summaryrefslogtreecommitdiff
path: root/gi/pipeline
diff options
context:
space:
mode:
authorphilblunsom@gmail.com <philblunsom@gmail.com@ec762483-ff6d-05da-a07a-a48fb63a330f>2010-06-29 17:06:03 +0000
committerphilblunsom@gmail.com <philblunsom@gmail.com@ec762483-ff6d-05da-a07a-a48fb63a330f>2010-06-29 17:06:03 +0000
commitfa91c76321460773e0de7fc077db5a3c919eb89b (patch)
tree097ba6452ddc389283d1fb2956357dcb9576ea46 /gi/pipeline
parent90f4ede0eea7ec5b73e7167dde4fe590485dcd67 (diff)
Added different executable for working with context models, changed input/output processing.
git-svn-id: https://ws10smt.googlecode.com/svn/trunk@54 ec762483-ff6d-05da-a07a-a48fb63a330f
Diffstat (limited to 'gi/pipeline')
-rwxr-xr-xgi/pipeline/local-gi-pipeline.pl3
1 files changed, 2 insertions, 1 deletions
diff --git a/gi/pipeline/local-gi-pipeline.pl b/gi/pipeline/local-gi-pipeline.pl
index af83beb8..be91f9ad 100755
--- a/gi/pipeline/local-gi-pipeline.pl
+++ b/gi/pipeline/local-gi-pipeline.pl
@@ -138,7 +138,8 @@ sub label_spans_with_topics {
print STDERR "$OUT_SPANS exists, reusing...\n";
} else {
safesystem("$ZCAT $IN_CLUSTERS > $OUTPUT/clusters.txt") or die "Failed to unzip";
- safesystem("$EXTRACTOR --base_phrase_spans -i $CORPUS -c $ITEMS_IN_MEMORY -L $BASE_PHRASE_MAX_SIZE -S $CONTEXT_SIZE | $S2L $OUTPUT/phrases.index $OUTPUT/contexts.index $OUTPUT/clusters.txt > $OUT_SPANS") or die "Failed to label spans";
+# safesystem("$EXTRACTOR --base_phrase_spans -i $CORPUS -c $ITEMS_IN_MEMORY -L $BASE_PHRASE_MAX_SIZE -S $CONTEXT_SIZE | $S2L $OUTPUT/phrases.index $OUTPUT/contexts.index $OUTPUT/clusters.txt > $OUT_SPANS") or die "Failed to label spans";
+ safesystem("$EXTRACTOR --base_phrase_spans -i $CORPUS -c $ITEMS_IN_MEMORY -L $BASE_PHRASE_MAX_SIZE -S $CONTEXT_SIZE | $S2L $OUTPUT/clusters.txt $CONTEXT_SIZE > $OUT_SPANS") or die "Failed to label spans";
unlink("$OUTPUT/clusters.txt") or warn "Failed to remove $OUTPUT/clusters.txt";
safesystem("paste -d ' ' $CORPUS $OUT_SPANS > $OUTPUT/corpus.src_trg_al") or die "Couldn't paste";
}