ezPAARSE is an open-source software that can ingest your (proxy) log files and show how users access suscribed electronic ressources. It filters, extracts and enriches the consultation events that were spotted and produces a CSV file following COUNTER codes of practice. This document describes how to install and run ezPAARSE on your computer.
Built-in proxies supported log formats are:
The tools you need to let ezPAARSE run are:
- Linux OS: See the prerequisites for those OSes
- Standard Linux tools: bash, make, grep, sed ...
- curl (used by nvm)
- unzip
- git >= 1.7.10 (required from github)
ezPAARSE then comes with all the elements it needs to run. When the prerequesites are met, you can launch the make command (see below) that will run all installation steps.
If you are a Windows user, you can install ezPAARSE on your computer by downloading the setup file and start the install like you would do for any other program.
To install the latest stable version of ezPAARSE on a Unix-type system, open a terminal and type:
git clone https://github.com/ezpaarse-project/ezpaarse.git
cd ezpaarse
git checkout `git describe --tags --abbrev=0`
make
You can also download a stable version in the tar.gz format and type in a terminal:
tar -xzfv ezpaarse-X.X.X.tar.gz
cd ezpaarse-X.X.X
make
If you want to install the version in development (unstable), open a terminal and type:
git clone https://github.com/ezpaarse-project/ezpaarse.git
cd ezpaarse
make
You can also download a deb archive and install it on your system this way:
sudo -E dpkg -i ezpaarse-X.X.X.deb
sudo /etc/init.d/ezpaarse start
This step allows you to validate that your install is working.
make start
make test
Anonymised example logfiles are made available in the repositories of ezPAARSE.
You need to make sure that ezPAARSE is started. To do so, type the following command:
make start
If you are not computer-savvy, the easiest way to work with ezPAARSE is to use its hmtl form, accessible from your favorite webbrowser and open the following URL: http://localhost:59599/
If you are computer-savvy, you can use an HTTP client (like curl) to send a logfile (for this example, we will use ./test/dataset/sd.2012-11-30.300.log) to ezPAARSE's Web service and get a CSV stream of consultation events as a response.
curl -X POST http://127.0.0.1:59599 \
-v --proxy "" --no-buffer \
--data-binary @./test/dataset/sd.2012-11-30.300.log
Or you can use the command ./bin/loginjector
ezPAARSE provides you with
to send the logfile to the web service in a simpler way:
. ./bin/env
cat ./test/dataset/sd.2012-11-30.300.log | ./bin/loginjector
You can also see quick countings on your data if you add the command
./bin/csvtotalizer
at the end of the command line.
Doing so, you will get an overview of the consultation events extracted
from your logs by ezPAARSE:
. ./bin/env
cat ./test/dataset/sd.2012-11-30.300.log | ./bin/loginjector | ./bin/csvtotalizer
To stop ezPAARSE, you have to type the following command:
make stop
The default ezPAARSE parameters can be found in the config.json
file. All these parameters can be changed. A good practice is to define a new file called config.local.json
containing just the parameters you need to override.
For example, to change the ezPAARSE listening port (59599 by default), you can override the EZPAARSE_NODEJS_PORT
by defining a new config.local.json
file this way:
{
"EZPAARSE_NODEJS_PORT": 45000
}