How to count file in function of their age in Python?

Question

I'm trying to count how many file are very young, young, old and very old in a directory passed by command line. I have some struggle counting the number of file in the directory and then to change the counter in function of the age of the file.

Here is what I did yet:

import sys
import os, os.path
import time

x = 7
file_count = 0
DIR = sys.argv[1]
age = 0
age_sum = 0
nb_vy = 0
nb_y = 0
nb_o = 0
nb_vo = 0
now = time.time()

for root, subFolders, files in os.walk(DIR):
    for file in (files):
        try:
            file_count += 1
            # here I want to do some if to add 1 in the counter nb_vy/nb_y/nb_o/nb_vo but I don't know how  
            age=now - os.stat(os.path.join(root,file)).st_mtime
            age_sum+=age
            if  now-timedelta(hours=24) <= age <= now :
                nb_vy += 1
            elif now-timedelta(days=7) <= age :
                nb_y += 1
            elif now-timedelta(days=30) <= age :
                nb_o += 1
            else:
                nb_vo += 1
        except Exception:
            pass


print("Filecount = %s" % file_count)


print("--------------------------------------------------------------------\n")
print("Scanned:  "+  str(file_count) +" files\n")
print("Average age:  "+  str(age/age_sum) + "\n")
print("--------------------------------------------------------------------\n")
print("Very young files (<=  1  day) |  "+  str(nb_vy/file_count) +"% (" + str(nb_vy) + ")\n")
print("Young files      (<=  1 week) |  "+  str(nb_y/file_count) +"% (" + str(nb_v) + ")\n")
print("Old files        (<= 30 days) |  "+  str(nb_o/file_count) +"% (" + str(nb_o) + ")\n")
print("Very old files   (>  30  days |  "+  str(nb_vo/file_count) +"% (" + str(nb_vo) + ")\n")
print("--------------------------------------------------------------------\n")

How can I manage the if cascade to increment the right counter ?

I feel like the part who is causing trouble is the `age = now - os.stat(file).st_mtime` but I don't know why. I fixed it with `age=now - os.stat(os.path.join(root,file)).st_mtime` but it's not very clean — J.erome, Dec 18 '19 at 10:09
`os.path.join(root,file)` is correct - else you only get the filename, and `os.stat()` then search the file in [the current working directory](https://en.wikipedia.org/wiki/Working_directory). — bruno desthuilliers, Dec 18 '19 at 10:24

Dan D. · Accepted Answer · 2019-12-18T10:51:47.470

4

You had the sides of the comparison swapped and the unneeded now - was still there. Once those are fixed and the timedelta converted to a duration of seconds for comparison:

        if  age <= timedelta(hours=24).total_seconds():
            nb_vy += 1
        elif age <= timedelta(days=7).total_seconds():
            nb_y += 1
        elif age <= timedelta(days=30).total_seconds():
            nb_o += 1
        else:
            nb_vo += 1

You should be using age < max_age_for_group as the condition. age is already now - mtime. age is in seconds.

Also except Exception: pass will harm debugging. If you must have it at least use:

except Exception:
    logging.exception('')

This will eat the exception but still print it. And then you can turn the printing off by changing the level on the root logger.

edited Dec 18 '19 at 10:51

answered Dec 18 '19 at 10:23

Dan D.

73,243
15
104
123

I have some struggle to find the max_age_for_group, I'm not used to work with time object I tried to follow [link][https://stackoverflow.com/questions/1345827/how-do-i-find-the-time-difference-between-two-datetime-objects-in-python] but I can't make it work fine – J.erome Dec 18 '19 at 10:24
Could you check my edit it seems like I did something wrong with comparing timestamp please @Dan D. – J.erome Dec 18 '19 at 10:39
I fixed like you did it but still no counter increments their value – J.erome Dec 18 '19 at 10:57

score 1 · Answer 2 · answered Dec 18 '19 at 10:45

You correctly calculate the age of the file with calculating the difference in time using:

now = time.time()
mt = os.stat(os.path.join(root,file)).st_mtime
age = now - mt

All of the above variable are stored as float and represent the time in seconds ellapsed since January 1st 1970. So the difference age is a time difference in seconds!

In order to categorize into the intervals specified, you need to convert this time difference to days, e.g. with:

# import datetime
td = datetime.timedelta(0, age)
print(td.days) # Prints the time delta in days (integer)

I think, this is exactly what you need for you comparison, as if the time difference is, e.g. 72000 seconds (=20 hours), td.days will be evalutated to 0.

I tried to use `timedelta` in my `if elif elif else` can you spot what I did wrong in ther e? — J.erome, Dec 18 '19 at 10:50
@Dan D. pointed out, what is wrong with your comparisons. You must ensure to compare the right units of time, thus convert to either seconds or days. — AnsFourtyTwo, Dec 18 '19 at 10:55

How to count file in function of their age in Python?

2 Answers2