Ñò
\ÐKc @ s‰ d Z d Z d d k Z d d k l Z d e f d „ ƒ YZ e ƒ Z d e i d „ Z d e f d
„ ƒ YZ
d e
f d „ ƒ YZ d S(
sÚ A file interface for handling local and remote data files.
The goal of datasource is to abstract some of the file system operations when
dealing with data files so the researcher doesn't have to know all the
low-level details. Through datasource, a researcher can obtain and use a
file with one function call, regardless of location of the file.
DataSource is meant to augment standard python libraries, not replace them.
It should work seemlessly with standard file IO operations and the os module.
DataSource files can originate locally or remotely:
- local files : '/home/guido/src/local/data.txt'
- URLs (http, ftp, ...) : 'http://www.scipy.org/not/real/data.txt'
DataSource files can also be compressed or uncompressed. Currently only gzip
and bz2 are supported.
Example::
>>> # Create a DataSource, use os.curdir (default) for local storage.
>>> ds = datasource.DataSource()
>>>
>>> # Open a remote file.
>>> # DataSource downloads the file, stores it locally in:
>>> # './www.google.com/index.html'
>>> # opens the file and returns a file object.
>>> fp = ds.open('http://www.google.com/index.html')
>>>
>>> # Use the file as you normally would
>>> fp.read()
>>> fp.close()
s restructuredtext eniÿÿÿÿN( t rmtreet _FileOpenersc B s2 e Z d Z d „ Z d „ Z d „ Z d „ Z RS( sÞ
Container for different methods to open (un-)compressed files.
`_FileOpeners` contains a dictionary that holds one method for each
supported file format. Attribute lookup is implemented in such a way that
an instance of `_FileOpeners` itself can be indexed with the keys of that
dictionary. Currently uncompressed files as well as files
compressed with ``gzip`` or ``bz2`` compression are supported.
Notes
-----
`_file_openers`, an instance of `_FileOpeners`, is made available for
use in the `_datasource` module.
Examples
--------
>>> np.lib._datasource._file_openers.keys()
[None, '.bz2', '.gz']
>>> np.lib._datasource._file_openers['.gz'] is gzip.open
True
c C s t | _ h t d 6| _ d S( N( t Falset _loadedt opent Nonet
_file_openers( t self( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt __init__D s c C sˆ | i o d Sy d d k } | i | i d >> repos = DataSource()
>>> repos.exists('www.google.com/index.html')
False
>>> repos.exists('http://www.google.com/index.html')
True
Temporary directories are deleted when the DataSource is deleted.
Examples
--------
::
>>> ds = DataSource('/home/guido')
>>> urlname = 'http://www.google.com/index.html'
>>> gfile = ds.open('http://www.google.com/index.html') # remote file
>>> ds.abspath(urlname)
'/home/guido/www.google.com/site/index.html'
>>> ds = DataSource(None) # use with temporary file
>>> ds.open('/home/guido/foobar.txt')
>>> ds.abspath('/home/guido/foobar.txt')
'/tmp/tmpy4pgsP/home/guido/foobar.txt'
c C sQ | o" t i i | ƒ | _ t | _ n% d d k } | i ƒ | _ t | _ d S( s2 Create a DataSource with a local path at destpath.iÿÿÿÿN( t osR t abspatht _destpathR t
_istmpdestt tempfilet mkdtempR
( R R R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR Ã s
c C s | i o t | i ƒ n d S( N( R R R ( R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt __del__Í s
c C s( t i i | ƒ \ } } | t i ƒ j S( sM Test if the filename is a zip file by looking at the file extension.
( R R t splitextR R ( R t filenamet fnamet ext( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _iszipÒ s c C s- d } x | D] } | | j o t Sq
Wt S( s4 Test if the given mode will open a file for writing.t wt +( R' R( ( R
R ( R R t _writemodest c( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _iswritemodeØ s
c C s/ | i | ƒ o t i i | ƒ S| d f Sd S( sx Split zip extension from filename and return filename.
