( )⚙ D11973 simplemerge: convert `merge_lines()` away from generator

This is an archive of the discontinued Mercurial Phabricator instance.

Differential D11973

simplemerge: convert `merge_lines()` away from generator
ClosedPublic

Authored by martinvonz on Jan 12 2022, 1:55 AM.

Details

Reviewers
Alphare
Group Reviewers
hg-reviewers
Commits
rHGce8c82a5cd65: simplemerge: convert `merge_lines()` away from generator
rHG3f5bc446357a: simplemerge: convert `merge_lines()` away from generator
Summary

We always consume all the lines and put them in a list anyway. By
making the function not a generator, we can later make it return an
additional value (to indicate if there were conflicts).

Diff Detail

Repository
rHG Mercurial
Lint
Automatic diff as part of commit; lint not applicable.
Unit
Automatic diff as part of commit; unit tests not applicable.

Revision Contents
Changeset List

Diff 31667

mercurial/debugcommands.py

p1 = repo[ps[0]] p1 = repo[ps[0]]
if len(ps) > 1: if len(ps) > 1:
p2 = repo[ps[1]] p2 = repo[ps[1]]
pa = p1.ancestor(p2) pa = p1.ancestor(p2)
base, local, other = [ base, local, other = [
x[fn].data() for x in (pa, p1, p2) x[fn].data() for x in (pa, p1, p2)
] ]
m3 = simplemerge.Merge3Text(base, local, other) m3 = simplemerge.Merge3Text(base, local, other)
ml = [l.strip() for l in m3.merge_lines()] ml = [l.strip() for l in m3.merge_lines()[0]]
ml.append(b"") ml.append(b"")
elif at > 0: elif at > 0:
ml = p1[fn].data().split(b"\n") ml = p1[fn].data().split(b"\n")
else: else:
ml = initialmergedlines ml = initialmergedlines
ml[id * linesperrev] += b" r%i" % id ml[id * linesperrev] += b" r%i" % id
mergedtext = b"\n".join(ml) mergedtext = b"\n".join(ml)
files.append(fn) files.append(fn)

