Bug #2869

Mis à jour par Serge Heiden il y a plus de 4 ans

TIGERSearch fails to open corpus configuration (corpus_config.xml file ) which path contains accents (e.g. "Télécharger").

h3. Solution 0

Display an error message if the path contains accents and abort the command.

Sample code to detect accents:

<pre>
import com.ibm.icu.text.Transliterator

// remove accents from characters ICU Transform, see http://userguide.icu-project.org/transforms/general
removeAccentsTransform = "NFD; [:M:] Remove; NFC"
path_no_accents = Transliterator.getInstance(removeAccentsTransform).transform(path)

if (!(path == path_no_accents)) then accents
</pre>

h3. Solution 1

Change the way XML files are opened&read in TIGERSearch core libraries and update the libraries of the TIGERSearch TXM extension

Retour