*Returns*:
base, zip_ext : {tuple}
N( R& R R R" R ( R R# ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _splitzipextâ s c C sT | g } | i | ƒ p7 x4 t i ƒ D]" } | o | i | | ƒ q& q& Wn | S( s9 Return a tuple containing compressed filename variations.( R& R R t append( R R# t namest zipext( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _possible_namesï s
c C s? d d k l } | | ƒ \ } } } } } } t | o | ƒ S( s= Test if path is a net location. Tests the scheme and netloc.iÿÿÿÿ( t urlparse( R1 t bool( R R R1 t schemet netloct upatht uparamst uqueryt ufrag( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _isurlø s c C s d d k l } l } | i | ƒ } t i i t i i | ƒ ƒ p t i t i i | ƒ ƒ n | i | ƒ oV y, | | ƒ } t
| d ƒ i | i ƒ ƒ Wq | j
o | d | ƒ ‚ q XnV y/ t
| d ƒ } t
| d ƒ i | i ƒ ƒ Wn# t
j
o t
d | ƒ ‚ n X| S( sh Cache the file specified by path.
Creates a copy of the file in the datasource cache.
iÿÿÿÿ( t urlopent URLErrorR' s URL not found: %sR s File not found: %s( t urllib2R: R; R R R t existst dirnamet makedirsR9 t filet writet readt IOError( R R R: R; R5 t openedurlt fp( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _cache s c C s· | i | ƒ p/ | i | ƒ } | | i | i | ƒ ƒ 7} n, | i | i | ƒ ƒ } | | i | ƒ } xF | D]> } | i | ƒ o( | i | ƒ o | i | ƒ } n | Sqq Wd S( s Searches for ``path`` and returns full path if found.
If path is an URL, _findfile will cache a local copy and return
the path to the cached file.
If path is a local file, _findfile will return a path to that local
file.
The search will include possible compressed versions of the file and
return the first occurence found.
N( R9 R0 R R= RF R ( R R t filelistt name( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _findfile' s c
C s› d d k l } | i | i d ƒ } t | ƒ d j o | d } n | | ƒ \ } } } } } } | i | ƒ } | i | ƒ } t i i | i | | ƒ S( sF
Return absolute path of file in the DataSource directory.
If `path` is an URL, then `abspath` will return either the location
the file exists locally or the location it would exist when opened
using the `open` method.
Parameters
----------
path : str
Can be a local file or a remote URL.
Returns
-------
out : str
Complete path, including the `DataSource` destination directory.
Notes
-----
The functionality is based on `os.path.abspath`.
iÿÿÿÿ( R1 i i ( R1 t splitR t lent _sanitize_relative_pathR R t join(
R R R1 t splitpathR3 R4 R5 R6 R7 R8 ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR G s c C s… d } t i i | ƒ } xf | | j oX | } | i t i ƒ i d ƒ } | i t i ƒ i d ƒ } t i i | ƒ \ } } q W| S( sv Return a sanitised relative path for which
os.path.abspath(os.path.join(base, path)).startswith(base)
t /s ..N( R R R t normpatht lstript sept pardirt
splitdrive( R R t lastt drive( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyRL r s
c C s› d d k l } l } t i i | ƒ o t S| i | ƒ } t i i | ƒ o t S| i | ƒ o2 y | | ƒ } ~ t SWq— | j
o t SXn t S( s(
Test if path exists.
Test if `path` exists as (and in this order):
- a local file.
- a remote URL that has been downloaded and stored locally in the
`DataSource` directory.
- a remote URL that has not been downloaded, but is valid and accessible.
Parameters
----------
path : str
Can be a local file or a remote URL.
Returns
-------
out : bool
True if `path` exists.
Notes
-----
When `path` is an URL, `exists` will return True if it's either stored
locally in the `DataSource` directory, or is a valid remote URL.
`DataSource` does not discriminate between the two, the file is accessible
if it exists in either location.
iÿÿÿÿ( R: R; (
R< R: R; R R R= R
R R9 R ( R R R: R; R5 t netfile( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR= € s
R c C s¥ | i | ƒ o | i | ƒ o t d ƒ ‚ n | i | ƒ } | oK | i | ƒ \ } } | d j o | i d d ƒ n t | | d | ƒSt d | ƒ ‚ d S( sR
Open and return file-like object.
If `path` is an URL, it will be downloaded, stored in the `DataSource`
directory and opened from there.
Parameters
----------
path : str
Local file path or URL to open.
mode : {'r', 'w', 'a'}, optional
Mode to open `path`. Mode 'r' for reading, 'w' for writing, 'a' to
append. Available modes depend on the type of object specified by
`path`. Default is 'r'.