mercurial/simplemerge.py

name_base=None, name_base=None,
start_marker=b'<<<<<<<', start_marker=b'<<<<<<<',
mid_marker=b'=======', mid_marker=b'=======',
end_marker=b'>>>>>>>', end_marker=b'>>>>>>>',
base_marker=None, base_marker=None,
minimize=False, minimize=False,
): ):
"""Return merge in cvs-like form.""" """Return merge in cvs-like form."""
self.conflicts = False conflicts = False
newline = b'\n' newline = b'\n'
if len(self.a) > 0: if len(self.a) > 0:
if self.a[0].endswith(b'\r\n'): if self.a[0].endswith(b'\r\n'):
newline = b'\r\n' newline = b'\r\n'
elif self.a[0].endswith(b'\r'): elif self.a[0].endswith(b'\r'):
newline = b'\r' newline = b'\r'
if name_a and start_marker: if name_a and start_marker:
start_marker = start_marker + b' ' + name_a start_marker = start_marker + b' ' + name_a
if name_b and end_marker: if name_b and end_marker:
end_marker = end_marker + b' ' + name_b end_marker = end_marker + b' ' + name_b
if name_base and base_marker: if name_base and base_marker:
base_marker = base_marker + b' ' + name_base base_marker = base_marker + b' ' + name_base
merge_groups = self.merge_groups() merge_groups = self.merge_groups()
if minimize: if minimize:
merge_groups = self.minimize(merge_groups) merge_groups = self.minimize(merge_groups)
for what, lines in merge_groups: lines = []
for what, group_lines in merge_groups:
if what == b'conflict': if what == b'conflict':
base_lines, a_lines, b_lines = lines base_lines, a_lines, b_lines = group_lines
self.conflicts = True conflicts = True
if start_marker is not None: if start_marker is not None:
yield start_marker + newline lines.append(start_marker + newline)
for line in a_lines: lines.extend(a_lines)
yield line
if base_marker is not None: if base_marker is not None:
yield base_marker + newline lines.append(base_marker + newline)
for line in base_lines: lines.extend(base_lines)
yield line
if mid_marker is not None: if mid_marker is not None:
yield mid_marker + newline lines.append(mid_marker + newline)
for line in b_lines: lines.extend(b_lines)
yield line
if end_marker is not None: if end_marker is not None:
yield end_marker + newline lines.append(end_marker + newline)
else: else:
for line in lines: lines.extend(group_lines)
yield line return lines, conflicts
def merge_groups(self): def merge_groups(self):
"""Yield sequence of line groups. Each one is a tuple: """Yield sequence of line groups. Each one is a tuple:
'unchanged', lines 'unchanged', lines
Lines unchanged from base Lines unchanged from base
'a', lines 'a', lines
else: else:
extrakwargs = { extrakwargs = {
'minimize': True, 'minimize': True,
} }
if mode == b'merge3': if mode == b'merge3':
extrakwargs['base_marker'] = b'|||||||' extrakwargs['base_marker'] = b'|||||||'
extrakwargs['name_base'] = name_base extrakwargs['name_base'] = name_base
extrakwargs['minimize'] = False extrakwargs['minimize'] = False
lines = list( lines, conflicts = m3.merge_lines(
m3.merge_lines(name_a=name_a, name_b=name_b, **extrakwargs) name_a=name_a, name_b=name_b, **extrakwargs
) )
conflicts = m3.conflicts
mergedtext = b''.join(lines) mergedtext = b''.join(lines)
if opts.get('print'): if opts.get('print'):
ui.fout.write(mergedtext) ui.fout.write(mergedtext)
else: else:
# localctx.flags() already has the merged flags (done in # localctx.flags() already has the merged flags (done in
# mergestate.resolve()) # mergestate.resolve())
localctx.write(mergedtext, localctx.flags()) localctx.write(mergedtext, localctx.flags())
if conflicts: if conflicts:
return 1 return 1

