Usage

docker run -v $(pwd):/input cloudtooling/data-anonymizer -t xml -i "input/my.xml:(type=street,xpath=//Strasse)"

or

inputFile=input/my.xml
docker run -v $(pwd):/input cloudtooling/data-anonymizer -t xml -i \
  "$inputFile:(type=street,xpath=//Strasse)" \
  "$inputFile:(type=number,xpath=//HausNr)" \
  "$inputFile:(type=zip,xpath=//PLZ)" \
  "$inputFile:(type=city,xpath=//Ort)" \
  "$inputFile:(type=city_suffix,xpath=//Ortsteil)" \
  "$inputFile:(type=city_suffix,xpath=//KreisRegion)" \
  "$inputFile:(type=last_name,xpath=//Nachname)" \
  "$inputFile:(type=first_name,xpath=//Vorname)" \
  "$inputFile:(type=name,xpath=//Ansprechpartner)" \
  "$inputFile:(type=name,xpath=//GeschFuehrer)" \
  "$inputFile:(type=passport_number,xpath=//HRNr)" \
  "$inputFile:(type=url,xpath=//Homepage)" \
  "$inputFile:(type=email,xpath=//Email)" \
  "$inputFile:(type=phone_number,xpath=//TelefonNr)" \
  "$inputFile:(type=phone_number,xpath=//MobilNr)" \
  "$inputFile:(type=phone_number,xpath=//FaxNr)"

# Anonymize email column in CSV
... -i data.csv:(type=email,column=2)

# Anonymize multiple columns
...  -i data.csv:(type=first_name,column=0) data.csv:(type=last_name,column=1)

# With wildcards
... -i data*.csv:(type=email,column=2)

# Overwrite original file
...  -i data.csv:(type=email,column=2) -o

# Custom delimiter (comma)
...  -i data.csv:(type=email,column=2) -d ','

# Different locale
...  -i data.csv:(type=name,column=0) -l en_US

CLI Arguments

Short	Long	Destination	Default	Action	Description
-i	--input	input	—	extend	One or more input sources. Examples: • CSV: inputfile1:(type=number,column=0) • XML: inputfile1:(type=last_name,xpath=./person/lastname) • SQLite: sqlite://[username:password@]server/database:(input_type=db,type=first_name,table=people,column=first_name) Use multiple arguments to anonymize across multiple files. Supports mixing types (csv, xml, json, …) and wildcards (*, ?).
-t	--type	type	number	—	Type of data to anonymize (e.g., name, first_name, last_name, email, zip, city, address, number, …).
-e	--encoding	encoding	ISO-8859-15	—	File encoding for reading/writing. Example: UTF-8.
-d	--delimiter	delimiter	;	—	CSV column delimiter. Use --delimiter $'\t' for tab-separated files.
-l	--locale	locale	de_DE	—	Locale for generating fake data (e.g., en_US, fr_FR).
-o	--overwrite	overwrite	False	store_true	Overwrite original file(s) with anonymized data.
-j	--ignore-missing-file	ignoreMissingFile	False	store_true	Ignore missing files instead of failing.
—	--header-lines	headerLines	0	—	Number of header lines in CSV files to skip.
—	--namespace	namespace	—	—	Define XML namespaces. Syntax: shortname=http://full-url-of-namespace.com Multiple can be provided, separated by spaces.

Check also .bin/tests.sh for some sample usages.

See here for supported faker types.

Extending

As Python Module

from faker import Factory
from jinja2 import Environment
from selector import Selector
from csv_anonymizer import anonymize_csv
from anonymizer import unidecode_filter

faker = Factory.create('de_DE')
template_env = Environment()
template_env.filters['unidecode'] = unidecode_filter

selector = Selector("(type=email,column=2)")
anonymize_csv(
    'input.csv', 'output.csv', [selector],
    0, 'utf-8', ';', faker, template_env
)