tests/test-linelog.py
author Raphaël Gomès <rgomes@octobus.net>
Tue, 05 Apr 2022 10:55:28 +0200
branchstable
changeset 49000 dd6b67d5c256
parent 45942 89a2afe31e82
child 48875 6000f5b25c9b
permissions -rw-r--r--
rust: fix unsound `OwningDirstateMap` As per the previous patch, `OwningDirstateMap` is unsound. Self-referential structs are difficult to implement correctly in Rust since the compiler is free to move structs around as much as it wants to. They are also very rarely needed in practice, so the state-of-the-art on how they should be done within the Rust rules is still a bit new. The crate `ouroboros` is an attempt at providing a safe way (in the Rust sense) of declaring self-referential structs. It is getting a lot attention and was improved very quickly when soundness issues were found in the past: rather than relying on our own (limited) review circle, we might as well use the de-facto common crate to fix this problem. This will give us a much better chance of finding issues should any new ones be discovered as well as the benefit of fewer `unsafe` APIs of our own. I was starting to think about how I would present a safe API to the old struct but soon realized that the callback-based approach was already done in `ouroboros`, along with a lot more care towards refusing incorrect structs. In short: we don't return a mutable reference to the `DirstateMap` anymore, we expect users of its API to pass a `FnOnce` that takes the map as an argument. This allows our `OwningDirstateMap` to control the input and output lifetimes of the code that modifies it to prevent such issues. Changing to `ouroboros` meant changing every API with it, but it is relatively low churn in the end. It correctly identified the example buggy modification of `copy_map_insert` outlined in the previous patch as violating the borrow rules. Differential Revision: https://phab.mercurial-scm.org/D12429

from __future__ import absolute_import, print_function

import difflib
import random
import unittest

from mercurial import linelog

vecratio = 3  # number of replacelines / number of replacelines_vec
maxlinenum = 0xFFFFFF
maxb1 = 0xFFFFFF
maxdeltaa = 10
maxdeltab = 10


def _genedits(seed, endrev):
    lines = []
    random.seed(seed)
    for rev in range(0, endrev):
        n = len(lines)
        a1 = random.randint(0, n)
        a2 = random.randint(a1, min(n, a1 + maxdeltaa))
        b1 = random.randint(0, maxb1)
        b2 = random.randint(b1, b1 + maxdeltab)
        usevec = not bool(random.randint(0, vecratio))
        if usevec:
            blines = [
                (random.randint(0, rev), random.randint(0, maxlinenum))
                for _ in range(b1, b2)
            ]
        else:
            blines = [(rev, bidx) for bidx in range(b1, b2)]
        lines[a1:a2] = blines
        yield lines, rev, a1, a2, b1, b2, blines, usevec


class linelogtests(unittest.TestCase):
    def testlinelogencodedecode(self):
        program = [
            linelog._eof(0, 0),
            linelog._jge(41, 42),
            linelog._jump(0, 43),
            linelog._eof(0, 0),
            linelog._jl(44, 45),
            linelog._line(46, 47),
        ]
        ll = linelog.linelog(program, maxrev=100)
        enc = ll.encode()
        # round-trips okay
        self.assertEqual(linelog.linelog.fromdata(enc)._program, ll._program)
        self.assertEqual(linelog.linelog.fromdata(enc), ll)
        # This encoding matches the encoding used by hg-experimental's
        # linelog file, or is supposed to if it doesn't.
        self.assertEqual(
            enc,
            (
                b'\x00\x00\x01\x90\x00\x00\x00\x06'
                b'\x00\x00\x00\xa4\x00\x00\x00*'
                b'\x00\x00\x00\x00\x00\x00\x00+'
                b'\x00\x00\x00\x00\x00\x00\x00\x00'
                b'\x00\x00\x00\xb1\x00\x00\x00-'
                b'\x00\x00\x00\xba\x00\x00\x00/'
            ),
        )

    def testsimpleedits(self):
        ll = linelog.linelog()
        # Initial revision: add lines 0, 1, and 2
        ll.replacelines(1, 0, 0, 0, 3)
        self.assertEqual(
            [(l.rev, l.linenum) for l in ll.annotate(1)],
            [
                (1, 0),
                (1, 1),
                (1, 2),
            ],
        )
        # Replace line 1 with a new line
        ll.replacelines(2, 1, 2, 1, 2)
        self.assertEqual(
            [(l.rev, l.linenum) for l in ll.annotate(2)],
            [
                (1, 0),
                (2, 1),
                (1, 2),
            ],
        )
        # delete a line out of 2
        ll.replacelines(3, 1, 2, 0, 0)
        self.assertEqual(
            [(l.rev, l.linenum) for l in ll.annotate(3)],
            [
                (1, 0),
                (1, 2),
            ],
        )
        # annotation of 1 is unchanged
        self.assertEqual(
            [(l.rev, l.linenum) for l in ll.annotate(1)],
            [
                (1, 0),
                (1, 1),
                (1, 2),
            ],
        )
        ll.annotate(3)  # set internal state to revision 3
        start = ll.getoffset(0)
        end = ll.getoffset(1)
        self.assertEqual(
            ll.getalllines(start, end),
            [
                (1, 0),
                (2, 1),
                (1, 1),
            ],
        )
        self.assertEqual(
            ll.getalllines(),
            [
                (1, 0),
                (2, 1),
                (1, 1),
                (1, 2),
            ],
        )

