Feature request ability to use faker to generate more robust data

[rsaylor73]

The process now that generates random hash data works but it would be useful if we could pipe to a faker data generator to be able to take the production data and replace with test data that is human readable.

https://github.com/joke2k/faker

[ppomes]

Hi @rsaylor73,

Sorry for the so looooong delay.... Just to let you know, I started to work on this request.

Pierre

[alexliam83]

Hi @ppomes,

Is there any documentation about this update?

Thank you in advance!

[ppomes]

Nope :-( Finally did not have any time to do it.... :-(

[ppomes]

@rsaylor73, @alexliam83 , @davidmr163,

I just added some code as a draft, to provide python support. Python faker can now be used.

You need to install python-dev libraries (for Ubuntu: apt-get install python3-dev)

Then, you need to build from branch issue5:

git checkout issue5
git pull
./autogen.sh
./configure --enable-python
make
make check

For the faker example, assuming faker is installed (for Ubuntu : apt-get install python3-fake-factory)

PYTHONPATH=$(pwd)/tests main/myanon -f tests/faker_test.conf < tests/faker_test.sql

Key points are:

• a python script is given in the global section of the config file (without .py extension)
• the path of this script must be set by the env variable PYTHONPATH
• this script must contain all the functions you need (ie : faker function for names, faker function for emails, etc...)
• for each field, you can select which faker function you want to use

Doc is not yet updated.

Feedback needed! If this new feature seems ok for you, I will do more tests/improvements, I will update the doc and merge into main branch.

Thanks!
Pierre