Returns
-------
out : file object
File object.
s URLs are not writeableR R( t R s
%s not found.N( R9 R+ t
ValueErrorRI R, t replaceR RC ( R R R t foundt _fnameR% ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR ´ s
( R R R R t curdirR R! R& R+ R, R0 R9 RF RI R RL R= R ( ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR ’ s /
+ 4t
Repositoryc B s_ e Z d Z e i d „ Z d „ Z d „ Z d „ Z d „ Z d „ Z
d d „ Z d „ Z RS(
sý
Repository(baseurl, destpath='.')
A data repository where multiple DataSource's share a base URL/directory.
`Repository` extends `DataSource` by prepending a base URL (or directory)
to all the files it handles. Use `Repository` when you will be working
with multiple files from one base URL. Initialize `Repository` with the
base URL, then refer to each file by its filename only.
Parameters
----------
baseurl : str
Path to the local directory or remote location that contains the
data files.
destpath : str or None, optional
Path to the directory where the source file gets downloaded to for use.
If `destpath` is None, a temporary directory will be created.
The default path is the current directory.
Examples
--------
To analyze all files in the repository, do something like this
(note: this is not self-contained code)::
>>> repos = np.lib._datasource.Repository('/home/user/data/dir/')
>>> for filename in filelist:
... fp = repos.open(filename)
... fp.analyze()
... fp.close()
Similarly you could use a URL for a repository::
>>> repos = np.lib._datasource.Repository('http://www.xyz.edu/data')
c C s t i | d | ƒ| | _ d S( s> Create a Repository with a shared url or directory of baseurl.R N( R R t _baseurl( R t baseurlR ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR s c C s t i | ƒ d S( N( R R! ( R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR!
s c C sN | i | i d ƒ } t | ƒ d j o t i i | i | ƒ } n | } | S( s> Return complete path for path. Prepends baseurl if necessary.i i ( RJ R_ RK R R RM ( R R RN t result( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt _fullpath
s
c C s t i | | i | ƒ ƒ S( s8 Extend DataSource method to prepend baseurl to ``path``.( R RI Rb ( R R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyRI s c C s t i | | i | ƒ ƒ S( s_
Return absolute path of file in the Repository directory.
If `path` is an URL, then `abspath` will return either the location
the file exists locally or the location it would exist when opened
using the `open` method.
Parameters
----------
path : str
Can be a local file or a remote URL. This may, but does not have
to, include the `baseurl` with which the `Repository` was initialized.
Returns
-------
out : str
Complete path, including the `DataSource` destination directory.
( R R Rb ( R R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR s c C s t i | | i | ƒ ƒ S( sÈ
Test if path exists prepending Repository base URL to path.
Test if `path` exists as (and in this order):
- a local file.
- a remote URL that has been downloaded and stored locally in the
`DataSource` directory.
- a remote URL that has not been downloaded, but is valid and
accessible.
Parameters
----------
path : str
Can be a local file or a remote URL. This may, but does not have
to, include the `baseurl` with which the `Repository` was initialized.
Returns
-------
out : bool
True if `path` exists.
Notes
-----
When `path` is an URL, `exists` will return True if it's either stored
locally in the `DataSource` directory, or is a valid remote URL.
`DataSource` does not discriminate between the two, the file is accessible
if it exists in either location.
( R R= Rb ( R R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR= 0 s R c C s t i | | i | ƒ | ƒ S( sÞ
Open and return file-like object prepending Repository base URL.
If `path` is an URL, it will be downloaded, stored in the DataSource
directory and opened from there.
Parameters
----------
path : str
Local file path or URL to open. This may, but does not have to,
include the `baseurl` with which the `Repository` was initialized.
mode : {'r', 'w', 'a'}, optional
Mode to open `path`. Mode 'r' for reading, 'w' for writing, 'a' to
append. Available modes depend on the type of object specified by
`path`. Default is 'r'.
Returns
-------
out : file object
File object.
( R R Rb ( R R R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR Q s c C s4 | i | i ƒ o
t d ‚ n t i | i ƒ Sd S( s
List files in the source Repository.
Returns
-------
files : list of str
List of file names (not containing a directory part).
Notes
-----
Does not currently work for remote repositories.
s- Directory listing of URLs, not supported yet.N( R9 R_ t NotImplementedErrorR t listdir( R ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyRd j s
(
R R R R R] R R! Rb RI R R= R Rd ( ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyR^ ß s $ ! ( R t
__docformat__R t shutilR t objectR R R] R R R^ ( ( ( s; /usr/lib64/python2.6/site-packages/numpy/lib/_datasource.pyt ! s ? $ÿ N