def blame_incremental(self, rev, file, **kwargs):
"""Iterator for blame information for the given file at the given revision.
Unlike .blame(), this does not return the actual file's contents, only
a stream of (commit, range) tuples.
:parm rev: revision specifier, see git-rev-parse for viable options.
:return: lazy iterator of (git.Commit, range) tuples, where the commit
indicates the commit to blame for the line, and range
indicates a span of line numbers in the resulting file.
If you combine all line number ranges outputted by this command, you
should get a continuous range spanning all line numbers in the file.
"""
data = self.git.blame(rev, '--', file, p=True, incremental=True, stdout_as_string=False, **kwargs)
commits = dict()
stream = iter(data.splitlines())
while True:
line = next(stream) # when exhausted, casues a StopIteration, terminating this function
hexsha, _, lineno, num_lines = line.split()
lineno = int(lineno)
num_lines = int(num_lines)
if hexsha not in commits:
# Now read the next few lines and build up a dict of properties
# for this commit
props = dict()
while True:
line = next(stream)
if line == b'boundary':
# "boundary" indicates a root commit and occurs
# instead of the "previous" tag
continue
tag, value = line.split(b' ', 1)
props[tag] = value
if tag == b'filename':
# "filename" formally terminates the entry for --incremental
break
c = Commit(self, hex_to_bin(hexsha),
author=Actor(safe_decode(props[b'author']),
safe_decode(props[b'author-mail'].lstrip(b'<').rstrip(b'>'))),
authored_date=int(props[b'author-time']),
committer=Actor(safe_decode(props[b'committer']),
safe_decode(props[b'committer-mail'].lstrip(b'<').rstrip(b'>'))),
committed_date=int(props[b'committer-time']),
message=safe_decode(props[b'summary']))
commits[hexsha] = c
else:
# Discard the next line (it's a filename end tag)
line = next(stream)
assert line.startswith(b'filename'), 'Unexpected git blame output'
yield commits[hexsha], range(lineno, lineno + num_lines)
def run_commit_hook(name, index, *args):
"""Run the commit hook of the given name. Silently ignores hooks that do not exist.
:param name: name of hook, like 'pre-commit'
:param index: IndexFile instance
:param args: arguments passed to hook file
:raises HookExecutionError: """
hp = hook_path(name, index.repo.git_dir)
if not os.access(hp, os.X_OK):
return
env = os.environ.copy()
env['GIT_INDEX_FILE'] = safe_decode(index.path) if PY3 else safe_encode(index.path)
env['GIT_EDITOR'] = ':'
try:
cmd = subprocess.Popen([hp] + list(args),
env=env,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
cwd=index.repo.working_dir,
close_fds=is_posix,
creationflags=PROC_CREATIONFLAGS,)
except Exception as ex:
raise HookExecutionError(hp, ex)
else:
stdout = []
stderr = []
handle_process_output(cmd, stdout.append, stderr.append, finalize_process)
stdout = ''.join(stdout)
stderr = ''.join(stderr)
if cmd.returncode != 0:
stdout = force_text(stdout, defenc)
stderr = force_text(stderr, defenc)
raise HookExecutionError(hp, cmd.returncode, stderr, stdout)
def run_commit_hook(name, index):
"""Run the commit hook of the given name. Silently ignores hooks that do not exist.
:param name: name of hook, like 'pre-commit'
:param index: IndexFile instance
:raises HookExecutionError: """
hp = hook_path(name, index.repo.git_dir)
if not os.access(hp, os.X_OK):
return
env = os.environ.copy()
env['GIT_INDEX_FILE'] = safe_decode(index.path) if PY3 else safe_encode(index.path)
env['GIT_EDITOR'] = ':'
try:
cmd = subprocess.Popen(hp,
env=env,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
cwd=index.repo.working_dir,
close_fds=is_posix,
creationflags=PROC_CREATIONFLAGS,)
except Exception as ex:
raise HookExecutionError(hp, ex)
else:
stdout = []
stderr = []
handle_process_output(cmd, stdout.append, stderr.append, finalize_process)
stdout = ''.join(stdout)
stderr = ''.join(stderr)
if cmd.returncode != 0:
stdout = force_text(stdout, defenc)
stderr = force_text(stderr, defenc)
raise HookExecutionError(hp, cmd.returncode, stdout, stderr)
def __init__(self,
command: Union[List[str], Tuple[str, ...], str],
status: Union[str, int, None, Exception] = None,
stderr: Union[bytes, str, None] = None,
stdout: Union[bytes, str, None] = None) -> None:
if not isinstance(command, (tuple, list)):
command = command.split()
self.command = command
self.status = status
if status:
if isinstance(status, Exception):
status = "%s('%s')" % (type(status).__name__,
safe_decode(str(status)))
else:
try:
status = 'exit code(%s)' % int(status)
except (ValueError, TypeError):
s = safe_decode(str(status))
status = "'%s'" % s if isinstance(status, str) else s
self._cmd = safe_decode(command[0])
self._cmdline = ' '.join(safe_decode(i) for i in command)
self._cause = status and " due to: %s" % status or "!"
