macports/macports-ports
0
0
Fork 0
mirror of https://github.com/macports/macports-ports.git synced 2026-07-12 18:20:25 -07:00
Code Issues Packages Projects Releases Wiki Activity
Files
python_variants
macports-ports/textproc/cicada/files/patch-samples.diff
T

565 lines
18 KiB
Diff
Raw Permalink Blame History

--- samples/kftt.30k/alignment/alignment.sh.orig 2013-07-30 14:14:40.000000000 +0900
+++ samples/kftt.30k/alignment/alignment.sh 2013-10-27 18:22:32.000000000 +0900
@@ -1,11 +1,11 @@
#!/bin/sh
-cicada=../../..
+bindir=@PREFIX@/bin
## We perform symmetized posterior constrained training, and
## perform smoothing by naive Bayes.
-exec ${cicada}/scripts/cicada-alignment.py \
+exec ${bindir}/cicada-alignment.py \
--f ../data/train.ja.bz2 \
--e ../data/train.en.bz2 \
--symmetric \
--- samples/kftt.30k/ngram/expgram.sh.orig 2013-07-30 14:14:34.000000000 +0900
+++ samples/kftt.30k/ngram/expgram.sh 2013-10-27 18:23:31.000000000 +0900
@@ -1,17 +1,12 @@
#!/bin/sh
-expgram=
-
-if test "$expgram" = ""; then
- echo "where is your expgram?"
- exit 1
-fi
+bindir=@PREFIX@/bin
### Following is a quick example of LM estimation on a small data set.
-$expgram/progs/expgram_counts_extract --corpus ../data/train.en.bz2 --output ngram.5.en.counts --order 5 --threads 4
+${bindir}/expgram_counts_extract --corpus ../data/train.en.bz2 --output ngram.5.en.counts --order 5 --threads 4
-$expgram/progs/expgram_counts_estimate --ngram ngram.5.en.counts --output ngram.5.en.lm --shard 4
+${bindir}/expgram_counts_estimate --ngram ngram.5.en.counts --output ngram.5.en.lm --shard 4
### A standard way is:
-# $expgram/scripts/expgram.py --corpus ../data/train.en --output ngram.5.en --threads 4
+# ${bindir}/expgram.py --corpus ../data/train.en --output ngram.5.en --threads 4
--- samples/kftt.30k/s2t/data/preprocess.sh.orig 2013-08-02 10:22:41.000000000 +0900
+++ samples/kftt.30k/s2t/data/preprocess.sh 2013-10-27 19:52:56.000000000 +0900
@@ -1,12 +1,8 @@
#!/bin/sh
-stanford=
-cicada=../../../..
-
-if test "$stanford" = ""; then
- echo "where is your stanford parser?"
- exit 1
-fi
+stanford=@PREFIX@/share/java/stanford-parser
+version=3.2.0
+bindir=@PREFIX@/bin
# Here, we use stanford-parser to parse training data in English
# cicada_filter_penntreebank to transform into hypergraph.
@@ -18,7 +14,7 @@
bzcat ../../data/train.en.bz2 | \
java \
-mx12g \
- -cp $stanford/stanford-parser.jar:$stanford/stanford-parser-3.2.0-models.jar \
+ -cp $stanford/stanford-parser.jar:$stanford/stanford-parser-${version}-models.jar \
-tLPP edu.stanford.nlp.parser.lexparser.EnglishTreebankParserParams \
-tokenized -sentences newline \
-escaper edu.stanford.nlp.process.PTBEscapingProcessor \
@@ -29,10 +25,10 @@
-outputFormatOptions includePunctuationDependencies \
edu/stanford/nlp/models/lexparser/englishFactored.ser.gz \
- | \
-$cicada/progs/cicada_filter_penntreebank \
+${bindir}/cicada_filter_penntreebank \
--map ../../data/train.en.bz2 \
--normalize | \
-$cicada/progs/cicada \
+${bindir}/cicada \
--input-forest \
--threads 8 \
--operation binarize:direction=left,order=2 \
--- samples/kftt.30k/s2t/model/extract.sh.orig 2013-07-31 09:19:40.000000000 +0900
+++ samples/kftt.30k/s2t/model/extract.sh 2013-10-27 19:53:42.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-extract.py \
+exec $bindir/cicada-extract.py \
--f ../../data/train.ja.bz2 \
--e ../../data/train.en.bz2 \
--a ../../alignment/model/aligned.posterior-itg \
--- samples/kftt.30k/s2t/model/index.sh.orig 2013-07-31 09:19:37.000000000 +0900
+++ samples/kftt.30k/s2t/model/index.sh 2013-10-27 19:54:07.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-index.py \
+exec $bindir/cicada-index.py \
--model-dir . \
\
--ghkm \
--- samples/kftt.30k/s2t/test/test.sh.orig 2013-08-02 10:23:48.000000000 +0900
+++ samples/kftt.30k/s2t/test/test.sh 2013-10-27 19:55:42.000000000 +0900
@@ -1,16 +1,11 @@
#!/bin/sh
-cicada=../../../..
