Statistics

[Jasata]

Probably requires a cron job to process Nginx logs.

Wanted statistics:
download counts for each image, over weeks (1 - 52/53). Count unfinished downloads separately.

DO NOT IMPLEMENT BEFORE UPLOAD SOLUTION HAS BEEN COMPLETED AND TESTED! Download might also be changed into HTML5 API implementation, and if so, then the statistics need to be maintained and recorded into a database.

Leave this until the download mechanism is truly fixed!

[Jasata]

Download has been stable for half a year now.

What was drafted and how was it going to be implemented, has been forgotten. Reason why it is not included here, is that Ari/Timo team was looking into it and what exists, is somewhere in GitLab...

[Jasata]

Backend implementation started as a student project. Frontend is not currently tasked to anyone.

[Jasata]

Mona's version needs to be found... This also exists:

#! /usr/bin/env python3
# -*- coding: utf-8 -*-
#
# Turku University (2019) Department of Future Technologies
#
# pal.py - Parse Access Log
#
#   0.1.0   2020.01.17  Initial version.
#
import os
import syslog
from datetime import datetime

LOGFILE = "vm.utu.fi.access.log.1"

syslog.openlog("access.log parser")
with open(LOGFILE, "r") as log:
    for line in log:
        try:
            item = line.split()
            if item[5] == '"GET' and item[6].startswith('/download/'):
                print(
                    datetime.strptime(
                        item[3][1:],
                        '%d/%b/%Y:%H:%M:%S'
                    ),
                    os.path.split(item[6])[1],
                    item[9]
                )
        except Exception as e:
            syslog.syslog("Rat bastards! Who screwed up the log format?")
            with open("troublemakers.log", "a") as trbl:
                trbl.write(line)