stdout_decode = safe_decode(stdout)
stderr_decode = safe_decode(stderr)
self.stdout = stdout_decode and "\n stdout: '%s'" % stdout_decode or ''
self.stderr = stderr_decode and "\n stderr: '%s'" % stderr_decode or ''
def run_commit_hook(name: str, index: 'IndexFile', *args: str) -> None:
"""Run the commit hook of the given name. Silently ignores hooks that do not exist.
:param name: name of hook, like 'pre-commit'
:param index: IndexFile instance
:param args: arguments passed to hook file
:raises HookExecutionError: """
hp = hook_path(name, index.repo.git_dir)
if not os.access(hp, os.X_OK):
return None
env = os.environ.copy()
env['GIT_INDEX_FILE'] = safe_decode(str(index.path))
env['GIT_EDITOR'] = ':'
cmd = [hp]
try:
if is_win and not _has_file_extension(hp):
# Windows only uses extensions to determine how to open files
# (doesn't understand shebangs). Try using bash to run the hook.
relative_hp = Path(hp).relative_to(
index.repo.working_dir).as_posix()
cmd = ["bash.exe", relative_hp]
cmd = subprocess.Popen(
cmd + list(args),
env=env,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
cwd=index.repo.working_dir,
close_fds=is_posix,
creationflags=PROC_CREATIONFLAGS,
)
except Exception as ex:
raise HookExecutionError(hp, ex) from ex
else:
stdout_list: List[str] = []
stderr_list: List[str] = []
handle_process_output(cmd, stdout_list.append, stderr_list.append,
finalize_process)
stdout = ''.join(stdout_list)
stderr = ''.join(stderr_list)
if cmd.returncode != 0:
stdout = force_text(stdout, defenc)
stderr = force_text(stderr, defenc)
raise HookExecutionError(hp, cmd.returncode, stderr, stdout)
def tree_entries_from_data(data):
"""Reads the binary representation of a tree and returns tuples of Tree items
:param data: data block with tree data (as bytes)
:return: list(tuple(binsha, mode, tree_relative_path), ...)"""
ord_zero = ord('0')
space_ord = ord(' ')
len_data = len(data)
i = 0
out = list()
while i < len_data:
mode = 0
# read mode
# Some git versions truncate the leading 0, some don't
# The type will be extracted from the mode later
while byte_ord(data[i]) != space_ord:
# move existing mode integer up one level being 3 bits
# and add the actual ordinal value of the character
mode = (mode << 3) + (byte_ord(data[i]) - ord_zero)
i += 1
# END while reading mode
# byte is space now, skip it
i += 1
# parse name, it is NULL separated
ns = i
while byte_ord(data[i]) != 0:
i += 1
# END while not reached NULL
# default encoding for strings in git is utf8
# Only use the respective unicode object if the byte stream was encoded
name = data[ns:i]
name = safe_decode(name)
# byte is NULL, get next 20
i += 1
sha = data[i:i + 20]
i = i + 20
out.append((sha, mode, name))
# END for each byte in data stream
return out
def tree_entries_from_data(data):
"""Reads the binary representation of a tree and returns tuples of Tree items
:param data: data block with tree data (as bytes)
:return: list(tuple(binsha, mode, tree_relative_path), ...)"""
