An tool to create textual reports out of Call Detail Records (CDR) as they are created by the Asterisk software.
Important: This is an after-hours hobby for me so do not expect me to have fast progress or fix any bugs. The company Digium owns the trademark of the name Asterisk. Digium is in no way affiliated (e.g. sponsoring, approving, …) with me or this project.
Reports are a great way to obtain concise statistics of your PBX. For example, you can count the number of calls and produce daily email reports to prevent telephone fraud.
To generate reports, you need to write two files, a report definition and a template definition. Then a report can be generated like this:
$cd cmd/report
go run report.go -template ~/reportTemplate.gohtml -definition ~/reportDefinition.json /var/log/asterisk/cdrcsv/Master.csv
Per default, this command escapes special HTML characters. If you do not want this, pass the -plain argument.
First, you need a JSON file containing a definition what you want to count. See reportDefinition.json for an example. The allowed column names can be found in in parse.go.
Second, you need the actual report template as text file or html file, for example reportTemplate.tmpl
The syntax is Go's template engine syntax and with the {{index .Stats "stat_name"}} you can access the statistics
defined in your JSON file. See the linked file for more examples on how to use template files. Additionally, you can
make use of the Go Template Specification.
In order to test report generation I had to write a tool that replaces personal data in CDR records by pseudonymified data. After pseudonymization all calls are the same relative to each other, only with other data. This means that, for example, you can still see that a particular number has called a lot of times, but the number is not the real number.
You can execute the tool with these commands:
$ cd cmd/pseudonymify
$ go run ./pseudo.go -contacts ../mockdata/persons.csv -years 1 -days -20 -minutes 5 -hours -3 -contexts hq,production,support,door,hr /tmp/cdr*.csv
The last arguments the csv files to by pseudonymified. The tool replaces all personal data with the data found in the
persons.csv provided with the -contact argument. It also shifts all calls by the time you specify with the arguments
-years, -days, -minutes and -hours. The -context parameter lets you define a list of context to be used to
replace the contexts in the real csv.