tests/test-simplemerge.py

def test_null_insert(self): def test_null_insert(self):
m3 = Merge3([], [b'aaa', b'bbb'], []) m3 = Merge3([], [b'aaa', b'bbb'], [])
# todo: should use a sentinel at end as from get_matching_blocks # todo: should use a sentinel at end as from get_matching_blocks
# to match without zz # to match without zz
self.assertEqual(list(m3.find_sync_regions()), [(0, 0, 2, 2, 0, 0)]) self.assertEqual(list(m3.find_sync_regions()), [(0, 0, 2, 2, 0, 0)])
self.assertEqual(list(m3.merge_regions()), [(b'a', 0, 2)]) self.assertEqual(list(m3.merge_regions()), [(b'a', 0, 2)])
self.assertEqual(list(m3.merge_lines()), [b'aaa', b'bbb']) self.assertEqual(m3.merge_lines(), ([b'aaa', b'bbb'], False))
def test_no_conflicts(self): def test_no_conflicts(self):
"""No conflicts because only one side changed""" """No conflicts because only one side changed"""
m3 = Merge3( m3 = Merge3(
[b'aaa', b'bbb'], [b'aaa', b'111', b'bbb'], [b'aaa', b'bbb'] [b'aaa', b'bbb'], [b'aaa', b'111', b'bbb'], [b'aaa', b'bbb']
) )
self.assertEqual( self.assertEqual(
list(m3.find_sync_regions()), list(m3.find_sync_regions()),
[(0, 1, 0, 1, 0, 1), (1, 2, 2, 3, 1, 2), (2, 2, 3, 3, 2, 2)], [(0, 1, 0, 1, 0, 1), (1, 2, 2, 3, 1, 2), (2, 2, 3, 3, 2, 2)],
) )
self.assertEqual( self.assertEqual(
list(m3.merge_regions()), list(m3.merge_regions()),
[(b'unchanged', 0, 1), (b'a', 1, 2), (b'unchanged', 1, 2)], [(b'unchanged', 0, 1), (b'a', 1, 2), (b'unchanged', 1, 2)],
) )
def test_append_a(self): def test_append_a(self):
m3 = Merge3( m3 = Merge3(
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
[b'aaa\n', b'bbb\n', b'222\n'], [b'aaa\n', b'bbb\n', b'222\n'],
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
) )
self.assertEqual(b''.join(m3.merge_lines()), b'aaa\nbbb\n222\n') self.assertEqual(b''.join(m3.merge_lines()[0]), b'aaa\nbbb\n222\n')
def test_append_b(self): def test_append_b(self):
m3 = Merge3( m3 = Merge3(
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
[b'aaa\n', b'bbb\n', b'222\n'], [b'aaa\n', b'bbb\n', b'222\n'],
) )
self.assertEqual(b''.join(m3.merge_lines()), b'aaa\nbbb\n222\n') self.assertEqual(b''.join(m3.merge_lines()[0]), b'aaa\nbbb\n222\n')
def test_append_agreement(self): def test_append_agreement(self):
m3 = Merge3( m3 = Merge3(
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
[b'aaa\n', b'bbb\n', b'222\n'], [b'aaa\n', b'bbb\n', b'222\n'],
[b'aaa\n', b'bbb\n', b'222\n'], [b'aaa\n', b'bbb\n', b'222\n'],
) )
self.assertEqual(b''.join(m3.merge_lines()), b'aaa\nbbb\n222\n') self.assertEqual(b''.join(m3.merge_lines()[0]), b'aaa\nbbb\n222\n')
def test_append_clash(self): def test_append_clash(self):
m3 = Merge3( m3 = Merge3(
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
[b'aaa\n', b'bbb\n', b'222\n'], [b'aaa\n', b'bbb\n', b'222\n'],
[b'aaa\n', b'bbb\n', b'333\n'], [b'aaa\n', b'bbb\n', b'333\n'],
) )
ml = m3.merge_lines( ml, conflicts = m3.merge_lines(
name_a=b'a', name_a=b'a',
name_b=b'b', name_b=b'b',
start_marker=b'<<', start_marker=b'<<',
mid_marker=b'--', mid_marker=b'--',
end_marker=b'>>', end_marker=b'>>',
) )
self.assertEqual( self.assertEqual(
b''.join(ml), b''.join(ml),
b'aaa\n' b'bbb\n' b'<< a\n' b'222\n' b'--\n' b'333\n' b'>> b\n', b'aaa\n' b'bbb\n' b'<< a\n' b'222\n' b'--\n' b'333\n' b'>> b\n',
) )
def test_insert_agreement(self): def test_insert_agreement(self):
m3 = Merge3( m3 = Merge3(
[b'aaa\n', b'bbb\n'], [b'aaa\n', b'bbb\n'],