    def testparseclinelogfile(self):
        # This data is what the replacements in testsimpleedits
        # produce when fed to the original linelog.c implementation.
        data = (
            b'\x00\x00\x00\x0c\x00\x00\x00\x0f'
            b'\x00\x00\x00\x00\x00\x00\x00\x02'
            b'\x00\x00\x00\x05\x00\x00\x00\x06'
            b'\x00\x00\x00\x06\x00\x00\x00\x00'
            b'\x00\x00\x00\x00\x00\x00\x00\x07'
            b'\x00\x00\x00\x06\x00\x00\x00\x02'
            b'\x00\x00\x00\x00\x00\x00\x00\x00'
            b'\x00\x00\x00\t\x00\x00\x00\t'
            b'\x00\x00\x00\x00\x00\x00\x00\x0c'
            b'\x00\x00\x00\x08\x00\x00\x00\x05'
            b'\x00\x00\x00\x06\x00\x00\x00\x01'
            b'\x00\x00\x00\x00\x00\x00\x00\x05'
            b'\x00\x00\x00\x0c\x00\x00\x00\x05'
            b'\x00\x00\x00\n\x00\x00\x00\x01'
            b'\x00\x00\x00\x00\x00\x00\x00\t'
        )
        llc = linelog.linelog.fromdata(data)
        self.assertEqual(
            [(l.rev, l.linenum) for l in llc.annotate(1)],
            [
                (1, 0),
                (1, 1),
                (1, 2),
            ],
        )
        self.assertEqual(
            [(l.rev, l.linenum) for l in llc.annotate(2)],
            [
                (1, 0),
                (2, 1),
                (1, 2),
            ],
        )
        self.assertEqual(
            [(l.rev, l.linenum) for l in llc.annotate(3)],
            [
                (1, 0),
                (1, 2),
            ],
        )
        # Check we emit the same bytecode.
        ll = linelog.linelog()
        # Initial revision: add lines 0, 1, and 2
        ll.replacelines(1, 0, 0, 0, 3)
        # Replace line 1 with a new line
        ll.replacelines(2, 1, 2, 1, 2)
        # delete a line out of 2
        ll.replacelines(3, 1, 2, 0, 0)
        diff = '\n   ' + '\n   '.join(
            difflib.unified_diff(
                ll.debugstr().splitlines(),
                llc.debugstr().splitlines(),
                'python',
                'c',
                lineterm='',
            )
        )
        self.assertEqual(ll._program, llc._program, 'Program mismatch: ' + diff)
        # Done as a secondary step so we get a better result if the
        # program is where the mismatch is.
        self.assertEqual(ll, llc)
        self.assertEqual(ll.encode(), data)

    def testanothersimplecase(self):
        ll = linelog.linelog()
        ll.replacelines(3, 0, 0, 0, 2)
        ll.replacelines(4, 0, 2, 0, 0)
        self.assertEqual([(l.rev, l.linenum) for l in ll.annotate(4)], [])
        self.assertEqual(
            [(l.rev, l.linenum) for l in ll.annotate(3)], [(3, 0), (3, 1)]
        )
        # rev 2 is empty because contents were only ever introduced in rev 3
        self.assertEqual([(l.rev, l.linenum) for l in ll.annotate(2)], [])

    def testrandomedits(self):
        # Inspired by original linelog tests.
        seed = random.random()
        numrevs = 2000
        ll = linelog.linelog()
        # Populate linelog
        for lines, rev, a1, a2, b1, b2, blines, usevec in _genedits(
            seed, numrevs
        ):
            if usevec:
                ll.replacelines_vec(rev, a1, a2, blines)
            else:
                ll.replacelines(rev, a1, a2, b1, b2)
            ll.annotate(rev)
            self.assertEqual(ll.annotateresult, lines)
        # Verify we can get back these states by annotating each rev
        for lines, rev, a1, a2, b1, b2, blines, usevec in _genedits(
            seed, numrevs
        ):
            ar = ll.annotate(rev)
            self.assertEqual([(l.rev, l.linenum) for l in ar], lines)

    def testinfinitebadprogram(self):
        ll = linelog.linelog.fromdata(
            b'\x00\x00\x00\x00\x00\x00\x00\x02'  # header
            b'\x00\x00\x00\x00\x00\x00\x00\x01'  # JUMP to self
        )
        with self.assertRaises(linelog.LineLogError):
            # should not be an infinite loop and raise
            ll.annotate(1)


if __name__ == '__main__':
    import silenttestrunner

    silenttestrunner.main(__name__)