[Tutor] Regex for Filesystem path (Asad)

[Asad]

Hi All,

         I tired seems its not working as required :

from os.path import dirname, join

testdir = dirname("/a/b/c/d/test/test_2814__2018_10_05_12_12_45/logA.log")

dirpath = join(testdir, '123456/789')

print dirpath

/a/b/c/d/test/test_2814__2018_10_05_12_12_45\123456/789

Instead i need the script to go to the location :

/a/b/c/d/test/123456/789

Please advice .


Thanks,









> ---------- Forwarded message ----------
> From: Cameron Simpson <cs at cskk.id.au>
> To: tutor at python.org
> Cc:
> Bcc:
> Date: Wed, 7 Nov 2018 06:47:33 +1100
> Subject: Re: [Tutor] Regex for Filesystem path
> On 06Nov2018 18:10, Alan Gauld <alan.gauld at yahoo.co.uk> wrote:
> >On 06/11/2018 13:13, Asad wrote:
> >
> >>         Can you provide some advice and code for the following problem :
> >
> >The first thing is to go read the documentation for the os.path module.
> >It is designed for reliable path manipulation.
> >
> >> /a/b/c/d/test/test_2814__2018_10_05_12_12_45/logA.log
> >>
> >> f3 = open ( r"/a/b/c/d/test/test_2814__2018_10_05_12_12_45/logA.log",
> 'r' )
> >> st1 = f3.readlines ()
> >
> >You hardly ever need readlines() any more, just iterate
> >over the file, its much easier.
> >
> >> for j in range ( len ( st1 ) ):
> >
> >for line in f3:
>
> Not to mention cheaper in memory usage.
>
> [...snip...]
> >>    a = mo.group()   ## 123456/789
> >>    ===================================================================
> >>    How to do I traverse to the required directory which is
> >> /a/b/c/d/test/123456/789 ?
> >
> >You can use relative paths in os.chdir.
> >So a payth of '..' will be one level up from the current
> >directory. Of course you need to chdir to that directory first
> >but os.path will tell you the dir you need.
>
> It is better to just construct the required path. Chdir there requires a
> chdir back, and chdir affects all the relative paths your programme may
> be using.
>
> I'd use os.path.dirname to get '/a/b/c/d/test' and then just append to
> it with os.path.join to contruct each directory path.
>
> [...]
> >But I'm guessing that's too obvious so the path may vary?
> >>    1) First I need to extract /a/b/c/d/test/  from
> >>  /a/b/c/d/test/test_2814__2018_10_05_12_12_45/logA.log  ?
>
> Use os.path.dirname:
>
>    # up the top
>    from os.path import dirname, join
>
>    # later
>    testdir = dirname(logfile_path)
>
> >get the dir then chdir to .. from there.
> >
> >>    2) Then add 123456/789 and create directory location as
> >> /a/b/c/d/test/123456/789
> >
> >Simple string manipulation or use the os.path functions.
>
> Eg dirpath = join(testdir, '123456/789')
>
> >>    3) cd /a/b/c/d/test/123456/789
> >
> >os.chdir()
>
> I still recommend avoiding this. Just construct the full path to what
> you need.
>
> >>    4) look for the latest file in the directory
> /a/b/c/d/test/123456/789
> >
> >Slightly more complex, you need the creation timestamp.
> >You can find that with os.path.getctime() (or several
> >other options, eg os.stat)
>
> Do not use ctime, it is _not_ "creation" time. It is "last change to
> inode" time. It _starts_ as creation time, but a chmod or even a
> link/unlink can change it: anything that changes the metadata.
>
> Generally people want mtime (last nmodified time), which is the last
> time the file data got changed. It is more meaningful.
>
> Cheers,
> Cameron Simpson <cs at cskk.id.au>
>
>
>
>
> --
Asad Hasan
+91 9582111698