From bb0ec1fc07628252f17d66a58a4a4b3654e571c3 Mon Sep 17 00:00:00 2001 From: Larry Hastings Date: Mon, 5 Sep 2016 15:11:23 -0700 Subject: Issue #27355: Removed support for Windows CE. It was never finished, and Windows CE is no longer a relevant platform for Python. --- Lib/tarfile.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'Lib/tarfile.py') diff --git a/Lib/tarfile.py b/Lib/tarfile.py index df4745d6ba..960c673067 100755 --- a/Lib/tarfile.py +++ b/Lib/tarfile.py @@ -144,7 +144,7 @@ PAX_NUMBER_FIELDS = { #--------------------------------------------------------- # initialization #--------------------------------------------------------- -if os.name in ("nt", "ce"): +if os.name == "nt": ENCODING = "utf-8" else: ENCODING = sys.getfilesystemencoding() -- cgit v1.2.1 From 4f05fcfb622ac6ed16a0b5a84cf1e657d98ed54e Mon Sep 17 00:00:00 2001 From: ?ukasz Langa Date: Fri, 9 Sep 2016 19:48:14 -0700 Subject: Issue #27199: TarFile expose copyfileobj bufsize to improve throughput Patch by Jason Fried. --- Lib/tarfile.py | 33 ++++++++++++++++++--------------- 1 file changed, 18 insertions(+), 15 deletions(-) (limited to 'Lib/tarfile.py') diff --git a/Lib/tarfile.py b/Lib/tarfile.py index 960c673067..a62ab82545 100755 --- a/Lib/tarfile.py +++ b/Lib/tarfile.py @@ -228,21 +228,21 @@ def calc_chksums(buf): signed_chksum = 256 + sum(struct.unpack_from("148b8x356b", buf)) return unsigned_chksum, signed_chksum -def copyfileobj(src, dst, length=None, exception=OSError): +def copyfileobj(src, dst, length=None, exception=OSError, bufsize=None): """Copy length bytes from fileobj src to fileobj dst. If length is None, copy the entire content. """ + bufsize = bufsize or 16 * 1024 if length == 0: return if length is None: - shutil.copyfileobj(src, dst) + shutil.copyfileobj(src, dst, bufsize) return - BUFSIZE = 16 * 1024 - blocks, remainder = divmod(length, BUFSIZE) + blocks, remainder = divmod(length, bufsize) for b in range(blocks): - buf = src.read(BUFSIZE) - if len(buf) < BUFSIZE: + buf = src.read(bufsize) + if len(buf) < bufsize: raise exception("unexpected end of data") dst.write(buf) @@ -1403,7 +1403,8 @@ class TarFile(object): def __init__(self, name=None, mode="r", fileobj=None, format=None, tarinfo=None, dereference=None, ignore_zeros=None, encoding=None, - errors="surrogateescape", pax_headers=None, debug=None, errorlevel=None): + errors="surrogateescape", pax_headers=None, debug=None, + errorlevel=None, copybufsize=None): """Open an (uncompressed) tar archive `name'. `mode' is either 'r' to read from an existing archive, 'a' to append data to an existing file or 'w' to create a new file overwriting an existing one. `mode' @@ -1459,6 +1460,7 @@ class TarFile(object): self.errorlevel = errorlevel # Init datastructures. + self.copybufsize = copybufsize self.closed = False self.members = [] # list of members as TarInfo objects self._loaded = False # flag if all members have been read @@ -1558,7 +1560,7 @@ class TarFile(object): saved_pos = fileobj.tell() try: return func(name, "r", fileobj, **kwargs) - except (ReadError, CompressionError) as e: + except (ReadError, CompressionError): if fileobj is not None: fileobj.seek(saved_pos) continue @@ -1963,10 +1965,10 @@ class TarFile(object): buf = tarinfo.tobuf(self.format, self.encoding, self.errors) self.fileobj.write(buf) self.offset += len(buf) - + bufsize=self.copybufsize # If there's data to follow, append it. if fileobj is not None: - copyfileobj(fileobj, self.fileobj, tarinfo.size) + copyfileobj(fileobj, self.fileobj, tarinfo.size, bufsize=bufsize) blocks, remainder = divmod(tarinfo.size, BLOCKSIZE) if remainder > 0: self.fileobj.write(NUL * (BLOCKSIZE - remainder)) @@ -2148,15 +2150,16 @@ class TarFile(object): """ source = self.fileobj source.seek(tarinfo.offset_data) + bufsize = self.copybufsize with bltn_open(targetpath, "wb") as target: if tarinfo.sparse is not None: for offset, size in tarinfo.sparse: target.seek(offset) - copyfileobj(source, target, size, ReadError) + copyfileobj(source, target, size, ReadError, bufsize) target.seek(tarinfo.size) target.truncate() else: - copyfileobj(source, target, tarinfo.size, ReadError) + copyfileobj(source, target, tarinfo.size, ReadError, bufsize) def makeunknown(self, tarinfo, targetpath): """Make a file from a TarInfo object with an unknown type @@ -2235,7 +2238,7 @@ class TarFile(object): os.lchown(targetpath, u, g) else: os.chown(targetpath, u, g) - except OSError as e: + except OSError: raise ExtractError("could not change owner") def chmod(self, tarinfo, targetpath): @@ -2244,7 +2247,7 @@ class TarFile(object): if hasattr(os, 'chmod'): try: os.chmod(targetpath, tarinfo.mode) - except OSError as e: + except OSError: raise ExtractError("could not change mode") def utime(self, tarinfo, targetpath): @@ -2254,7 +2257,7 @@ class TarFile(object): return try: os.utime(targetpath, (tarinfo.mtime, tarinfo.mtime)) - except OSError as e: + except OSError: raise ExtractError("could not change modification time") #-------------------------------------------------------------------------- -- cgit v1.2.1 From 12201a6874ee891b263f9c405d93a4359f038e67 Mon Sep 17 00:00:00 2001 From: Xavier de Gaye Date: Fri, 9 Dec 2016 09:33:09 +0100 Subject: Issue #26937: The chown() method of the tarfile.TarFile class does not fail now when the grp module cannot be imported, as for example on Android platforms. --- Lib/tarfile.py | 31 +++++++++++++++++++------------ 1 file changed, 19 insertions(+), 12 deletions(-) (limited to 'Lib/tarfile.py') diff --git a/Lib/tarfile.py b/Lib/tarfile.py index b78b1b1f4b..5d4c86ce36 100755 --- a/Lib/tarfile.py +++ b/Lib/tarfile.py @@ -50,9 +50,13 @@ import copy import re try: - import grp, pwd + import pwd except ImportError: - grp = pwd = None + pwd = None +try: + import grp +except ImportError: + grp = None # os.symlink on Windows prior to 6.0 raises NotImplementedError symlink_exception = (AttributeError, NotImplementedError) @@ -2219,22 +2223,25 @@ class TarFile(object): def chown(self, tarinfo, targetpath, numeric_owner): """Set owner of targetpath according to tarinfo. If numeric_owner - is True, use .gid/.uid instead of .gname/.uname. + is True, use .gid/.uid instead of .gname/.uname. If numeric_owner + is False, fall back to .gid/.uid when the search based on name + fails. """ - if pwd and hasattr(os, "geteuid") and os.geteuid() == 0: + if hasattr(os, "geteuid") and os.geteuid() == 0: # We have to be root to do so. - if numeric_owner: - g = tarinfo.gid - u = tarinfo.uid - else: + g = tarinfo.gid + u = tarinfo.uid + if not numeric_owner: try: - g = grp.getgrnam(tarinfo.gname)[2] + if grp: + g = grp.getgrnam(tarinfo.gname)[2] except KeyError: - g = tarinfo.gid + pass try: - u = pwd.getpwnam(tarinfo.uname)[2] + if pwd: + u = pwd.getpwnam(tarinfo.uname)[2] except KeyError: - u = tarinfo.uid + pass try: if tarinfo.issym() and hasattr(os, "lchown"): os.lchown(targetpath, u, g) -- cgit v1.2.1