ord_zero = ord('0')
space_ord = ord(' ')
len_data = len(data)
i = 0
out = list()
while i < len_data:
mode = 0
# read mode
# Some git versions truncate the leading 0, some don't
# The type will be extracted from the mode later
while byte_ord(data[i]) != space_ord:
# move existing mode integer up one level being 3 bits
# and add the actual ordinal value of the character
mode = (mode << 3) + (byte_ord(data[i]) - ord_zero)
i += 1
# END while reading mode
# byte is space now, skip it
i += 1
# parse name, it is NULL separated
ns = i
while byte_ord(data[i]) != 0:
i += 1
# END while not reached NULL
# default encoding for strings in git is utf8
# Only use the respective unicode object if the byte stream was encoded
name = data[ns:i]
name = safe_decode(name)
# byte is NULL, get next 20
i += 1
sha = data[i:i + 20]
i = i + 20
out.append((sha, mode, name))
# END for each byte in data stream
return out
def run_commit_hook(name: str, index: 'IndexFile', *args: str) -> None:
"""Run the commit hook of the given name. Silently ignores hooks that do not exist.
:param name: name of hook, like 'pre-commit'
:param index: IndexFile instance
:param args: arguments passed to hook file
:raises HookExecutionError: """
hp = hook_path(name, index.repo.git_dir)
if not os.access(hp, os.X_OK):
return None
env = os.environ.copy()
env['GIT_INDEX_FILE'] = safe_decode(str(index.path))
env['GIT_EDITOR'] = ':'
try:
cmd = subprocess.Popen(
[hp] + list(args),
env=env,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
cwd=index.repo.working_dir,
close_fds=is_posix,
creationflags=PROC_CREATIONFLAGS,
)
except Exception as ex:
raise HookExecutionError(hp, ex) from ex
else:
stdout_list = [] # type: List[str]
stderr_list = [] # type: List[str]
handle_process_output(cmd, stdout_list.append, stderr_list.append,
finalize_process)
stdout = ''.join(stdout_list)
stderr = ''.join(stderr_list)
if cmd.returncode != 0:
stdout = force_text(stdout, defenc)
stderr = force_text(stderr, defenc)
raise HookExecutionError(hp, cmd.returncode, stderr, stdout)
def __init__(self, command, status=None, stderr=None, stdout=None):
if not isinstance(command, (tuple, list)):
command = command.split()
self.command = command
self.status = status
if status:
if isinstance(status, Exception):
status = u"%s('%s')" % (type(status).__name__, safe_decode(str(status)))
else:
try:
status = u'exit code(%s)' % int(status)
except (ValueError, TypeError):
s = safe_decode(str(status))
status = u"'%s'" % s if isinstance(status, str) else s
self._cmd = safe_decode(command[0])
self._cmdline = u' '.join(safe_decode(i) for i in command)
self._cause = status and u" due to: %s" % status or "!"
self.stdout = stdout and u"\n stdout: '%s'" % safe_decode(stdout) or ''
self.stderr = stderr and u"\n stderr: '%s'" % safe_decode(stderr) or ''
def __init__(self, command, status=None, stderr=None, stdout=None):
if not isinstance(command, (tuple, list)):
command = command.split()
self.command = command
self.status = status
if status:
if isinstance(status, Exception):
status = u"%s('%s')" % (type(status).__name__, safe_decode(str(status)))
else:
try:
status = u'exit code(%s)' % int(status)
except (ValueError, TypeError):
s = safe_decode(str(status))
status = u"'%s'" % s if isinstance(status, string_types) else s
self._cmd = safe_decode(command[0])
self._cmdline = u' '.join(safe_decode(i) for i in command)
self._cause = status and u" due to: %s" % status or "!"
self.stdout = stdout and u"\n stdout: '%s'" % safe_decode(stdout) or ''
self.stderr = stderr and u"\n stderr: '%s'" % safe_decode(stderr) or ''
def blame_incremental(self, rev, file, **kwargs):
"""Iterator for blame information for the given file at the given revision.
Unlike .blame(), this does not return the actual file's contents, only
a stream of BlameEntry tuples.
:parm rev: revision specifier, see git-rev-parse for viable options.
:return: lazy iterator of BlameEntry tuples, where the commit
indicates the commit to blame for the line, and range
indicates a span of line numbers in the resulting file.
If you combine all line number ranges outputted by this command, you
should get a continuous range spanning all line numbers in the file.