-weights=
-
-if test "$weights" = ""; then
- echo "where is your weights? (for instance ../tune/learn.10.weights)"
- exit 1
-fi
+bindir=@PREFIX@/bin
+weights=../tune/learn.10.weights
### generate config file
-$cicada/progs/cicada_filter_config \
+$bindir/cicada_filter_config \
--input ../tune/cicada.config \
--output cicada.config \
--weights "weights=$weights" \
@@ -18,9 +13,9 @@
--file "file=-"
### perform translation
-$cicada/progs/cicada --config cicada.config --threads 4 --debug < ../../data/dev.ja > dev.ja-en
+$bindir/cicada --config cicada.config --threads 4 --debug < ../../data/dev.ja > dev.ja-en
### evaluation
-$cicada/progs/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
+$bindir/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
--- samples/kftt.30k/s2t/tune/config.sh.orig 2013-08-02 10:24:29.000000000 +0900
+++ samples/kftt.30k/s2t/tune/config.sh 2013-10-27 19:56:16.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-config.py \
+exec $bindir/cicada-config.py \
--tree-grammar ../model/ghkm-index \
--max-span 10 \
--goal '[ROOT]' \
--- samples/kftt.30k/s2t/tune/tune.sh.orig 2013-09-19 17:38:55.000000000 +0900
+++ samples/kftt.30k/s2t/tune/tune.sh 2013-10-27 19:56:32.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../../
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-learn.py \
+exec $bindir/cicada-learn.py \
--srcset ../../data/tune.ja \
--refset ../../data/tune.en.ref \
--config cicada.config \
--- samples/kftt.30k/scfg/model/extract.sh.orig 2013-07-30 14:14:16.000000000 +0900
+++ samples/kftt.30k/scfg/model/extract.sh 2013-10-27 19:57:17.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-extract.py \
+exec $bindir/cicada-extract.py \
--f ../../data/train.ja.bz2 \
--e ../../data/train.en.bz2 \
--a ../../alignment/model/aligned.posterior-itg \
--- samples/kftt.30k/scfg/model/index.sh.orig 2013-07-30 14:14:21.000000000 +0900
+++ samples/kftt.30k/scfg/model/index.sh 2013-10-27 19:57:11.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-index.py \
+exec $bindir/cicada-index.py \
--model-dir . \
\
--scfg \
--- samples/kftt.30k/scfg/test/test.sh.orig 2013-07-30 13:00:31.000000000 +0900
+++ samples/kftt.30k/scfg/test/test.sh 2013-10-27 19:57:55.000000000 +0900
@@ -1,16 +1,11 @@
#!/bin/sh
-cicada=../../../..
-weights=
-
-if test "$weights" = ""; then
- echo "where is your weights? (for instance ../tune/learn.10.weights)"
- exit 1
-fi
+bindir=@PREFIX@/bin
+weights=../tune/learn.10.weights
### generate config file
-$cicada/progs/cicada_filter_config \
+$bindir/cicada_filter_config \
--input ../tune/cicada.config \
--output cicada.config \
--weights "weights=$weights" \
@@ -18,9 +13,9 @@
--file "file=-"
### perform translation
-$cicada/progs/cicada --config cicada.config --threads 4 --debug < ../../data/dev.ja > dev.ja-en
+$bindir/cicada --config cicada.config --threads 4 --debug < ../../data/dev.ja > dev.ja-en
### evaluation
-$cicada/progs/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
+$bindir/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
--- samples/kftt.30k/scfg/tune/config.sh.orig 2013-07-29 13:01:44.000000000 +0900
+++ samples/kftt.30k/scfg/tune/config.sh 2013-10-27 19:58:21.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-config.py \
+exec $bindir/cicada-config.py \
--grammar ../model/scfg-index \
--max-span 15 \
--straight \
--- samples/kftt.30k/scfg/tune/tune.sh.orig 2013-09-19 17:38:30.000000000 +0900
+++ samples/kftt.30k/scfg/tune/tune.sh 2013-10-27 19:58:32.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-learn.py \
+exec $bindir/cicada-learn.py \
--srcset ../../data/tune.ja \
--refset ../../data/tune.en.ref \
--config cicada.config \
--- samples/kftt.30k/t2s/data/preprocess.sh.orig 2013-07-30 15:15:17.000000000 +0900
+++ samples/kftt.30k/t2s/data/preprocess.sh 2013-10-27 20:00:35.000000000 +0900
@@ -4,20 +4,9 @@
# Note that the Japanese side was segmented by mecab (https://code.google.com/p/mecab/).
# Thus, it is a natural choice to use cabocha (https://code.google.com/p/cabocha/).
-cicada=../../../..
