Feature #1597

Updated by Alexey Lavrentev over 2 years ago

XML milestones are not implemented in the structural attributes of the CQP corpus data model.

Some corpora need to be queried directly by some milestones related encodings because it is not pertinent to transform the information in usual structural attributes.

h3. Solution

A first step, to just allow using the position information of milestones with respect to words, is to encode them in word properties. Like for example the distance of the word to a previous and to a next specific milestone.

4 word properties are created for each milestone specified in the import form:

* <milestone>start : distance in tokens to the 1st preceding milestone (implemented)
* <milestone>end : distance in tokens to the 1st following milestone (implemented)
* <milestone>id : identifier (@*:id) of the 1st preceding milestone (implemented)
* <milestone>n : number (@n) of the 1st preceding milestone (not implemented)