"""
data = self.git.blame(rev,
'--',
file,
p=True,
incremental=True,
stdout_as_string=False,
**kwargs)
commits = dict()
stream = (line for line in data.split(b'\n') if line)
while True:
line = next(
stream
) # when exhausted, causes a StopIteration, terminating this function
hexsha, orig_lineno, lineno, num_lines = line.split()
lineno = int(lineno)
num_lines = int(num_lines)
orig_lineno = int(orig_lineno)
if hexsha not in commits:
# Now read the next few lines and build up a dict of properties
# for this commit
props = dict()
while True:
line = next(stream)
if line == b'boundary':
# "boundary" indicates a root commit and occurs
# instead of the "previous" tag
continue
tag, value = line.split(b' ', 1)
props[tag] = value
if tag == b'filename':
# "filename" formally terminates the entry for --incremental
orig_filename = value
break
c = Commit(
self,
hex_to_bin(hexsha),
author=Actor(
safe_decode(props[b'author']),
safe_decode(
props[b'author-mail'].lstrip(b'<').rstrip(b'>'))),
authored_date=int(props[b'author-time']),
committer=Actor(
safe_decode(props[b'committer']),
safe_decode(props[b'committer-mail'].lstrip(
b'<').rstrip(b'>'))),
committed_date=int(props[b'committer-time']))
commits[hexsha] = c
else:
# Discard all lines until we find "filename" which is
# guaranteed to be the last line
while True:
line = next(
stream) # will fail if we reach the EOF unexpectedly
tag, value = line.split(b' ', 1)
if tag == b'filename':
orig_filename = value
break
yield BlameEntry(commits[hexsha], range(lineno,
lineno + num_lines),
safe_decode(orig_filename),
range(orig_lineno, orig_lineno + num_lines))
def as_text(stdout_value):
return not output_stream and safe_decode(stdout_value) or '<OUTPUT_STREAM>'
def execute(self, command,
istream=None,
with_extended_output=False,
with_exceptions=True,
as_process=False,
output_stream=None,
stdout_as_string=True,
kill_after_timeout=None,
with_stdout=True,
universal_newlines=False,
shell=None,
**subprocess_kwargs
):
"""Handles executing the command on the shell and consumes and returns
the returned information (stdout)
:param command:
The command argument list to execute.
It should be a string, or a sequence of program arguments. The
program to execute is the first item in the args sequence or string.
:param istream:
Standard input filehandle passed to subprocess.Popen.
:param with_extended_output:
Whether to return a (status, stdout, stderr) tuple.
:param with_exceptions:
Whether to raise an exception when git returns a non-zero status.
:param as_process:
Whether to return the created process instance directly from which
streams can be read on demand. This will render with_extended_output and
with_exceptions ineffective - the caller will have
to deal with the details himself.
It is important to note that the process will be placed into an AutoInterrupt
wrapper that will interrupt the process once it goes out of scope. If you
use the command in iterators, you should pass the whole process instance
instead of a single stream.
:param output_stream:
If set to a file-like object, data produced by the git command will be
output to the given stream directly.
This feature only has any effect if as_process is False. Processes will
always be created with a pipe due to issues with subprocess.
This merely is a workaround as data will be copied from the
output pipe to the given output stream directly.
Judging from the implementation, you shouldn't use this flag !
:param stdout_as_string:
if False, the commands standard output will be bytes. Otherwise, it will be
decoded into a string using the default encoding (usually utf-8).
The latter can fail, if the output contains binary data.
:param subprocess_kwargs:
Keyword arguments to be passed to subprocess.Popen. Please note that
some of the valid kwargs are already set by this method, the ones you
specify may not be the same ones.
:param with_stdout: If True, default True, we open stdout on the created process
:param universal_newlines:
if True, pipes will be opened as text, and lines are split at
all known line endings.
:param shell:
Whether to invoke commands through a shell (see `Popen(..., shell=True)`).
It overrides :attr:`USE_SHELL` if it is not `None`.
:param kill_after_timeout:
To specify a timeout in seconds for the git command, after which the process
should be killed. This will have no effect if as_process is set to True. It is
set to None by default and will let the process run until the timeout is
explicitly specified. This feature is not supported on Windows. It's also worth
noting that kill_after_timeout uses SIGKILL, which can have negative side
effects on a repository. For example, stale locks in case of git gc could
render the repository incapable of accepting changes until the lock is manually
removed.
:return:
* str(output) if extended_output = False (Default)
* tuple(int(status), str(stdout), str(stderr)) if extended_output = True
if ouput_stream is True, the stdout value will be your output stream:
* output_stream if extended_output = False
* tuple(int(status), output_stream, str(stderr)) if extended_output = True
Note git is executed with LC_MESSAGES="C" to ensure consistent
output regardless of system language.
