id.
1830
# The basis tree is changing this record. If this is a
1831
# rename, then we don't want the record here at all
1832
# anymore. If it is just an in-place change, we want the
1833
# record here, but we'll add it if we need to. So we just
1835
if active_kind == 'r':
1836
active_path = entry[1][0][1]
1837
active_entry = self._get_entry(0, file_id, active_path)
1838
if active_entry[1][1][0] != 'r':
1839
self._raise_invalid(old_path, file_id,
1840
"Dirstate did not have matching rename entries")
1841
elif active_entry[1][0][0] in 'ar':
1842
self._raise_invalid(old_path, file_id,
1843
"Dirstate had a rename pointing at an inactive"
1845
active_entry[1][1] = null
1677
1846
del self._dirblocks[block_index][1][entry_index]
1848
# This was a directory, and the active tree says it
1849
# doesn't exist, and now the basis tree says it doesn't
1850
# exist. Remove its dirblock if present
1852
present) = self._find_block_index_from_key(
1855
dir_block = self._dirblocks[dir_block_index][1]
1857
# This entry is empty, go ahead and just remove it
1858
del self._dirblocks[dir_block_index]
1679
if entry[1][0][0] == 'a':
1680
self._changes_aborted = True
1681
raise errors.InconsistentDelta(old_path, file_id,
1682
'The entry was considered a rename, but the source path'
1683
' is marked as absent.')
1684
# For whatever reason, we were asked to rename an entry
1685
# that was originally marked as deleted. This could be
1686
# because we are renaming the parent directory, and the WT
1687
# current state has the file marked as deleted.
1688
elif entry[1][0][0] == 'r':
1689
# implement the rename
1690
del self._dirblocks[block_index][1][entry_index]
1692
# it is being resurrected here, so blank it out temporarily.
1693
self._dirblocks[block_index][1][entry_index][1][1] = null
1860
# There is still an active record, so just mark this
1863
block_i, entry_i, d_present, f_present = \
1864
self._get_block_entry_index(old_path, '', 1)
1866
dir_block = self._dirblocks[block_i][1]
1867
for child_entry in dir_block:
1868
child_basis_kind = child_entry[1][1][0]
1869
if child_basis_kind not in 'ar':
1870
self._raise_invalid(old_path, file_id,
1871
"The file id was deleted but its children were "
1695
1874
def _after_delta_check_parents(self, parents, index):
1696
1875
"""Check that parents required by the delta are all intact.
2145
2327
def _get_id_index(self):
2146
"""Get an id index of self._dirblocks."""
2328
"""Get an id index of self._dirblocks.
2330
This maps from file_id => [(directory, name, file_id)] entries where
2331
that file_id appears in one of the trees.
2147
2333
if self._id_index is None:
2149
2335
for key, tree_details in self._iter_entries():
2150
id_index.setdefault(key[2], set()).add(key)
2336
self._add_to_id_index(id_index, key)
2151
2337
self._id_index = id_index
2152
2338
return self._id_index
2340
def _add_to_id_index(self, id_index, entry_key):
2341
"""Add this entry to the _id_index mapping."""
2342
# This code used to use a set for every entry in the id_index. However,
2343
# it is *rare* to have more than one entry. So a set is a large
2344
# overkill. And even when we do, we won't ever have more than the
2345
# number of parent trees. Which is still a small number (rarely >2). As
2346
# such, we use a simple tuple, and do our own uniqueness checks. While
2347
# the 'in' check is O(N) since N is nicely bounded it shouldn't ever
2348
# cause quadratic failure.
2349
file_id = entry_key[2]
2350
entry_key = static_tuple.StaticTuple.from_sequence(entry_key)
2351
if file_id not in id_index:
2352
id_index[file_id] = static_tuple.StaticTuple(entry_key,)
2354
entry_keys = id_index[file_id]
2355
if entry_key not in entry_keys:
2356
id_index[file_id] = entry_keys + (entry_key,)
2358
def _remove_from_id_index(self, id_index, entry_key):
2359
"""Remove this entry from the _id_index mapping.
2361
It is an programming error to call this when the entry_key is not
2364
file_id = entry_key[2]
2365
entry_keys = list(id_index[file_id])
2366
entry_keys.remove(entry_key)
2367
id_index[file_id] = static_tuple.StaticTuple.from_sequence(entry_keys)
2154
2369
def _get_output_lines(self, lines):
2155
2370
"""Format lines for final output.
2465
2725
new_details = []
2466
2726
for lookup_index in xrange(tree_index):
2467
2727
# boundary case: this is the first occurence of file_id
2468
# so there are no id_indexs, possibly take this out of
2728
# so there are no id_indexes, possibly take this out of
2470
if not len(id_index[file_id]):
2730
if not len(entry_keys):
2471
2731
new_details.append(DirState.NULL_PARENT_DETAILS)
2473
2733
# grab any one entry, use it to find the right path.
2474
# TODO: optimise this to reduce memory use in highly
2475
# fragmented situations by reusing the relocation
2477
a_key = iter(id_index[file_id]).next()
2734
a_key = iter(entry_keys).next()
2478
2735
if by_path[a_key][lookup_index][0] in ('r', 'a'):
2479
# its a pointer or missing statement, use it as is.
2736
# its a pointer or missing statement, use it as
2480
2738
new_details.append(by_path[a_key][lookup_index])
2482
2740
# we have the right key, make a pointer to it.
2483
2741
real_path = ('/'.join(a_key[0:2])).strip('/')
2484
new_details.append(('r', real_path, 0, False, ''))
2742
new_details.append(st('r', real_path, 0, False,
2485
2744
new_details.append(self._inv_entry_to_details(entry))
2486
2745
new_details.extend(new_location_suffix)
2487
2746
by_path[new_entry_key] = new_details
2488
id_index[file_id].add(new_entry_key)
2747
self._add_to_id_index(id_index, new_entry_key)
2489
2748
# --- end generation of full tree mappings
2491
2750
# sort and output all the entries
added
removed
bzrlib/dirstate.py
