summaryrefslogtreecommitdiff
path: root/python/tests/extractor/run.sh
diff options
context:
space:
mode:
authorMichael Denkowski <michael.j.denkowski@gmail.com>2013-01-07 23:11:17 -0500
committerMichael Denkowski <michael.j.denkowski@gmail.com>2013-01-07 23:11:17 -0500
commitf10edb43ceaca8284282707097aaaebb6c1aedde (patch)
tree617f84fe832c288fbf25c9e37d26946140fea31d /python/tests/extractor/run.sh
parent88efef0b3f24e6defd37fac207d6761bffa4678f (diff)
Rule extraction unit test
Diffstat (limited to 'python/tests/extractor/run.sh')
-rwxr-xr-xpython/tests/extractor/run.sh14
1 files changed, 14 insertions, 0 deletions
diff --git a/python/tests/extractor/run.sh b/python/tests/extractor/run.sh
new file mode 100755
index 00000000..f44da9f8
--- /dev/null
+++ b/python/tests/extractor/run.sh
@@ -0,0 +1,14 @@
+#!/usr/bin/env bash
+
+# Make sure that the sa and online extractors are producing the same (correct) output
+
+set -x verbose
+
+python -m cdec.sa.compile -a corpus.al.gz -b corpus.fr-en.gz -o extract >| extract.ini
+
+cat test.in | python -m cdec.sa.extract -c extract.ini -g gold -o 2>&1 | egrep '\[X\].+\|\|\|.+\|\|\|.+\|\|\|.+\|\|\|'|sed -re 's/INFO.+://g' | ./refmt.py | LC_ALL=C sort >| rules.sort
+
+cd gold && cat grammar.0|sed -re 's/Egiv.+(IsSingletonF=)/\1/g'|LC_ALL=C sort >| rules.sort && cd ..
+
+diff gold/rules.sort gold-rules.sort
+diff rules.sort gold-rules.sort