Converter from CSV file to a multilingual IMS VDEX vocabulary XML file

VDEX is a very good standardized format for multilingual vocabularies, ontologies, etc. It just sucks to create its XML manually. There is poor editor support. But everybody has Excel, well, but almost everybody knows how to create tables. So let the user create a sheet with a column of keys for each term and for each language a column with the translated terms value.

A flat vocabulary

As a CSV this looks like:

"key";"english";"german";"italian"
"k01";"ant";"Ameise";"formica"
"k02";"bee";"Biene";"ape"
"k03";"wasp";"Wespe";"vespa"
"k04";"hornet";"Hornisse";"calabrone"

After running through csv2vdex, called like so:

csv2vdex insects 'insects,Insekten,insetto' \
         insects.csv insects.xml --languages en,de,it --startrow 1

This results in such a VDEX XML:

<vdex xmlns="http://www.imsglobal.org/xsd/imsvdex_v1p0" orderSignificant="true">
  <vocabIdentifier>insects</vocabIdentifier>
  <vocabName>
    <langstring language="en">insects</langstring>
    <langstring language="de">Insekten</langstring>
    <langstring language="it">insetto</langstring>
  </vocabName>
  <term>
    <termIdentifier>k01</termIdentifier>
    <caption>
      <langstring language="en">ant</langstring>
      <langstring language="de">Ameise</langstring>
      <langstring language="it">formica</langstring>
    </caption>
  </term>
  <term>
    <termIdentifier>k02</termIdentifier>
    <caption>
      <langstring language="en">bee</langstring>
      <langstring language="de">Biene</langstring>
      <langstring language="it">ape</langstring>
    </caption>
  </term>
  <term>
    <termIdentifier>k03</termIdentifier>
    <caption>
      <langstring language="en">wasp</langstring>
      <langstring language="de">Wespe</langstring>
      <langstring language="it">vespa</langstring>
    </caption>
  </term>
  <term>
    <termIdentifier>k04</termIdentifier>
    <caption>
      <langstring language="en">hornet</langstring>
      <langstring language="de">Hornisse</langstring>
      <langstring language="it">calabrone</langstring>
    </caption>
  </term>
</vdex>

A tree vocabulary

If we want to have a tree-like vocabulary, the key is used to define the level. Here a dot is used as delimiter.

As a CSV it looks like:

"key";"term value"
"nwe";"North-west of Europe"
"nwe.1";"A. m. iberica"
"nwe.2";"A. m. intermissa"
"nwe.3";"A. m. lihzeni"
"nwe.4";"A. m. mellifera"
"nwe.5";"A. m. sahariensis"
"swe";"South-west of Europe"
"swe.1";"A. m. carnica"
"swe.2";"A. m. cecropia"
"swe.3";"A. m. ligustica"
"swe.4";"A. m. macedonica"
"swe.5";"A. m. ruttneri"
"swe.6";"A. m. sicula"

After running through csv2vdex, called like so:

csv2vdex beeeurope 'European Honey Bees' bees.csv bees.xml -s 1

The result is:

<vdex xmlns="http://www.imsglobal.org/xsd/imsvdex_v1p0" orderSignificant="true">
  <vocabIdentifier>beeeurope</vocabIdentifier>
  <vocabName>
    <langstring language="en">European Honey Bees</langstring>
  </vocabName>
  <term>
    <termIdentifier>nwe</termIdentifier>
    <caption>
      <langstring language="en">North-west of Europe</langstring>
    </caption>
    <term>
      <termIdentifier>nwe.1</termIdentifier>
      <caption>
        <langstring language="en">A. m. iberica</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>nwe.2</termIdentifier>
      <caption>
        <langstring language="en">A. m. intermissa</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>nwe.3</termIdentifier>
      <caption>
        <langstring language="en">A. m. lihzeni</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>nwe.4</termIdentifier>
      <caption>
        <langstring language="en">A. m. mellifera</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>nwe.5</termIdentifier>
      <caption>
        <langstring language="en">A. m. sahariensis</langstring>
      </caption>
    </term>
  </term>
  <term>
    <termIdentifier>swe</termIdentifier>
    <caption>
      <langstring language="en">South-west of Europe</langstring>
    </caption>
    <term>
      <termIdentifier>swe.1</termIdentifier>
      <caption>
        <langstring language="en">A. m. carnica</langstring>
      </caption>
    </term>
    <term>
   <term>
      <termIdentifier>swe.2</termIdentifier>
      <caption>
        <langstring language="en">A. m. cecropia</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>swe.3</termIdentifier>
      <caption>
        <langstring language="en">A. m. ligustica</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>swe.4</termIdentifier>
      <caption>
        <langstring language="en">A. m. macedonica</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>swe.5</termIdentifier>
      <caption>
        <langstring language="en">A. m. ruttneri</langstring>
      </caption>
    </term>
    <term>
      <termIdentifier>swe.6</termIdentifier>
      <caption>
        <langstring language="en">A. m. sicula</langstring>
      </caption>
    </term>
  </term>
</vdex>

A tree-vocabulary with descriptions

As a CSV this looks like:

field_work_terms,Field work terms,
field_work_terms.1,Acidification,"Acidification is a process. It happens naturally ..."
field_work_terms.2,Aquifer,"If you get a shovel and dig at the ground below your ..."
field_work_terms.3,Biodiversity,"This has many contentious meanings but for our ..."