:raise GitCommandError:
:note:
If you add additional keyword arguments to the signature of this method,
you must update the execute_kwargs tuple housed in this module."""
if self.GIT_PYTHON_TRACE and (self.GIT_PYTHON_TRACE != 'full' or as_process):
log.info(' '.join(command))
# Allow the user to have the command executed in their working dir.
cwd = self._working_dir or os.getcwd()
# Start the process
env = os.environ.copy()
# Attempt to force all output to plain ascii english, which is what some parsing code
# may expect.
# According to stackoverflow (http://goo.gl/l74GC8), we are setting LANGUAGE as well
# just to be sure.
env["LANGUAGE"] = "C"
env["LC_ALL"] = "C"
env.update(self._environment)
if is_win:
cmd_not_found_exception = OSError
if kill_after_timeout:
raise GitCommandError(command, '"kill_after_timeout" feature is not supported on Windows.')
else:
if sys.version_info[0] > 2:
cmd_not_found_exception = FileNotFoundError # NOQA # exists, flake8 unknown @UndefinedVariable
else:
cmd_not_found_exception = OSError
# end handle
stdout_sink = (PIPE
if with_stdout
else getattr(subprocess, 'DEVNULL', None) or open(os.devnull, 'wb'))
log.debug("Popen(%s, cwd=%s, universal_newlines=%s, shell=%s)",
command, cwd, universal_newlines, shell)
try:
proc = Popen(command,
env=env,
cwd=cwd,
bufsize=-1,
stdin=istream,
stderr=PIPE,
stdout=stdout_sink,
shell=shell is not None and shell or self.USE_SHELL,
close_fds=is_posix, # unsupported on windows
universal_newlines=universal_newlines,
creationflags=PROC_CREATIONFLAGS,
**subprocess_kwargs
)
except cmd_not_found_exception as err:
raise GitCommandNotFound(command, err)
if as_process:
return self.AutoInterrupt(proc, command)
def _kill_process(pid):
""" Callback method to kill a process. """
p = Popen(['ps', '--ppid', str(pid)], stdout=PIPE,
creationflags=PROC_CREATIONFLAGS)
child_pids = []
for line in p.stdout:
if len(line.split()) > 0:
local_pid = (line.split())[0]
if local_pid.isdigit():
child_pids.append(int(local_pid))
try:
# Windows does not have SIGKILL, so use SIGTERM instead
sig = getattr(signal, 'SIGKILL', signal.SIGTERM)
os.kill(pid, sig)
for child_pid in child_pids:
try:
os.kill(child_pid, sig)
except OSError:
pass
kill_check.set() # tell the main routine that the process was killed
except OSError:
# It is possible that the process gets completed in the duration after timeout
# happens and before we try to kill the process.
pass
return
# end
if kill_after_timeout:
kill_check = threading.Event()
watchdog = threading.Timer(kill_after_timeout, _kill_process, args=(proc.pid,))
# Wait for the process to return
status = 0
stdout_value = b''
stderr_value = b''
try:
if output_stream is None:
if kill_after_timeout:
watchdog.start()
stdout_value, stderr_value = proc.communicate()
if kill_after_timeout:
watchdog.cancel()
if kill_check.isSet():
stderr_value = 'Timeout: the command "%s" did not complete in %d ' \
'secs.' % (" ".join(command), kill_after_timeout)
# strip trailing "\n"
if stdout_value.endswith(b"\n"):
stdout_value = stdout_value[:-1]
if stderr_value.endswith(b"\n"):
stderr_value = stderr_value[:-1]
status = proc.returncode
else:
stream_copy(proc.stdout, output_stream, self.max_chunk_size)
stdout_value = output_stream
stderr_value = proc.stderr.read()
# strip trailing "\n"
if stderr_value.endswith(b"\n"):
stderr_value = stderr_value[:-1]
status = proc.wait()
# END stdout handling
finally:
proc.stdout.close()
proc.stderr.close()
if self.GIT_PYTHON_TRACE == 'full':
cmdstr = " ".join(command)
def as_text(stdout_value):
return not output_stream and safe_decode(stdout_value) or '<OUTPUT_STREAM>'
# end
if stderr_value:
log.info("%s -> %d; stdout: '%s'; stderr: '%s'",
cmdstr, status, as_text(stdout_value), safe_decode(stderr_value))
elif stdout_value:
log.info("%s -> %d; stdout: '%s'", cmdstr, status, as_text(stdout_value))
else:
log.info("%s -> %d", cmdstr, status)
# END handle debug printing
if with_exceptions and status != 0:
raise GitCommandError(command, status, stderr_value, stdout_value)
if isinstance(stdout_value, bytes) and stdout_as_string: # could also be output_stream
stdout_value = safe_decode(stdout_value)
# Allow access to the command's status code
if with_extended_output:
return (status, stdout_value, safe_decode(stderr_value))
else:
return stdout_value
def blame_incremental(self, rev, file, **kwargs):
"""Iterator for blame information for the given file at the given revision.