-mecab=
-cabocha=
-
-if test "$mecab" = ""; then
- echo "where is your mecab?"
- exit 1
-fi
-
-if test "$cabocha" = ""; then
- echo "where is your cabocha?"
- exit 1
-fi
-
+bindir=@PREFIX@/bin
+mecab=$bindir/mecab
+cabocha=$bindir/cabocha
for data in train dev tune; do
@@ -34,12 +23,12 @@
awk '{for (i=1;i<=NF;++i) {printf "%s\t*\n", $i } print "EOS";}' | \
$mecab -p | \
$cabocha -f1 -I 1 | \
- $cicada/progs/cicada_filter_dependency \
+ $bindir/cicada_filter_dependency \
--cabocha \
--func \
--forest \
--head | \
- $cicada/progs/cicada \
+ $bindir/cicada \
--input-forest \
--threads 4 \
--operation binarize:direction=cyk,order=1 \
--- samples/kftt.30k/t2s/model/extract.sh.orig 2013-07-30 15:26:16.000000000 +0900
+++ samples/kftt.30k/t2s/model/extract.sh 2013-10-27 20:01:00.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-extract.py \
+exec $bindir/cicada-extract.py \
--f ../../data/train.ja.bz2 \
--e ../../data/train.en.bz2 \
--a ../../alignment/model/aligned.posterior-itg \
--- samples/kftt.30k/t2s/model/index.sh.orig 2013-07-30 14:14:12.000000000 +0900
+++ samples/kftt.30k/t2s/model/index.sh 2013-10-27 20:01:08.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-index.py \
+exec $bindir/cicada-index.py \
--root-dir . \
--model-dir . \
\
--- samples/kftt.30k/t2s/test/test.sh.orig 2013-07-30 13:00:41.000000000 +0900
+++ samples/kftt.30k/t2s/test/test.sh 2013-10-27 20:01:41.000000000 +0900
@@ -1,16 +1,11 @@
#!/bin/sh
-cicada=../../../..
-weights=
-
-if test "$weights" = ""; then
- echo "where is your weights? (for instance ../tune/learn.10.weights)"
- exit 1
-fi
+bindir=@PREFIX@/bin
+weights=../tune/learn.10.weights
### generate config file
-$cicada/progs/cicada_filter_config \
+$bindir/cicada_filter_config \
--input ../tune/cicada.config \
--output cicada.config \
--weights "weights=$weights" \
@@ -18,9 +13,9 @@
--file "file=-"
### perform translation
-$cicada/progs/cicada --config cicada.config --threads 4 --debug --input ../data/dev.forest.ja.gz > dev.ja-en
+$bindir/cicada --config cicada.config --threads 4 --debug --input ../data/dev.forest.ja.gz > dev.ja-en
### evaluation
-$cicada/progs/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
+$bindir/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
--- samples/kftt.30k/t2s/tune/config.sh.orig 2013-07-29 13:52:34.000000000 +0900
+++ samples/kftt.30k/t2s/tune/config.sh 2013-10-27 20:02:02.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-config.py \
+exec $bindir/cicada-config.py \
--tree-grammar ../model/ghkm-index \
--goal '[x]' \
--glue '[x]' \
--- samples/kftt.30k/t2s/tune/tune.sh.orig 2013-09-19 17:38:49.000000000 +0900
+++ samples/kftt.30k/t2s/tune/tune.sh 2013-10-27 20:02:19.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-learn.py \
+exec $bindir/cicada-learn.py \
--srcset ../data/tune.forest.ja.gz \
--refset ../../data/tune.en.ref \
--config cicada.config \
--- samples/kftt.30k/t2t/model/extract.sh.orig 2013-08-01 08:37:16.000000000 +0900
+++ samples/kftt.30k/t2t/model/extract.sh 2013-10-27 20:02:40.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-extract.py \
+exec $bindir/cicada-extract.py \
--f ../../data/train.ja \
--e ../../data/train.en \
--a ../../alignment/model/aligned.posterior-itg \
--- samples/kftt.30k/t2t/model/index.sh.orig 2013-07-31 09:36:12.000000000 +0900
+++ samples/kftt.30k/t2t/model/index.sh 2013-10-27 20:02:49.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-index.py \
+exec $bindir/cicada-index.py \
--model-dir . \
\
--tree \
--- samples/kftt.30k/t2t/test/test.sh.orig 2013-07-31 09:35:38.000000000 +0900
+++ samples/kftt.30k/t2t/test/test.sh 2013-10-27 20:03:19.000000000 +0900
@@ -1,16 +1,11 @@
#!/bin/sh
-cicada=../../../..
