def as_entry(infos):
"""
Return an Entry built from 7zip path data.
"""
e = extractcode.Entry()
e.path = infos.get('Path')
e.size = infos.get('Size', 0)
e.packed_size = infos.get('Packed Size', 0)
e.date = infos.get('Modified', 0)
e.is_dir = infos.get('Folder', False) == '+'
e.is_file = not e.is_dir
e.is_broken_link = False
e.mode = infos.get('Mode', '')
e.user = infos.get('User')
e.group = infos.get('Group')
e.is_special = False
e.is_hardlink = False
sl = infos.get('Symbolic Link')
if sl:
e.is_symlink = True
e.link_target = sl
hl = infos.get('Hard Link')
if hl:
e.is_hardlink = True
e.link_target = hl
if sl and hl:
raise ExtractWarningIncorrectEntry('A Symlink cannot be a hardlink too')
e.linkcount = infos.get('Links', 0)
e.host = infos.get('Host OS')
e.comment = infos.get('Comment')
e.encrypted = infos.get('Encrypted')
return e
def from_dict(cls, infos, errors=None):
"""
Return an Entry built from a 7zip path listing data in the `infos` mapping.
"""
is_symlink = False
is_hardlink = False
link_target = None
sl = infos.get('Symbolic Link')
if sl:
is_symlink = True
link_target = sl
hl = infos.get('Hard Link')
if hl:
is_hardlink = True
link_target = hl
if sl and hl:
from pprint import pformat
raise ExtractWarningIncorrectEntry(
'A symlink cannot be also a hardlink: {}'.format(
pformat(infos)))
# depending on the type of arhcive the file vs dir flags are in
# diiferent attributes :|
is_dir = (
# in some listings we have this: Mode = drwxrwxr-x
infos.get('Mode', '').lower().startswith('d') or
# in cpio and a few more we have a Folder attrib
infos.get('Folder', '').startswith('+') or
# in 7z listing we have this: Attributes = D_ drwxrwxr-x
infos.get('Attributes', '').lower().startswith('d_')) or False
is_file = not is_dir
e = cls(
path=infos.get('Path'),
size=infos.get('Size', 0),
date=infos.get('Modified', None),
is_dir=is_dir,
is_file=is_file,
is_symlink=is_symlink,
is_hardlink=is_hardlink,
link_target=link_target,
errors=errors or [],
)
return e
def parse_7z_listing(location, utf=False):
"""
Parse a long format 7zip listing and return an iterable of entry.
The 7zip -slt format is:
- copyright and version details
- '--' line
- archive header info, varying based on the archive types and subtype
- lines of key=value pairs
- Errors: followed by one or more message lines
- Warnings: followed by one or more message lines
- Open Warning: : followed by one or more message lines
- sometimes a '---' line
- blank line
- '----------' line
- for each archive member:
- lines of either
- key = value pairs
- Errors: followed by one or more message lines
- Warnings: followed by one or more message lines
- Open Warning: : followed by one or more message lines
- blank line
- two blank lines
- footer sometimes with lines with summary stats
such as Warnings: 1 Errors: 1
- a line with two or more dashes or an empty line
"""
if utf:
text = codecs.open(location, encoding='UTF-8').read()
text = text.replace(u'\r\n', u'\n')
else:
text = open(location, 'rb').read()
header_tail = re.split('\n----------\n', text, flags=re.MULTILINE)
if len(header_tail) != 2:
# we more than one a header, confusion entails.
raise ExtractWarningIncorrectEntry('Incorrect 7zip listing with multiple headers')
if len(header_tail) == 1:
# we have only a header, likely an error condition or an empty archive
return []
_header, body = header_tail
body_and_footer = re.split('\n\n\n', body, flags=re.MULTILINE)
no_footer = len(body_and_footer) == 1
multiple_footers = len(body_and_footer) > 2
_footer = ''
if no_footer:
body = body_and_footer[0]
elif multiple_footers:
raise ExtractWarningIncorrectEntry('Incorrect 7zip listing with multiple footers')
else:
body, _footer == body_and_footer
# FIXME: do something with header and footer?
entries = []
paths = re.split('\n\n', body, flags=re.MULTILINE)
for path in paths:
is_err = False
errors = []
infos = {}
lines = path.splitlines(False)
for line in lines:
line = line.strip()
if not line:
continue
if line.startswith(('Open Warning:', 'Errors:', 'Warnings:')):
is_err = True
messages = line.split(':', 1)
errors.append(messages)
continue
if '=' not in line and is_err:
# not a key = value line, an error message
errors.append(line)
continue
parts = line.split('=', 1)
if len(parts) != 2:
raise ExtractWarningIncorrectEntry('Incorrect 7zip listing line with no key=value')
is_err = False
key, value = parts
assert key not in infos, 'Duplicate keys in 7zip listing'
infos[key.strip()] = value.strip() or ''
if infos:
entries.append(as_entry(infos))
return entries
def parse_7z_listing(location, utf=False):
"""
Parse a long format 7zip listing and return an iterable of entry.
If `utf` is True or if on Python 3, the console output will treated as
utf-8-encoded text. Otherwise it is treated as bytes.
The 7zip -slt format is:
- copyright and version details
- '--' line
- archive header info, varying based on the archive types and subtype
- lines of key=value pairs
- Errors: followed by one or more message lines
- Warnings: followed by one or more message lines
- Open Warning: : followed by one or more message lines
- sometimes a '---' line
- blank line
- '----------' line
- for each archive member:
- lines of either
- key = value pairs
- Errors: followed by one or more message lines
- Warnings: followed by one or more message lines
- Open Warning: : followed by one or more message lines
- blank line
- two blank lines
- footer sometimes with lines with summary stats
such as Warnings: 1 Errors: 1
- a line with two or more dashes or an empty line
"""
if utf or py3:
# read to unicode
with io.open(location, 'r', encoding='utf-8') as listing:
text = listing.read()
text = text.replace(u'\r\n', u'\n')
header_sep = u'\n----------\n'
empty = u''
body_sep = u'\n\n\n'
path_sep = u'\n\n'
msg_sep = u':'
equal_sep = u'='
errror_line_starters = 'Open Warning:', 'Errors:', 'Warnings:'
else:
# read to bytes
with io.open(location, 'rb') as listing:
text = listing.read()
header_sep = b'\n----------\n'
empty = b''
body_sep = b'\n\n\n'
path_sep = b'\n\n'
msg_sep = b':'
equal_sep = b'='
errror_line_starters = b'Open Warning:', b'Errors:', b'Warnings:'
if TRACE:
logger.debug('parse_7z_listing: initial text: type: ' +
repr(type(text)))
print('--------------------------------------')
print(text)
print('--------------------------------------')
header_tail = re.split(header_sep, text, flags=re.MULTILINE)
if len(header_tail) != 2:
# we more than one a header, confusion entails.
raise ExtractWarningIncorrectEntry(
'Incorrect 7zip listing with multiple headers')
if len(header_tail) == 1:
# we have only a header, likely an error condition or an empty archive
return []
# FIXME: do something with header and footer?
_header, body = header_tail
body_and_footer = re.split(body_sep, body, flags=re.MULTILINE)
no_footer = len(body_and_footer) == 1
multiple_footers = len(body_and_footer) > 2
_footer = empty
if no_footer:
body = body_and_footer[0]
elif multiple_footers:
raise ExtractWarningIncorrectEntry(
'Incorrect 7zip listing with multiple footers')
else:
body, _footer == body_and_footer
entries = []
if TRACE:
logger.debug('parse_7z_listing: body:')
print(body)
paths = re.split(path_sep, body, flags=re.MULTILINE)
if TRACE:
logger.debug('parse_7z_listing: paths:')
pprint(paths)
for path in paths:
is_err = False
errors = []
infos = {}
lines = path.splitlines(False)
for line in lines:
line = line.strip()
if not line:
continue
if line.startswith(errror_line_starters):
is_err = True
messages = line.split(msg_sep, 1)
errors.append(messages)
continue
if equal_sep not in line and is_err:
# not a key = value line, an error message
errors.append(line)
continue
parts = line.split(equal_sep, 1)
if len(parts) != 2:
raise ExtractWarningIncorrectEntry(
'Incorrect 7zip listing line with no key=value')
is_err = False
key, value = parts
assert key not in infos, 'Duplicate keys in 7zip listing'
infos[key.strip()] = value.strip() or empty
if infos:
entries.append(as_entry(infos))
return entries