gsoc-stats

Created 7 years ago
Maintained by skamath
A mini CLI tool to automatically generate user activity graphs using datagrepper
Members 3
Sachin Kamath committed 6 years ago

fedstats-gsoc

A simple CLI tool to gather statistics from datagrepper

Description

This tool helps pull statistics of any Fedora user with an active FAS Account account.

Usage

This tool usesargparse to parse arguments. This can be used in two ways, one-liner / interactive method. The interactive mode can be enabled by using the --interactive or -i flag. This mode of usage does not require any additional argument. Please note that the arguments passed (if any) will be invalid.

One-liner uses the classic argument parsing method to generate output. This is useful for automating the report generation process. The only mandatory argument is --user / -u which takes the FAS username as input.

python stats.py --user=nobody or python stats.py -u nobody will generate text based statistics of user nobody from datagrepper.

Arguments

--interactive / -i

  • Launches the tool in interactive mode. Does not require any further arguments. Some of the argparse variables are yet to be implemented.

--user / -u

  • Takes any FAS Username as argument. There is no default value and the tool will throw an error if this argument is left blank/not used. This argument can also take the value all. If user is set as all, then the unfiltered data is pulled from datagrepper.

--start / -s

  • Takes a date as input in the format MM/DD/YYYY as input, that determines the starting date for which the data is required. This will be internally converted to the epoch time. If this parameter is set along with --end, --delta is ignored.

--end / -e

  • Takes a date as input in the format MM/DD/YYYY as input, that determines the end date for which the data is required. This will be internally converted to the epoch time.

--weeks / -w

  • Takes an integer value to represent number of weeks. Converts it into time-delta. (1week = 604,800 seconds). Default value is 1. This values is ignored if the --start and --end values are set.

--mode / -m

  • Takes a single word string input. Supported input modes are : json, text, csv, markdown, svg and png. The default value is text. (More features will be added soon)

--category / -c

  • Take a single word string input. This will define the category for which deeper analytics are required. Some example categories : pagure, irc, mailman, etc.

--output / -o

  • Takes a single word string input. This will define the output file name. This option is to be combined with the --mode/-m argument. Please note that this option DOES NOT require an extension type. For instance, if you need an SVG output with the name nobody.svg, the --output flag should be set as nobody and not nobody.svg. the default value is stats.

--group / -g

  • Takes a single word string input. This will take a group name of FAS as an argument. Useful for pulling statistics of an entire group. Please note that this requires authentication and the credentials are to be put in fas_credentials.cfg file. Not to be combined with --user / -u

--logging / -l

  • Stores the value as True if called, does not require any argument. When logging is set, all the logs from start to end / mentioned weeks will be pulled from datagrepper and dumped into a text file according to the naming convention.

File naming convention

All files generated by this tool are systematically named to avoid confusion. Combined with the file extension, there won't be two files with similar file names, hence preventing unexpected over-writes.

The naming convention is as follows :

  • All the main report files (i.e : Category Overview), text files are named as <username>_main.<extension>

  • The sub-category report (i.e the Category bar-chart) is named as <username>_<category>.<extension>

  • The further category interaction report (i.e The sub-categories chart) is named as <username>_<category>_<sub_category>.<extension>

Examples

  • Generate statistics of user nobody for a week and view the text logs :

python main.py --user=nobody

  • Generate statistics of user foo in .svg format :

python main.py --user=foo --mode=svg

This will create stats.svg in your $pwd.

  • Generate statistics of user foo in .png format with output name as foo_stats.png

python main.py --user=foo --mode=png --output=foo_stats

  • Generate data between specific dates and locally save as JSON

python main.py --user=foo --mode=json --start=05/23/2016 --end=05/28/2016

Basic Troubleshooting :

Please take a look at this blog-post.