@iqb/csv2ttl v0.7.7

Use this package to transform vocabularies from csv files of specific format to ttl files.

It's a pain to write ttl-files manually. But it's also a great idea to provide metadata vocabularies via ttl-files, especially if one uses skos or even skohub to generate shiny html or handy json.

So we support only a special use case of ttf/rdf: Hierarchical vocabularies. Here we use these attributes of rtf:

• prefLabel (required): Multilingual name of the entry,
• definition: Multilingual description to give more information about the entry, and
• notation: Semver-like enumeration mainly to describe the hierarchy, like 2.4.3.

Workflow

1. First write the csv-File(s)

Look at the simplest example:

notation,title,description,id
,geschlossen,,
,halboffen,,
,offen,,

The first line and therefor the description of data structure is required. You see, that only the title is given.

2. Write csv2ttl_config.json

There is one json-file to provide with some parameters for the generation process. The idea is, that one (git)-repository contains of multiple vocabularies describing one common issue. We call this repo a 'vocabulary-group'. The config file lists all vocabularies with id and name.

3. Add IDs

Adding IDs is a precondition for the generation of ttl-files. This csv2ttl-package contains of two commands: addidcsv and csv2ttl. Use the first one to add IDs to the csv-files. Every entry gets an id randomly. Existing ids will not be overwritten.

4. Generate

Use the csv2ttl-command to get the ttl-files.

Create repo and install

First create a node repo. Precondition for that is of course a working node installation. Just create a new folder, go into this folder and type:

npm init

You will be prompted for some parameters. After that, type:

npm install @iqb/csv2ttl

Put all csv-files in that folder and write the csv2ttl_config.json.

After that, edit the package.json to have short access to the commands:

{
  "scripts": {
    "add": "addidcsv",
    "csv": "csv2ttl"
  }
}

To run these scripts, use your IDE or type

npm run add
npm run csv

If you like to store the source files in a separate folder, add a parameter when calling a script:

npm run add csv-src

Customise: csv2ttl_config.json

This json-file lists all vocabularies (i.e. csv-files) to convert and provides additional parameters for the generation process.

Before processing the config file it is validated. You'll get errors if it's malformed.

5. Multi-language support

From the 0.7.6 csv2ttl supports multilanguage. The csv files should have the following format:

notation,title,description,id,title_en,description_en
,geschlossen,,,closed,
,halboffen,,,half open,
,offen,,,open,

The configuration example for multi-language support is available sample_data/csv2ttl_configEn.json.