Mercurial > public > mercurial-scm > hg

comparison mercurial/revlog.py @ 47253:b876f0bf7366

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
revlog: introduce a plain compression mode That mode is simple it means the chunk contains uncompressed data and can be used directly. Differential Revision: https://phab.mercurial-scm.org/D10650
author Pierre-Yves David <pierre-yves.david@octobus.net>
date Mon, 03 May 2021 19:46:25 +0200
parents e340b556a13e
children eac3591abbf4
comparison
equal deleted inserted replaced
47252:e340b556a13e 47253:b876f0bf7366
34 from .i18n import _ 34 from .i18n import _
35 from .pycompat import getattr 35 from .pycompat import getattr
36 from .revlogutils.constants import ( 36 from .revlogutils.constants import (
37 ALL_KINDS, 37 ALL_KINDS,
38 COMP_MODE_INLINE, 38 COMP_MODE_INLINE,
39 COMP_MODE_PLAIN,
39 FEATURES_BY_VERSION, 40 FEATURES_BY_VERSION,
40 FLAG_GENERALDELTA, 41 FLAG_GENERALDELTA,
41 FLAG_INLINE_DATA, 42 FLAG_INLINE_DATA,
42 INDEX_HEADER, 43 INDEX_HEADER,
43 REVLOGV0, 44 REVLOGV0,
1755 to be used for reading. If used, the seek position of the file will not 1756 to be used for reading. If used, the seek position of the file will not
1756 be preserved. 1757 be preserved.
1757 1758
1758 Returns a str holding uncompressed data for the requested revision. 1759 Returns a str holding uncompressed data for the requested revision.
1759 """ 1760 """
1760 return self.decompress(self._getsegmentforrevs(rev, rev, df=df)[1]) 1761 compression_mode = self.index[rev][10]
1762 data = self._getsegmentforrevs(rev, rev, df=df)[1]
1763 if compression_mode == COMP_MODE_PLAIN:
1764 return data
1765 elif compression_mode == COMP_MODE_INLINE:
1766 return self.decompress(data)
1767 else:
1768 msg = 'unknown compression mode %d'
1769 msg %= compression_mode
1770 raise error.RevlogError(msg)
1761 1771
1762 def _chunks(self, revs, df=None, targetsize=None): 1772 def _chunks(self, revs, df=None, targetsize=None):
1763 """Obtain decompressed chunks for the specified revisions. 1773 """Obtain decompressed chunks for the specified revisions.
1764 1774
1765 Accepts an iterable of numeric revisions that are assumed to be in 1775 Accepts an iterable of numeric revisions that are assumed to be in
1808 for rev in revschunk: 1818 for rev in revschunk:
1809 chunkstart = start(rev) 1819 chunkstart = start(rev)
1810 if inline: 1820 if inline:
1811 chunkstart += (rev + 1) * iosize 1821 chunkstart += (rev + 1) * iosize
1812 chunklength = length(rev) 1822 chunklength = length(rev)
1823 comp_mode = self.index[rev][10]
1813 c = buffer(data, chunkstart - offset, chunklength) 1824 c = buffer(data, chunkstart - offset, chunklength)
1814 ladd(decomp(c)) 1825 if comp_mode == COMP_MODE_PLAIN:
1826 ladd(c)
1827 elif comp_mode == COMP_MODE_INLINE:
1828 ladd(decomp(c))
1829 else:
1830 msg = 'unknown compression mode %d'
1831 msg %= comp_mode
1832 raise error.RevlogError(msg)
1815 1833
1816 return l 1834 return l
1817 1835
1818 def _chunkclear(self): 1836 def _chunkclear(self):
1819 """Clear the raw chunk cache.""" 1837 """Clear the raw chunk cache."""
2459 2477
2460 revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags) 2478 revinfo = _revisioninfo(node, p1, p2, btext, textlen, cachedelta, flags)
2461 2479
2462 deltainfo = deltacomputer.finddeltainfo(revinfo, fh) 2480 deltainfo = deltacomputer.finddeltainfo(revinfo, fh)
2463 2481
2482 compression_mode = COMP_MODE_INLINE
2483 if self._docket is not None:
2484 h, d = deltainfo.data
2485 if not h and not d:
2486 # not data to store at all... declare them uncompressed
2487 compression_mode = COMP_MODE_PLAIN
2488 elif not h and d[0:1] == b'\0':
2489 compression_mode = COMP_MODE_PLAIN
2490 elif h == b'u':
2491 # we have a more efficient way to declare uncompressed
2492 h = b''
2493 compression_mode = COMP_MODE_PLAIN
2494 deltainfo = deltautil.drop_u_compression(deltainfo)
2495
2464 if sidedata and self.hassidedata: 2496 if sidedata and self.hassidedata:
2465 serialized_sidedata = sidedatautil.serialize_sidedata(sidedata) 2497 serialized_sidedata = sidedatautil.serialize_sidedata(sidedata)
2466 sidedata_offset = offset + deltainfo.deltalen 2498 sidedata_offset = offset + deltainfo.deltalen
2467 else: 2499 else:
2468 serialized_sidedata = b"" 2500 serialized_sidedata = b""
2480 p1r, 2512 p1r,
2481 p2r, 2513 p2r,
2482 node, 2514 node,
2483 sidedata_offset, 2515 sidedata_offset,
2484 len(serialized_sidedata), 2516 len(serialized_sidedata),
2485 COMP_MODE_INLINE, 2517 compression_mode,
2486 ) 2518 )
2487 2519
2488 self.index.append(e) 2520 self.index.append(e)
2489 entry = self.index.entry_binary(curr) 2521 entry = self.index.entry_binary(curr)
2490 if curr == 0 and self._docket is None: 2522 if curr == 0 and self._docket is None: