mercurial/utils/storageutil.py
author Gregory Szorc <gregory.szorc@gmail.com>
Mon, 24 Sep 2018 14:31:31 -0700
changeset 39878 3e896b51aa5d
parent 39877 f8eb71f9e3bd
child 39879 d269ddbf54f0
permissions -rw-r--r--
storageutil: move metadata parsing and packing from revlog (API) Parsing and writing of revision text metadata is likely identical across storage backends. Let's move the code out of revlog so we don't need to import the revlog module in order to use it. Differential Revision: https://phab.mercurial-scm.org/D4754
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
39877
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     1
# storageutil.py - Storage functionality agnostic of backend implementation.
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     2
#
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     3
# Copyright 2018 Gregory Szorc <gregory.szorc@gmail.com>
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     4
#
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     5
# This software may be used and distributed according to the terms of the
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     6
# GNU General Public License version 2 or any later version.
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     7
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     8
from __future__ import absolute_import
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
     9
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    10
import hashlib
39878
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    11
import re
39877
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    12
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    13
from ..node import (
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    14
    nullid,
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    15
)
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    16
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    17
_nullhash = hashlib.sha1(nullid)
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    18
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    19
def hashrevisionsha1(text, p1, p2):
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    20
    """Compute the SHA-1 for revision data and its parents.
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    21
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    22
    This hash combines both the current file contents and its history
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    23
    in a manner that makes it easy to distinguish nodes with the same
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    24
    content in the revision graph.
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    25
    """
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    26
    # As of now, if one of the parent node is null, p2 is null
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    27
    if p2 == nullid:
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    28
        # deep copy of a hash is faster than creating one
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    29
        s = _nullhash.copy()
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    30
        s.update(p1)
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    31
    else:
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    32
        # none of the parent nodes are nullid
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    33
        if p1 < p2:
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    34
            a = p1
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    35
            b = p2
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    36
        else:
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    37
            a = p2
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    38
            b = p1
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    39
        s = hashlib.sha1(a)
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    40
        s.update(b)
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    41
    s.update(text)
f8eb71f9e3bd storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff changeset
    42
    return s.digest()
39878
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    43
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    44
METADATA_RE = re.compile(b'\x01\n')
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    45
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    46
def parsemeta(text):
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    47
    """Parse metadata header from revision data.
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    48
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    49
    Returns a 2-tuple of (metadata, offset), where both can be None if there
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    50
    is no metadata.
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    51
    """
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    52
    # text can be buffer, so we can't use .startswith or .index
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    53
    if text[:2] != b'\x01\n':
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    54
        return None, None
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    55
    s = METADATA_RE.search(text, 2).start()
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    56
    mtext = text[2:s]
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    57
    meta = {}
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    58
    for l in mtext.splitlines():
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    59
        k, v = l.split(b': ', 1)
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    60
        meta[k] = v
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    61
    return meta, s + 2
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    62
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    63
def packmeta(meta, text):
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    64
    """Add metadata to fulltext to produce revision text."""
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    65
    keys = sorted(meta)
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    66
    metatext = b''.join(b'%s: %s\n' % (k, meta[k]) for k in keys)
3e896b51aa5d storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents: 39877
diff changeset
    67
    return b'\x01\n%s\x01\n%s' % (metatext, text)