|
1 | | -# mapper-dnb-ubo |
| 1 | +# mapper-dnb |
| 2 | + |
| 3 | +## Overview |
| 4 | + |
| 5 | +The [dnb_mapper.py](dnb_mapper.py) python script converts Dun & Bradstreet (DNB) files to json files ready to load into Senzing. This includes the following formats ... |
| 6 | +- Companies and their principles **(CMPCVF)** json format |
| 7 | +- Global contacts **(GCA)** tab delimited csv format |
| 8 | +- Ultimate beneficial owners **(UBO)** tab delinited csv format |
| 9 | + |
| 10 | +Normally these are provided by DNB on request and placed on an FTP server for you to download. |
| 11 | + |
| 12 | +*Warning: the [dnb_formats.json](dnb_formats.json) file contains the exact structure of these files. You may need to send these formats to DNB so they know exactly how to create them!* |
| 13 | + |
| 14 | +Loading DNB data into Senzing requires additional features and configurations. These are contained in the |
| 15 | +[dnb_config_updates.json](dnb_config_updates.json) file. |
| 16 | + |
| 17 | +Usage: |
| 18 | +```console |
| 19 | +python3 dnb_mapper.py --help |
| 20 | +usage: dnb_mapper.py [-h] [-f DNB_FORMAT] [-i INPUT_SPEC] [-o OUTPUT_PATH] |
| 21 | + [-l LOG_FILE] |
| 22 | + |
| 23 | +optional arguments: |
| 24 | + -h, --help show this help message and exit |
| 25 | + -f DNB_FORMAT, --dnb_format DNB_FORMAT |
| 26 | + choose CMPCVF, UBO, or GCA |
| 27 | + -i INPUT_SPEC, --input_spec INPUT_SPEC |
| 28 | + the name of one or more DNB files to map (place in |
| 29 | + quotes if you use wild cards) |
| 30 | + -o OUTPUT_PATH, --output_path OUTPUT_PATH |
| 31 | + output directory or file name for mapped json records |
| 32 | + -l LOG_FILE, --log_file LOG_FILE |
| 33 | + optional statistics filename (json format). |
| 34 | +``` |
| 35 | + |
| 36 | +## Contents |
| 37 | + |
| 38 | +1. [Prerequisites](#Prerequisites) |
| 39 | +2. [Installation](#Installation) |
| 40 | +3. [Configuring Senzing](#Configuring-Senzing) |
| 41 | +4. [Running the mapper](#Running-the-mapper) |
| 42 | +5. [Loading into Senzing](#Loading-into-Senzing) |
| 43 | + |
| 44 | +### Prerequisites |
| 45 | +- python 3.6 or higher |
| 46 | +- Senzing API version 1.7 or higher |
| 47 | +- https://github.com/Senzing/mapper-base |
| 48 | + |
| 49 | +### Installation |
| 50 | + |
| 51 | +Place the the following files on a directory of your choice ... |
| 52 | +- [dnb_mapper.py](dnb_mapper.py) |
| 53 | +- [dnb_config_updates.json](dnb_config_updates.json) |
| 54 | +- [dnb_formats.json](dnb_formats.json) |
| 55 | + |
| 56 | +*Note: Since the mapper-base project referenced above is required by this mapper, it is necessary to place them in a common directory structure like so ...* |
| 57 | +```Console |
| 58 | +/senzing/mappers/mapper-base |
| 59 | +/senzing/mappers/mapper-dnb <-- |
| 60 | +``` |
| 61 | +You will also need to set the PYTHONPATH to where the base mapper is as follows ... (assumuing the directory structure above) |
| 62 | +```Console |
| 63 | +export PYTHONPATH=$PYTHONPATH:/senzing/mappers/mapper-base |
| 64 | +``` |
| 65 | + |
| 66 | +### Configuring Senzing |
| 67 | + |
| 68 | +*Note:* This only needs to be performed one time! In fact you may want to add these configuration updates to a master configuration file for all your data sources. |
| 69 | + |
| 70 | +From the /opt/senzing/g2/python directory ... |
| 71 | +```console |
| 72 | +python3 G2ConfigTool.py <path-to-file>/dnb_config_updates.json |
| 73 | +``` |
| 74 | +This will step you through the process of adding the data sources, entity types, features, attributes and other settings needed to load this watch list data into Senzing. After each command you will see a status message saying "success" or "already exists". For instance, if you run the script twice, the second time through they will all say "already exists" which is OK. |
| 75 | + |
| 76 | +Configuration updates include: |
| 77 | +- addDataSource **DNB-COMPANY** used when when mapping companies from CMPCVF json files |
| 78 | +- addDataSource **DNB-PRINCIPLE** used when when mapping principles from CMPCVF json files |
| 79 | +- addDataSource **DNB-OWNER** used when when mapping owners from UBO csv files |
| 80 | +- addDataSource **DNB-CONTACT** used when when mapping contacts from GCA csv files |
| 81 | +- addEntityType **PERSON** |
| 82 | +- addEntityType **ORGANIZATION** |
| 83 | +- add features and attributes for ... |
| 84 | + - **DNB_OWNER_ID** This is used to help prevent owners from resolving to each other and so that you can search on it. |
| 85 | + |
| 86 | +### Running the mapper |
| 87 | + |
| 88 | +First, download the DNB files you want to load from the DNB FTP site. Since the data files are so large, these are normally split into multiple files. |
| 89 | + |
| 90 | +Second, run the mapper. Example usage: |
| 91 | +```console |
| 92 | +python3 dnb_mapper.py -f CMPCVF -i "./input/CMPCVF*.txt" -o ./output -l cmpcvf_stats.json |
| 93 | + |
| 94 | +python3 dnb_mapper.py -f GCA -i "./input/GCA*.txt" -o ./output -l gca_stats.json |
| 95 | + |
| 96 | +python3 dnb_mapper.py -f UBO -i "./input/UBO*.txt" -o ./output -l ubo_stats.json |
| 97 | +``` |
| 98 | +The output file defaults to the same name and location as the input file and a .json extension is added. |
| 99 | + |
| 100 | +*It is critical that the -f file format match the input files exactly!* |
| 101 | + |
| 102 | +### Loading into Senzing |
| 103 | + |
| 104 | +If you use the G2Loader program to load your data, its best to list the mapped json files you want to load in a project file. There is an example of one in your senzing instalation here: /opt/senzing/g2/python/demo/sample/project.csv. Then from from the /opt/senzing/g2/python directory ... |
| 105 | +```console |
| 106 | +python3 G2Loader.py -p <name of project file> |
| 107 | + |
| 108 | +If you use the API directly, then you just need to perform an process() or addRecord() for each line of each mapped file. |
| 109 | +``` |
0 commit comments