[b'aaa\n', b'222\n', b'bbb\n'], [b'aaa\n', b'222\n', b'bbb\n'],
[b'aaa\n', b'222\n', b'bbb\n'], [b'aaa\n', b'222\n', b'bbb\n'],
) )
ml = m3.merge_lines( ml, conflicts = m3.merge_lines(
name_a=b'a', name_a=b'a',
name_b=b'b', name_b=b'b',
start_marker=b'<<', start_marker=b'<<',
mid_marker=b'--', mid_marker=b'--',
end_marker=b'>>', end_marker=b'>>',
) )
self.assertEqual(b''.join(ml), b'aaa\n222\nbbb\n') self.assertEqual(b''.join(ml), b'aaa\n222\nbbb\n')
list(m3.merge_groups()), list(m3.merge_groups()),
[ [
(b'unchanged', [b'aaa\n']), (b'unchanged', [b'aaa\n']),
(b'conflict', ([], [b'111\n'], [b'222\n'])), (b'conflict', ([], [b'111\n'], [b'222\n'])),
(b'unchanged', [b'bbb\n']), (b'unchanged', [b'bbb\n']),
], ],
) )
ml = m3.merge_lines( ml, conflicts = m3.merge_lines(
name_a=b'a', name_a=b'a',
name_b=b'b', name_b=b'b',
start_marker=b'<<', start_marker=b'<<',
mid_marker=b'--', mid_marker=b'--',
end_marker=b'>>', end_marker=b'>>',
) )
self.assertEqual( self.assertEqual(
b''.join(ml), b''.join(ml),
self.assertEqual( self.assertEqual(
list(m3.find_sync_regions()), list(m3.find_sync_regions()),
[(0, 1, 0, 1, 0, 1), (3, 4, 4, 5, 5, 6), (4, 4, 5, 5, 6, 6)], [(0, 1, 0, 1, 0, 1), (3, 4, 4, 5, 5, 6), (4, 4, 5, 5, 6, 6)],
) )
def test_merge_poem(self): def test_merge_poem(self):
"""Test case from diff3 manual""" """Test case from diff3 manual"""
m3 = Merge3(TZU, LAO, TAO) m3 = Merge3(TZU, LAO, TAO)
ml = list(m3.merge_lines(b'LAO', b'TAO')) ml, conflicts = m3.merge_lines(b'LAO', b'TAO')
self.log(b'merge result:') self.log(b'merge result:')
self.log(b''.join(ml)) self.log(b''.join(ml))
self.assertEqual(ml, MERGED_RESULT) self.assertEqual(ml, MERGED_RESULT)
def test_binary(self): def test_binary(self):
with self.assertRaises(error.Abort): with self.assertRaises(error.Abort):
Merge3([b'\x00'], [b'a'], [b'b']) Merge3([b'\x00'], [b'a'], [b'b'])
def test_dos_text(self): def test_dos_text(self):
base_text = b'a\r\n' base_text = b'a\r\n'
this_text = b'b\r\n' this_text = b'b\r\n'
other_text = b'c\r\n' other_text = b'c\r\n'
m3 = Merge3( m3 = Merge3(
base_text.splitlines(True), base_text.splitlines(True),
other_text.splitlines(True), other_text.splitlines(True),
this_text.splitlines(True), this_text.splitlines(True),
) )
m_lines = m3.merge_lines(b'OTHER', b'THIS') m_lines, conflicts = m3.merge_lines(b'OTHER', b'THIS')
self.assertEqual( self.assertEqual(
b'<<<<<<< OTHER\r\nc\r\n=======\r\nb\r\n' b'<<<<<<< OTHER\r\nc\r\n=======\r\nb\r\n'
b'>>>>>>> THIS\r\n'.splitlines(True), b'>>>>>>> THIS\r\n'.splitlines(True),
list(m_lines), m_lines,
) )
def test_mac_text(self): def test_mac_text(self):
base_text = b'a\r' base_text = b'a\r'
this_text = b'b\r' this_text = b'b\r'
other_text = b'c\r' other_text = b'c\r'
m3 = Merge3( m3 = Merge3(
base_text.splitlines(True), base_text.splitlines(True),
other_text.splitlines(True), other_text.splitlines(True),
this_text.splitlines(True), this_text.splitlines(True),
) )
m_lines = m3.merge_lines(b'OTHER', b'THIS') m_lines, conflicts = m3.merge_lines(b'OTHER', b'THIS')
self.assertEqual( self.assertEqual(
b'<<<<<<< OTHER\rc\r=======\rb\r' b'<<<<<<< OTHER\rc\r=======\rb\r'
b'>>>>>>> THIS\r'.splitlines(True), b'>>>>>>> THIS\r'.splitlines(True),
list(m_lines), m_lines,
) )
if __name__ == '__main__': if __name__ == '__main__':
import silenttestrunner import silenttestrunner
silenttestrunner.main(__name__) silenttestrunner.main(__name__)