def from_line(cls, line):
""":return: New RefLogEntry instance from the given revlog line.
:param line: line without trailing newline
:raise ValueError: If line could not be parsed"""
try:
info, msg = line.split("\t", 2)
except ValueError:
raise ValueError("line is missing tab separator")
# END handle first plit
oldhexsha = info[:40]
newhexsha = info[41:81]
for hexsha in (oldhexsha, newhexsha):
if not cls._re_hexsha_only.match(hexsha):
raise ValueError("Invalid hexsha: %s" % hexsha)
# END if hexsha re doesn't match
# END for each hexsha
email_end = info.find(">", 82)
if email_end == -1:
raise ValueError("Missing token: >")
# END handle missing end brace
actor = Actor._from_string(info[82 : email_end + 1])
time, tz_offset = parse_date(info[email_end + 2 :])
return RefLogEntry((oldhexsha, newhexsha, actor, (time, tz_offset), msg))
def parse_actor_and_date(line): """Parse out the actor (author or committer) info from a line like:: author Tom Preston-Werner <*****@*****.**> 1191999972 -0700 :return: [Actor, int_seconds_since_epoch, int_timezone_offset]""" m = _re_actor_epoch.search(line) actor, epoch, offset = m.groups() return (Actor._from_string(actor), int(epoch), utctz_to_altz(offset))
def append_entry(cls, config_reader, filepath, oldbinsha, newbinsha, message):
"""Append a new log entry to the revlog at filepath.
:param config_reader: configuration reader of the repository - used to obtain
user information. May be None
:param filepath: full path to the log file
:param oldbinsha: binary sha of the previous commit
:param newbinsha: binary sha of the current commit
:param message: message describing the change to the reference
:param write: If True, the changes will be written right away. Otherwise
the change will not be written
:return: RefLogEntry objects which was appended to the log
:note: As we are append-only, concurrent access is not a problem as we
do not interfere with readers."""
if len(oldbinsha) != 20 or len(newbinsha) != 20:
raise ValueError("Shas need to be given in binary format")
# END handle sha type
assure_directory_exists(filepath, is_file=True)
entry = RefLogEntry(
(
bin_to_hex(oldbinsha),
bin_to_hex(newbinsha),
Actor.committer(config_reader),
(int(time.time()), time.altzone),
message,
)
)
lf = LockFile(filepath)
lf._obtain_lock_or_raise()
fd = open(filepath, "a")
try:
fd.write(repr(entry))
finally:
fd.close()
lf._release_lock()
# END handle write operation
return entry
def create_from_tree(cls, repo, tree, message, parent_commits=None, head=False):
"""Commit the given tree, creating a commit object.
:param repo: Repo object the commit should be part of
:param tree: Tree object or hex or bin sha
the tree of the new commit
:param message: Commit message. It may be an empty string if no message is provided.
It will be converted to a string in any case.
:param parent_commits:
Optional Commit objects to use as parents for the new commit.
If empty list, the commit will have no parents at all and become
a root commit.
If None , the current head commit will be the parent of the
new commit object
:param head:
If True, the HEAD will be advanced to the new commit automatically.
Else the HEAD will remain pointing on the previous commit. This could
lead to undesired results when diffing files.
:return: Commit object representing the new commit
:note:
Additional information about the committer and Author are taken from the
environment or from the pygit configuration, see git-commit-tree for
more information"""
parents = parent_commits
if parent_commits is None:
try:
parent_commits = [ repo.head.commit ]
except ValueError:
# empty repositories have no head commit
parent_commits = list()
# END handle parent commits
# END if parent commits are unset
# retrieve all additional information, create a commit object, and
# serialize it
# Generally:
# * Environment variables override configuration values
# * Sensible defaults are set according to the pygit documentation
# COMMITER AND AUTHOR INFO
cr = repo.config_reader()
env = os.environ
committer = Actor.committer(cr)
author = Actor.author(cr)
# PARSE THE DATES
unix_time = int(time())
offset = altzone
author_date_str = env.get(cls.env_author_date, '')
if author_date_str:
author_time, author_offset = parse_date(author_date_str)
else:
author_time, author_offset = unix_time, offset
# END set author time
committer_date_str = env.get(cls.env_committer_date, '')
if committer_date_str:
committer_time, committer_offset = parse_date(committer_date_str)
else:
committer_time, committer_offset = unix_time, offset
# END set committer time
# assume utf8 encoding
enc_section, enc_option = cls.conf_encoding.split('.')