After running through csv2vdex, called like so:

csv2vdex --description True --csvdelimiter "," terms "Terminology" terms.csv terms.xml

This results in such a VDEX XML:

<vdex xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://www.imsglobal.org/xsd/imsvdex_v1p0" xsi:schemaLocation="http://www.imsglobal.org/imsvdex_v1p0 imsvdex_v1p0.xsd" profileType="lax" orderSignificant="true">
  <vocabIdentifier>terms</vocabIdentifier>
  <vocabName>
    <langstring language="en">Terminology</langstring>
  </vocabName>
  <term>
    <termIdentifier>field_work_terms</termIdentifier>
    <caption>
      <langstring language="en">Field work terms</langstring>
    </caption>
    <description>
      <langstring language="en"></langstring>
    </description>
    <term>
      <termIdentifier>field_work_terms.1</termIdentifier>
      <caption>
        <langstring language="en">Acidification</langstring>
      </caption>
      <description>
        <langstring language="en">Acidification is a process. It happens naturally ...</langstring>
      </description>
    </term>
    <term>
      <termIdentifier>field_work_terms.2</termIdentifier>
      <caption>
        <langstring language="en">Aquifer</langstring>
      </caption>
      <description>
        <langstring language="en">If you get a shovel and dig at the ground below your ...</langstring>
      </description>
    </term>
    <term>
      <termIdentifier>field_work_terms.3</termIdentifier>
      <caption>
        <langstring language="en">Biodiversity</langstring>
      </caption>
      <description>
        <langstring language="en">This has many contentious meanings but for our ...</langstring>
      </description>
    </term>
  </term>
</vdex>

Help Text

  usage: csv2vdex [-h] [--languages [LANGUAGES]] [--startrow [STARTROW]]
      [--description [DESCRIPTION]] [--keycolumn [KEYCOLUMN]]
      [--startcolumn [STARTCOLUMN]]
                  [--ordered [ORDERED]] [--dialect [DIALECT]]
                  [--csvdelimiter [CSVDELIMITER]]
                  [--treedelimiter [TREEDELIMITER]] [--encoding [ENCODING]]
                  id name source target
  csv2vdex: error: too few arguments
  jensens@minime:~/workspace/vdexcsv$ ./bin/csv2vdex --help
  usage: csv2vdex [-h] [--languages [LANGUAGES]] [--startrow [STARTROW]]
      [--description [DESCRIPTION]] [--keycolumn [KEYCOLUMN]]
      [--startcolumn [STARTCOLUMN]]
                  [--ordered [ORDERED]] [--dialect [DIALECT]]
                  [--csvdelimiter [CSVDELIMITER]]
                  [--treedelimiter [TREEDELIMITER]] [--encoding [ENCODING]]
                  id name source target

  Converts CSV files to VDEX XML

  positional arguments:
    id                    unique identifier of vocabulary
    name                  Human readable name of vocabulary. If more than one
                          language is given separate each langstring by a comma
                          and provide same order as argument --languages
    source                CSV file to read from
    target                XML target file

  optional arguments:
    -h, --help            show this help message and exit
    --languages [LANGUAGES], -l [LANGUAGES]
                          Comma separated list of ISO-language codes. Default:
                          en
--description
          Whether the terms have descriptions. If so, each term takes
          up two columns per language: one for the caption and one for
          the description.
    --startrow [STARTROW], -r [STARTROW]
                          number of row in CSV file where to begin reading,
                          starts with 0, default 0.
    --keycolumn [KEYCOLUMN], -k [KEYCOLUMN]
                          number of column with the keys of the vocabulary,
                          start with 0, default 0.
    --startcolumn [STARTCOLUMN], -s [STARTCOLUMN]
                          number of column with the first langstring of the
                          vocabulary. It assumes n + number languages of columns
                          after this, starts counting with 0, default 1.
          If terms include description, it assumes two columns
          per language.
    --ordered [ORDERED], -o [ORDERED]
                          Whether vocabulary is ordered or not, Default: True
    --dialect [DIALECT]   CSV dialect, default excel.
    --csvdelimiter [CSVDELIMITER]
                          CSV delimiter of the source file, default semicolon.
    --treedelimiter [TREEDELIMITER]
                          Delimiter used to split the key the vocabulary into a
                          path to determine the position in the tree, default
                          dot.
    --encoding [ENCODING], -e [ENCODING]
                          Encoding of input file. Default: utf-8

Source Code

The sources are in a GIT DVCS with its main branches at github.

We'd be happy to see many forks and pull-requests to make vdexcsv even better.

Contributors

• Jens W. Klein <jens@bluedynamics.com>
• Peter Holzer <hpeter@agitator.com>
• Jean Jordaan <jean.jordaan@gmail.com>