Unlike .blame(), this does not return the actual file's contents, only
a stream of BlameEntry tuples.
:parm rev: revision specifier, see git-rev-parse for viable options.
:return: lazy iterator of BlameEntry tuples, where the commit
indicates the commit to blame for the line, and range
indicates a span of line numbers in the resulting file.
If you combine all line number ranges outputted by this command, you
should get a continuous range spanning all line numbers in the file.
"""
data = self.git.blame(rev, '--', file, p=True, incremental=True, stdout_as_string=False, **kwargs)
commits = dict()
stream = (line for line in data.split(b'\n') if line)
while True:
line = next(stream) # when exhausted, casues a StopIteration, terminating this function
hexsha, orig_lineno, lineno, num_lines = line.split()
lineno = int(lineno)
num_lines = int(num_lines)
orig_lineno = int(orig_lineno)
if hexsha not in commits:
# Now read the next few lines and build up a dict of properties
# for this commit
props = dict()
while True:
line = next(stream)
if line == b'boundary':
# "boundary" indicates a root commit and occurs
# instead of the "previous" tag
continue
tag, value = line.split(b' ', 1)
props[tag] = value
if tag == b'filename':
# "filename" formally terminates the entry for --incremental
orig_filename = value
break
c = Commit(self, hex_to_bin(hexsha),
author=Actor(safe_decode(props[b'author']),
safe_decode(props[b'author-mail'].lstrip(b'<').rstrip(b'>'))),
authored_date=int(props[b'author-time']),
committer=Actor(safe_decode(props[b'committer']),
safe_decode(props[b'committer-mail'].lstrip(b'<').rstrip(b'>'))),
committed_date=int(props[b'committer-time']))
commits[hexsha] = c
else:
# Discard all lines until we find "filename" which is
# guaranteed to be the last line
while True:
line = next(stream) # will fail if we reach the EOF unexpectedly
tag, value = line.split(b' ', 1)
if tag == b'filename':
orig_filename = value
break
yield BlameEntry(commits[hexsha],
range(lineno, lineno + num_lines),
safe_decode(orig_filename),
range(orig_lineno, orig_lineno + num_lines))
def execute(self,
command,
istream=None,
with_extended_output=False,
with_exceptions=True,
as_process=False,
output_stream=None,
stdout_as_string=True,
kill_after_timeout=None,
with_stdout=True,
universal_newlines=False,
shell=None,
env=None,
**subprocess_kwargs):
"""Handles executing the command on the shell and consumes and returns
the returned information (stdout)
:param command:
The command argument list to execute.
It should be a string, or a sequence of program arguments. The
program to execute is the first item in the args sequence or string.
:param istream:
Standard input filehandle passed to subprocess.Popen.
:param with_extended_output:
Whether to return a (status, stdout, stderr) tuple.
:param with_exceptions:
Whether to raise an exception when git returns a non-zero status.
:param as_process:
Whether to return the created process instance directly from which
streams can be read on demand. This will render with_extended_output and
with_exceptions ineffective - the caller will have
to deal with the details himself.
It is important to note that the process will be placed into an AutoInterrupt
wrapper that will interrupt the process once it goes out of scope. If you
use the command in iterators, you should pass the whole process instance
instead of a single stream.
:param output_stream:
If set to a file-like object, data produced by the git command will be
output to the given stream directly.
This feature only has any effect if as_process is False. Processes will
always be created with a pipe due to issues with subprocess.
This merely is a workaround as data will be copied from the
output pipe to the given output stream directly.
Judging from the implementation, you shouldn't use this flag !
