Mercurial > public > mercurial-scm > hg-stable

comparison mercurial/revlog.py @ 42808:e91411fcc697

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
revlog: split `rawtext` retrieval out of _revisiondata This part is reasonably independent. Having it on its own clarify the code flow and will help code that inherit from revlog to overwrite specific area only.
author Pierre-Yves David <pierre-yves.david@octobus.net>
date Tue, 20 Aug 2019 17:12:36 +0200
parents 90f5dfc9c42a
children 268662aac075
comparison
equal deleted inserted replaced
42807:90f5dfc9c42a 42808:e91411fcc697
1621 1621
1622 # fast path the special `nullid` rev 1622 # fast path the special `nullid` rev
1623 if node == nullid: 1623 if node == nullid:
1624 return "" 1624 return ""
1625 1625
1626 # revision in the cache (could be useful to apply delta)
1627 cachedrev = None
1628 # the revlog's flag for this revision
1629 # (usually alter its state or content)
1630 flags = None
1631 # The text as stored inside the revlog. Might be the revision or might 1626 # The text as stored inside the revlog. Might be the revision or might
1632 # need to be processed to retrieve the revision. 1627 # need to be processed to retrieve the revision.
1633 rawtext = None 1628 rawtext = None
1629
1630 rev, rawtext, validated = self._rawtext(node, rev, _df=_df)
1631
1632 if raw and validated:
1633 # if we don't want to process the raw text and that raw
1634 # text is cached, we can exit early.
1635 return rawtext
1636 if rev is None:
1637 rev = self.rev(node)
1638 # the revlog's flag for this revision
1639 # (usually alter its state or content)
1640 flags = self.flags(rev)
1641
1642 if validated and flags == REVIDX_DEFAULT_FLAGS:
1643 # no extra flags set, no flag processor runs, text = rawtext
1644 return rawtext
1645
1646 text, validatehash = self._processflags(rawtext, flags, 'read', raw=raw)
1647 if validatehash:
1648 self.checkhash(text, node, rev=rev)
1649 if not validated:
1650 self._revisioncache = (node, rev, rawtext)
1651
1652 return text
1653
1654 def _rawtext(self, node, rev, _df=None):
1655 """return the possibly unvalidated rawtext for a revision
1656
1657 returns (rev, rawtext, validated)
1658 """
1659
1660 # revision in the cache (could be useful to apply delta)
1661 cachedrev = None
1634 # An intermediate text to apply deltas to 1662 # An intermediate text to apply deltas to
1635 basetext = None 1663 basetext = None
1636 # Do we need to update the rawtext cache once it is validated ?
1637 needcaching = True
1638 1664
1639 # Check if we have the entry in cache 1665 # Check if we have the entry in cache
1640 # The cache entry looks like (node, rev, rawtext) 1666 # The cache entry looks like (node, rev, rawtext)
1641 if self._revisioncache: 1667 if self._revisioncache:
1642 if self._revisioncache[0] == node: 1668 if self._revisioncache[0] == node:
1643 needcaching = False 1669 return (rev, self._revisioncache[2], True)
1644 # _cache only stores rawtext
1645 # rawtext is reusable. but we might need to run flag processors
1646 rawtext = self._revisioncache[2]
1647 if raw:
1648 # if we don't want to process the raw text and that raw
1649 # text is cached, we can exit early.
1650 return rawtext
1651 # duplicated, but good for perf
1652 if rev is None:
1653 rev = self.rev(node)
1654 if flags is None:
1655 flags = self.flags(rev)
1656 # no extra flags set, no flag processor runs, text = rawtext
1657 if flags == REVIDX_DEFAULT_FLAGS:
1658 return rawtext
1659
1660 cachedrev = self._revisioncache[1] 1670 cachedrev = self._revisioncache[1]
1661 1671
1662 # look up what we need to read 1672 if rev is None:
1663 if rawtext is None: 1673 rev = self.rev(node)
1664 if rev is None: 1674
1665 rev = self.rev(node) 1675 chain, stopped = self._deltachain(rev, stoprev=cachedrev)
1666 1676 if stopped:
1667 chain, stopped = self._deltachain(rev, stoprev=cachedrev) 1677 basetext = self._revisioncache[2]
1668 if stopped: 1678
1669 basetext = self._revisioncache[2] 1679 # drop cache to save memory, the caller is expected to
1670 1680 # update self._revisioncache after validating the text
1671 # drop cache to save memory 1681 self._revisioncache = None
1672 self._revisioncache = None 1682
1673 1683 targetsize = None
1674 targetsize = None 1684 rawsize = self.index[rev][2]
1675 rawsize = self.index[rev][2] 1685 if 0 <= rawsize:
1676 if 0 <= rawsize: 1686 targetsize = 4 * rawsize
1677 targetsize = 4 * rawsize 1687
1678 1688 bins = self._chunks(chain, df=_df, targetsize=targetsize)
1679 bins = self._chunks(chain, df=_df, targetsize=targetsize) 1689 if basetext is None:
1680 if basetext is None: 1690 basetext = bytes(bins[0])
1681 basetext = bytes(bins[0]) 1691 bins = bins[1:]
1682 bins = bins[1:] 1692
1683 1693 rawtext = mdiff.patches(basetext, bins)
1684 rawtext = mdiff.patches(basetext, bins) 1694 del basetext # let us have a chance to free memory early
1685 del basetext # let us have a chance to free memory early 1695 return (rev, rawtext, False)
1686
1687 if flags is None:
1688 if rev is None:
1689 rev = self.rev(node)
1690 flags = self.flags(rev)
1691
1692 text, validatehash = self._processflags(rawtext, flags, 'read', raw=raw)
1693 if validatehash:
1694 self.checkhash(text, node, rev=rev)
1695
1696 if needcaching:
1697 self._revisioncache = (node, rev, rawtext)
1698
1699 return text
1700 1696
1701 def rawdata(self, nodeorrev, _df=None): 1697 def rawdata(self, nodeorrev, _df=None):
1702 """return an uncompressed raw data of a given node or revision number. 1698 """return an uncompressed raw data of a given node or revision number.
1703 1699
1704 _df - an existing file handle to read from. (internal-only) 1700 _df - an existing file handle to read from. (internal-only)