fix: rw scripts

engisalor · Oct 11, 2024 · ba35b18 · ba35b18
1 parent 212f88a
commit ba35b18
Showing 1 changed file with 4 additions and 4 deletions.
diff --git a/rw_corpora_update.sh b/rw_corpora_update.sh
@@ -10,13 +10,13 @@ fi
 echo "... get and process source data"
 python3 rw_corpora_update.py "${1}" "${2}" || exit 1
 echo "... convert source files to .conllu (run Stanza NLP)"
-ES_FILES=$(find -name reliefweb_es*.txt)
-FR_FILES=$(find -name reliefweb_fr*.txt)
-EN_FILES=$(find -name reliefweb_en*.txt)
+ES_FILES=$(find -name "reliefweb_es*.txt")
+FR_FILES=$(find -name "reliefweb_fr*.txt")
+EN_FILES=$(find -name "reliefweb_en*.txt")
 python3 ./pipeline/stanza/base_pipeline.py to-conll -l es $ES_FILES || exit 1
 python3 ./pipeline/stanza/base_pipeline.py to-conll -l fr $FR_FILES || exit 1
 python3 ./pipeline/stanza/base_pipeline.py to-conll -l en $EN_FILES || exit 1
 echo "... convert files to .vert and compress"
-CONLLU_FILES=$(find -name reliefweb_*.txt.conllu)
+CONLLU_FILES=$(find -name "reliefweb_*.txt.conllu")
 python3 ./pipeline/stanza/base_pipeline.py conll-to-vert $CONLLU_FILES || exit 1
 echo "... rw_corpora_update.sh completed"