:param stdout_as_string:
if False, the commands standard output will be bytes. Otherwise, it will be
decoded into a string using the default encoding (usually utf-8).
The latter can fail, if the output contains binary data.
:param env:
A dictionary of environment variables to be passed to `subprocess.Popen`.
:param subprocess_kwargs:
Keyword arguments to be passed to subprocess.Popen. Please note that
some of the valid kwargs are already set by this method, the ones you
specify may not be the same ones.
:param with_stdout: If True, default True, we open stdout on the created process
:param universal_newlines:
if True, pipes will be opened as text, and lines are split at
all known line endings.
:param shell:
Whether to invoke commands through a shell (see `Popen(..., shell=True)`).
It overrides :attr:`USE_SHELL` if it is not `None`.
:param kill_after_timeout:
To specify a timeout in seconds for the git command, after which the process
should be killed. This will have no effect if as_process is set to True. It is
set to None by default and will let the process run until the timeout is
explicitly specified. This feature is not supported on Windows. It's also worth
noting that kill_after_timeout uses SIGKILL, which can have negative side
effects on a repository. For example, stale locks in case of git gc could
render the repository incapable of accepting changes until the lock is manually
removed.
:return:
* str(output) if extended_output = False (Default)
* tuple(int(status), str(stdout), str(stderr)) if extended_output = True
if output_stream is True, the stdout value will be your output stream:
* output_stream if extended_output = False
* tuple(int(status), output_stream, str(stderr)) if extended_output = True
Note git is executed with LC_MESSAGES="C" to ensure consistent
output regardless of system language.
:raise GitCommandError:
:note:
If you add additional keyword arguments to the signature of this method,
you must update the execute_kwargs tuple housed in this module."""
if self.GIT_PYTHON_TRACE and (self.GIT_PYTHON_TRACE != 'full'
or as_process):
log.info(' '.join(command))
# Allow the user to have the command executed in their working dir.
cwd = self._working_dir or os.getcwd()
# Start the process
inline_env = env
env = os.environ.copy()
# Attempt to force all output to plain ascii english, which is what some parsing code
# may expect.
# According to stackoverflow (http://goo.gl/l74GC8), we are setting LANGUAGE as well
# just to be sure.
env["LANGUAGE"] = "C"
env["LC_ALL"] = "C"
env.update(self._environment)
if inline_env is not None:
env.update(inline_env)
if is_win:
cmd_not_found_exception = OSError
if kill_after_timeout:
raise GitCommandError(
command,
'"kill_after_timeout" feature is not supported on Windows.'
)
else:
if sys.version_info[0] > 2:
cmd_not_found_exception = FileNotFoundError # NOQA # exists, flake8 unknown @UndefinedVariable
else:
cmd_not_found_exception = OSError
# end handle
stdout_sink = (PIPE if with_stdout else getattr(
subprocess, 'DEVNULL', None) or open(os.devnull, 'wb'))
log.debug("Popen(%s, cwd=%s, universal_newlines=%s, shell=%s)",
command, cwd, universal_newlines, shell)
try:
proc = Popen(
command,
env=env,
cwd=cwd,
bufsize=-1,
stdin=istream,
stderr=PIPE,
stdout=stdout_sink,
shell=shell is not None and shell or self.USE_SHELL,
close_fds=is_posix, # unsupported on windows
universal_newlines=universal_newlines,
creationflags=PROC_CREATIONFLAGS,
**subprocess_kwargs)
except cmd_not_found_exception as err:
raise GitCommandNotFound(command, err)
if as_process:
return self.AutoInterrupt(proc, command)
def _kill_process(pid):
""" Callback method to kill a process. """
p = Popen(['ps', '--ppid', str(pid)],
stdout=PIPE,
creationflags=PROC_CREATIONFLAGS)
child_pids = []
for line in p.stdout:
if len(line.split()) > 0:
local_pid = (line.split())[0]
if local_pid.isdigit():
child_pids.append(int(local_pid))
try:
# Windows does not have SIGKILL, so use SIGTERM instead
sig = getattr(signal, 'SIGKILL', signal.SIGTERM)
os.kill(pid, sig)
for child_pid in child_pids:
try:
os.kill(child_pid, sig)
except OSError:
pass
kill_check.set(
) # tell the main routine that the process was killed
except OSError:
