Laboratoire ICAR » Plateforme TXM

Since the introduction of new clitic rules management in the tokenizer, some graphical forms are segmented.

For exemple in English language ('en'): I don't -> I do n't

The corresponding word properties are (form/pos/lemma): I/PP/I do/VVP/do n't/RB/n't

The XML/w page outputs the following surface (graphical forms): I do n't

The correct surface should be : I don't -> like in the source

Solution¶

Introduce clitic management in the typographic rules management in the Page renderer to prevent the space between "do" and "n't".

MD: rules added from clitics : "'s","'re","'ve","'d","'m","'em","'ll","n't"