def test_hungarian(self):
matrices = [
# Square
([[400, 150, 400],
[400, 450, 600],
[300, 225, 300]],
850 # expected cost
),
# Rectangular variant
([[400, 150, 400, 1],
[400, 450, 600, 2],
[300, 225, 300, 3]],
452 # expected cost
),
# Square
([[10, 10, 8],
[ 9, 8, 1],
[ 9, 7, 4]],
18
),
# Rectangular variant
([[10, 10, 8, 11],
[ 9, 8, 1, 1],
[ 9, 7, 4, 10]],
15
),
]
for matrix, expected_cost in matrices:
indexes = lapacho.hungarian(matrix)
total_cost = 0
for i, j in indexes:
total_cost += matrix[i][j]
self.assertEqual(expected_cost, total_cost)
def build(repo, store, options):
for c in repo.walk(repo.head.target, GIT_SORT_REVERSE):
logger.info("commit %s", c.hex)
# Store commit
commit = store.find(Commit, Commit.hex == c.hex.decode("utf-8")).one()
if commit is None:
commit = store.add(Commit(c.hex.decode("utf-8"), c.commit_time))
store.flush()
# Check commit message
for i in test_regexp(c.message):
bug = store.find(Bug, Bug.bug_no == i).one()
if not bug:
bug = store.add(Bug(i))
bug.commits.add(commit)
logger.debug("bug %d", i)
store.flush()
# Check parents
for p in c.parents:
parent = store.find(Commit, Commit.hex == p.hex.decode("utf-8")).one()
if parent is None:
parent = store.add(Commit(p.hex.decode("utf-8"), p.commit_time))
commit.parents.add(parent)
logger.debug("parent %s", p.hex)
diff = p.tree.diff_to_tree(c.tree, GIT_DIFF_IGNORE_WHITESPACE_EOL)
for p in [p for p in diff if splitext(p.old_file_path)[1] in [".c", ".h"]]:
logger.debug("path %s", p.old_file_path)
old_file = store.find(File, File.path == p.old_file_path.decode("utf-8")).one()
if old_file is None:
old_file = store.add(File(p.old_file_path.decode("utf-8")))
new_file = store.find(File, File.path == p.new_file_path.decode("utf-8")).one()
if new_file is None:
new_file = store.add(File(p.new_file_path.decode("utf-8")))
store.flush()
for h in p.hunks:
old_data = [l for l in h.lines if l[0] != "+"]
new_data = [l for l in h.lines if l[0] != "-"]
logger.debug("hunk old:%d-%d, new:%d-%d", h.old_start, h.old_lines, h.new_start, h.new_lines)
deletions = [(l[1], i) for i, l in enumerate(old_data) if l[0] == "-"]
additions = [(l[1], i) for i, l in enumerate(new_data) if l[0] == "+"]
# Identify changed lines
changed_lines = []
if len(deletions) != 0 and len(additions) != 0:
d = map(lambda (x, y): levenshtein_distance(x[0], y[0], True), product(deletions, additions))
step = len(additions)
matrix = [d[x : x + step] for x in xrange(0, len(d), step)]
indexes = hungarian(matrix)
changed_lines = [
(deletions[i][1], additions[j][1])
for i, j in indexes
if 0.0 < matrix[i][j] and matrix[i][j] < 0.4
]
# Store all changed lines
for x, y in changed_lines:
store.add(Edit(commit, old_file, new_file, h.old_start + x, h.new_start + y))
# Store all deleted lines
for l, x in deletions:
if x not in [i for i, j in changed_lines]:
store.add(Edit(commit, old_file, new_file, h.old_start + x, None))
# Store all added lines
for l, y in additions:
if y not in [j for i, j in changed_lines]:
store.add(Edit(commit, old_file, new_file, None, h.new_start + y))
store.flush()
# Save the changes
store.commit()
def build(repo, store, options):
for c in repo.walk(repo.head.oid, GIT_SORT_REVERSE):
logger.info('commit %s', c.hex)
# Store commit
commit = store.find(Commit, Commit.hex == c.hex.decode('utf-8')).one()
if commit is None:
commit = store.add(Commit(c.hex.decode('utf-8'), c.commit_time))
store.flush()
# Check commit message
for i in test_regexp(c.message):
bug = store.find(Bug, Bug.bug_no == i).one()
if not bug: bug = store.add(Bug(i))
bug.commits.add(commit)
logger.debug('bug %d', i)
store.flush()
# Check parents
for p in c.parents:
parent = store.find(Commit, Commit.hex == p.hex.decode('utf-8')).one()
if parent is None:
parent = store.add(Commit(p.hex.decode('utf-8'), p.commit_time))
commit.parents.add(parent)
logger.debug('parent %s', p.hex)
diff = p.tree.diff(c.tree)
if 'hunks' not in diff.changes:
continue
for h in [f for f in diff.changes['hunks'] if splitext(f.old_file)[1] in ['.c', '.h']]:
old_file = store.find(File, File.path == h.old_file.decode('utf-8')).one()
if old_file is None:
old_file = store.add(File(h.old_file.decode('utf-8')))
new_file = store.find(File, File.path == h.new_file.decode('utf-8')).one()
if new_file is None:
new_file = store.add(File(h.new_file.decode('utf-8')))
store.flush()
old_data = [l for l in h.data if l[1] != GIT_DIFF_LINE_ADDITION]
new_data = [l for l in h.data if l[1] != GIT_DIFF_LINE_DELETION]
logger.debug('hunk old:%d-%d, new:%d-%d', h.old_start, h.old_lines, h.new_start, h.new_lines)
deletions = [(l[0], i) for i, l in enumerate(old_data) if l[1] == GIT_DIFF_LINE_DELETION]
additions = [(l[0], i) for i, l in enumerate(new_data) if l[1] == GIT_DIFF_LINE_ADDITION]
# Identify changed lines
changed_lines = []
if len(deletions) != 0 and len(additions) != 0:
d = map(lambda (x, y): levenshtein_distance(x[0], y[0], True), product(deletions, additions))
step = len(additions)
matrix = [d[x:x+step] for x in xrange(0, len(d), step)]
indexes = hungarian(matrix)
changed_lines = [(deletions[i][1], additions[j][1]) for i, j in indexes if 0.0 < matrix[i][j] and matrix[i][j] < 0.4]
# Store all changed lines
for x, y in changed_lines:
store.add(Edit(commit, old_file, new_file, h.old_start + x, h.new_start + y))
# Store all deleted lines
for l, x in deletions:
if x not in [i for i, j in changed_lines]:
store.add(Edit(commit, old_file, new_file, h.old_start + x, None))
# Store all added lines
for l, y in additions:
if y not in [j for i, j in changed_lines]:
store.add(Edit(commit, old_file, new_file, None, h.new_start + y))
store.flush()
# Save the changes
store.commit()