conf_encoding = cr.get_value(enc_section, enc_option, cls.default_encoding)
# if the tree is no object, make sure we create one - otherwise
# the created commit object is invalid
if isinstance(tree, str):
tree = repo.tree(tree)
# END tree conversion
# CREATE NEW COMMIT
new_commit = cls(repo, cls.NULL_BIN_SHA, tree,
author, author_time, author_offset,
committer, committer_time, committer_offset,
message, parent_commits, conf_encoding)
stream = StringIO()
new_commit._serialize(stream)
streamlen = stream.tell()
stream.seek(0)
istream = repo.odb.store(IStream(cls.type, streamlen, stream))
new_commit.binsha = istream.binsha
if head:
# need late import here, importing pygit at the very beginning throws
# as well ...
import pygit.refs
try:
repo.head.set_commit(new_commit, logmsg="commit: %s" % message)
except ValueError:
# head is not yet set to the ref our HEAD points to
# Happens on first commit
import pygit.refs
master = pygit.refs.Head.create(repo, repo.head.ref, new_commit, logmsg="commit (initial): %s" % message)
repo.head.set_reference(master, logmsg='commit: Switching to %s' % master)
# END handle empty repositories
# END advance head handling
return new_commit
def blame(self, rev, file):
"""The blame information for the given file at the given revision.
:parm rev: revision specifier, see pygit-rev-parse for viable options.
:return:
list: [git.Commit, list: [<line>]]
A list of tuples associating a Commit object with a list of lines that
changed within the given commit. The Commit objects will be given in order
of appearance."""
data = self.git.blame(rev, '--', file, p=True)
commits = dict()
blames = list()
info = None
for line in data.splitlines(False):
parts = self.re_whitespace.split(line, 1)
firstpart = parts[0]
if self.re_hexsha_only.search(firstpart):
# handles
# 634396b2f541a9f2d58b00be1a07f0c358b999b3 1 1 7 - indicates blame-data start
# 634396b2f541a9f2d58b00be1a07f0c358b999b3 2 2
digits = parts[-1].split(" ")
if len(digits) == 3:
info = {'id': firstpart}
blames.append([None, []])
# END blame data initialization
else:
m = self.re_author_committer_start.search(firstpart)
if m:
# handles:
# author Tom Preston-Werner
# author-mail <*****@*****.**>
# author-time 1192271832
# author-tz -0700
# committer Tom Preston-Werner
# committer-mail <*****@*****.**>
# committer-time 1192271832
# committer-tz -0700 - IGNORED BY US
role = m.group(0)
if firstpart.endswith('-mail'):
info["%s_email" % role] = parts[-1]
elif firstpart.endswith('-time'):
info["%s_date" % role] = int(parts[-1])
elif role == firstpart:
info[role] = parts[-1]
# END distinguish mail,time,name
else:
# handle
# filename lib/grit.rb
# summary add Blob
# <and rest>
if firstpart.startswith('filename'):
info['filename'] = parts[-1]
elif firstpart.startswith('summary'):
info['summary'] = parts[-1]
elif firstpart == '':
if info:
sha = info['id']
c = commits.get(sha)
if c is None:
c = Commit( self, hex_to_bin(sha),
author=Actor._from_string(info['author'] + ' ' + info['author_email']),
authored_date=info['author_date'],
committer=Actor._from_string(info['committer'] + ' ' + info['committer_email']),
committed_date=info['committer_date'],
message=info['summary'])
commits[sha] = c
# END if commit objects needs initial creation
m = self.re_tab_full_line.search(line)
text, = m.groups()
blames[-1][0] = c
blames[-1][1].append( text )
info = None
# END if we collected commit info
# END distinguish filename,summary,rest
# END distinguish author|committer vs filename,summary,rest
# END distinguish hexsha vs other information
return blames