# It is possible that the process gets completed in the duration after timeout
# happens and before we try to kill the process.
pass
return
# end
if kill_after_timeout:
kill_check = threading.Event()
watchdog = threading.Timer(kill_after_timeout,
_kill_process,
args=(proc.pid, ))
# Wait for the process to return
status = 0
stdout_value = b''
stderr_value = b''
try:
if output_stream is None:
if kill_after_timeout:
watchdog.start()
stdout_value, stderr_value = proc.communicate()
if kill_after_timeout:
watchdog.cancel()
if kill_check.isSet():
stderr_value = (
'Timeout: the command "%s" did not complete in %d '
'secs.' % (" ".join(command),
kill_after_timeout)).encode(defenc)
# strip trailing "\n"
if stdout_value.endswith(b"\n"):
stdout_value = stdout_value[:-1]
if stderr_value.endswith(b"\n"):
stderr_value = stderr_value[:-1]
status = proc.returncode
else:
stream_copy(proc.stdout, output_stream, self.max_chunk_size)
stdout_value = output_stream
stderr_value = proc.stderr.read()
# strip trailing "\n"
if stderr_value.endswith(b"\n"):
stderr_value = stderr_value[:-1]
status = proc.wait()
# END stdout handling
finally:
proc.stdout.close()
proc.stderr.close()
if self.GIT_PYTHON_TRACE == 'full':
cmdstr = " ".join(command)
def as_text(stdout_value):
return not output_stream and safe_decode(
stdout_value) or '<OUTPUT_STREAM>'
# end
if stderr_value:
log.info("%s -> %d; stdout: '%s'; stderr: '%s'", cmdstr,
status, as_text(stdout_value),
safe_decode(stderr_value))
elif stdout_value:
log.info("%s -> %d; stdout: '%s'", cmdstr, status,
as_text(stdout_value))
else:
log.info("%s -> %d", cmdstr, status)
# END handle debug printing
if with_exceptions and status != 0:
raise GitCommandError(command, status, stderr_value, stdout_value)
if isinstance(
stdout_value,
bytes) and stdout_as_string: # could also be output_stream
stdout_value = safe_decode(stdout_value)
# Allow access to the command's status code
if with_extended_output:
return (status, stdout_value, safe_decode(stderr_value))
else:
return stdout_value
def as_text(stdout_value):
return not output_stream and safe_decode(
stdout_value) or '<OUTPUT_STREAM>'
def blame_incremental(self, rev, file, **kwargs):
"""Iterator for blame information for the given file at the given revision.
Unlike .blame(), this does not return the actual file's contents, only
a stream of (commit, range) tuples.
:parm rev: revision specifier, see git-rev-parse for viable options.
:return: lazy iterator of (git.Commit, range) tuples, where the commit
indicates the commit to blame for the line, and range
indicates a span of line numbers in the resulting file.
If you combine all line number ranges outputted by this command, you
should get a continuous range spanning all line numbers in the file.
"""
data = self.git.blame(rev,
'--',
file,
p=True,
incremental=True,
stdout_as_string=False,
**kwargs)
commits = dict()
stream = iter(data.splitlines())
while True:
line = next(
stream
) # when exhausted, casues a StopIteration, terminating this function
hexsha, _, lineno, num_lines = line.split()
lineno = int(lineno)
num_lines = int(num_lines)
if hexsha not in commits:
# Now read the next few lines and build up a dict of properties
# for this commit
props = dict()
while True:
line = next(stream)
if line == b'boundary':
# "boundary" indicates a root commit and occurs
# instead of the "previous" tag
continue
tag, value = line.split(b' ', 1)
props[tag] = value
if tag == b'filename':
# "filename" formally terminates the entry for --incremental
break
c = Commit(
self,
hex_to_bin(hexsha),
author=Actor(
safe_decode(props[b'author']),
safe_decode(
props[b'author-mail'].lstrip(b'<').rstrip(b'>'))),
authored_date=int(props[b'author-time']),
committer=Actor(
safe_decode(props[b'committer']),
safe_decode(props[b'committer-mail'].lstrip(
b'<').rstrip(b'>'))),
committed_date=int(props[b'committer-time']),
message=safe_decode(props[b'summary']))
commits[hexsha] = c
else:
# Discard the next line (it's a filename end tag)
line = next(stream)
assert line.startswith(
b'filename'), 'Unexpected git blame output'
yield commits[hexsha], range(lineno, lineno + num_lines)