* Moving pyqmc.utils.file_utils to wpylib project.

wirawan 14 years ago
parent 631a152b8e
commit 090d9f7b00
  1. 9
  2. 168

@ -0,0 +1,9 @@
# $Id: __init__.py,v 1.1 2010-09-27 19:47:19 wirawan Exp $
# wpylib.file module
# Created: 20100927
# Wirawan Purwanto
"""File-related stuff."""

@ -0,0 +1,168 @@
# $Id: file_utils.py,v 1.1 2010-09-27 19:47:19 wirawan Exp $
# pyqmc.utils.file_utils module
# File-manipulation utilities
# Wirawan Purwanto
# Created: 20090601
# Routines put here are commonly used in my own scripts.
# They are not necessarily suitable for general-purpose uses; evaluate
# your needs and see if they can them as well.
import bz2
import glob
import gzip
import os
import subprocess
has_subprocess = True
has_subprocess = False
class super_file(object):
'''"Super-file" hack wrapper for a file-like object.
Intended to allow extra capabilities to file-like iterators such as:
* ability to push back text lines for the subsequent next() calls.
This is to provide some level of rewinding in parsing text files.
* what else?
def __init__(self, obj):
'''Creates a super_file wrapper around the "obj" object.'''
self.obj = obj
self.pushback = []
def __iter__(self):
return self
def close(self):
return self.obj.close()
def flush(self):
return self.obj.flush()
def next(self):
if len(self.pushback) > 0:
return self.pushback.pop()
return self.obj.next()
def push(self, s):
def open_input_file(fname, superize=0):
if fname.endswith(".bz2"):
fobj = bz2.BZ2File(fname, "r")
elif fname.endswith(".gz") or fname.endswith(".Z"):
fobj = gzip.GzipFile(fname, "r")
elif fname.endswith(".lzma"):
# until lzma has a "standard" python module, we use "lzma" executable:
if has_subprocess:
px = subprocess.Popen(("lzma", "-dc", fname), stdout=subprocess.PIPE)
fobj = px.stdout
px = os.popen('lzma -dc "' + fname + '"', "r")
fobj = open(fname, "r")
if superize:
return super_file(fobj)
return fobj
# Miscellaneous functions
def glob_files(filespec):
'''Processes a glob string, or does nothing (pass-on only) if an iterable object
(e.g. list or tuple) is already given.
When globbing is done, the result is sorted for predictability.'''
if getattr(filespec, "__iter__", False):
return filespec # no re-sorting
elif isinstance(filespec, str):
return sorted(glob.glob(filespec))
raise ValueError, "Don't know how to glob for an object of " + type(filespec)
def path_search(*specs, **opts):
'''Generalized path search.
Multiple paths can be specified for different parts of the sought filename,
and the first file found is returned.
Additional options:
* pathsep="/" -- path separator
* filetest=os.path.isfile -- filetest operator to be used
* raise_error=False -- do we want to raise an exception if the file
is not found?
# FIXME: this can be extremely expensive!
xspecs = []
xlen = []
xstride = []
xtot = 1
pathsep = opts.get("pathsep", "/")
filetest = opts.get("filetest", os.path.isfile)
for spec in specs:
if not getattr(spec, "__iter__", False):
xspecs.append(tuple([ x for x in spec ]))
xtot *= xlen[-1]
for idx in xrange(xtot):
idx0 = idx
# Construct the filename based on the index: we reconstruct
# the indices for all the parts given in the argument, then
# concatenate them to get the full pathname
s = ""
for d in xrange(len(xspecs)-1,-1,-1):
a = idx0 / xstride[d]
if s == "":
s = xspecs[d][a]
s = xspecs[d][a] + pathsep + s
idx0 = idx0 % xstride[d]
#print a,
#print s
if filetest(s):
return s
if opts.get("raise_error", False):
raise ValueError, "Cannot find file with specified combination"
return None
def untar(archive, subdir=None, verbose=None, files=[]):
'''Extracts a TAR archive. The destination directory can be given; otherwise
the files are extracted to the current directory.
Assuming GNU tar which accepts -z and -j switches.
LZMA compression is supported via lzma program.
opts = [ 'tar' ]
# Python doc says: "the arguments to the child process must start with the
# name of the command being run"
if subdir:
opts += [ "-C", subdir ]
if archive.endswith(".tar.bz2") or archive.endswith(".tbz2") or archive.endswith(".tbz"):
elif archive.endswith(".tar.Z") or archive.endswith(".tar.gz") or archive.endswith(".tgz"):
elif archive.endswith(".tar.lzma") or archive.endswith(".tza"):
if verbose:
for i in xrange(verbose): opts.append("-v")
opts += [ "-xf", archive ]
opts += files
return os.spawnvp(os.P_WAIT, "tar", opts)