-weights=
-
-if test "$weights" = ""; then
- echo "where is your weights? (for instance ../tune/learn.10.weights)"
- exit 1
-fi
+bindir=@PREFIX@/bin
+weights=../tune/learn.10.weights
### generate config file
-$cicada/progs/cicada_filter_config \
+$bindir/cicada_filter_config \
--input ../tune/cicada.config \
--output cicada.config \
--weights "weights=$weights" \
@@ -18,9 +13,9 @@
--file "file=-"
### perform translation
-$cicada/progs/cicada --config cicada.config --threads 4 --debug --input ../../t2s/data/dev.forest.ja.gz > dev.ja-en
+$bindir/cicada --config cicada.config --threads 4 --debug --input ../../t2s/data/dev.forest.ja.gz > dev.ja-en
### evaluation
-$cicada/progs/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
+$bindir/cicada_eval --tstset dev.ja-en --refset ../../data/dev.en.ref
--- samples/kftt.30k/t2t/tune/config.sh.orig 2013-07-31 09:52:14.000000000 +0900
+++ samples/kftt.30k/t2t/tune/config.sh 2013-10-27 20:03:41.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-config.py \
+exec $bindir/cicada-config.py \
--tree-grammar ../model/tree-index \
--goal '[ROOT]' \
--glue '[x]' \
--- samples/kftt.30k/t2t/tune/tune.sh.orig 2013-09-19 17:38:41.000000000 +0900
+++ samples/kftt.30k/t2t/tune/tune.sh 2013-10-27 20:03:55.000000000 +0900
@@ -1,8 +1,8 @@
#!/bin/sh
-cicada=../../../..
+bindir=@PREFIX@/bin
-exec $cicada/scripts/cicada-learn.py \
+exec $bindir/cicada-learn.py \
--srcset ../../t2s/data/tune.forest.ja.gz \
--refset ../../data/tune.en.ref \
--config cicada.config \
--- samples/s2t/sample.sh.orig 2013-08-03 10:37:31.000000000 +0900
+++ samples/s2t/sample.sh 2013-10-27 20:07:14.000000000 +0900
@@ -1,6 +1,7 @@
#!/bin/sh
-cicada=../..
+bindir=@PREFIX@/bin
+exdir=@PREFIX@/share/examples/cicada
##
##
@@ -18,16 +19,16 @@
## apply features
## output kbests
-exec $cicada/progs/cicada \
- --input $cicada/samples/scfg/input.txt \
+exec $bindir/cicada \
+ --input $exdir/scfg/input.txt \
--goal '[ROOT]' \
- --tree-grammar $cicada/samples/s2t/grammar.bin:max-span=10 \
+ --tree-grammar $exdir/s2t/grammar.bin:max-span=10 \
--tree-grammar "glue:goal-source=[ROOT],goal-target=[ROOT],non-terminal-source=[x],non-terminal-target=[x],straight=true,invert=false" \
--grammar "insertion:non-terminal=[x]" \
- --feature-function "ngram:file=$cicada/samples/scfg/ngram.bin" \
+ --feature-function "ngram:file=$exdir/scfg/ngram.bin" \
--feature-function word-penalty \
--feature-function rule-penalty \
--feature-function glue-tree-penalty \
--operation compose-tree-cky \
- --operation apply:prune=true,size=100,weights=$cicada/samples/s2t/weights \
- --operation output:file=-,kbest=10,weights=$cicada/samples/s2t/weights
+ --operation apply:prune=true,size=100,weights=$exdir/s2t/weights \
+ --operation output:file=-,kbest=10,weights=$exdir/s2t/weights
--- samples/scfg/sample.sh.orig 2013-07-30 14:00:13.000000000 +0900
+++ samples/scfg/sample.sh 2013-10-27 20:07:02.000000000 +0900
@@ -1,15 +1,16 @@
#!/bin/sh
-cicada=../..
+bindir=@PREFIX@/bin
+exdir=@PREFIX@/share/examples/cicada
-exec $cicada/progs/cicada \
- --input $cicada/samples/scfg/input.txt \
- --grammar $cicada/samples/scfg/grammar.bin \
+exec $bindir/cicada \
+ --input $exdir/scfg/input.txt \
+ --grammar $exdir/scfg/grammar.bin \
--grammar "glue:straight=true,inverted=false,non-terminal=[x],goal=[s]" \
--grammar "insertion:non-terminal=[x]" \
- --feature-function "ngram:file=$cicada/samples/scfg/ngram.bin" \
+ --feature-function "ngram:file=$exdir/scfg/ngram.bin" \
--feature-function word-penalty \
--feature-function rule-penalty \
--operation compose-cky \
- --operation apply:prune=true,size=100,weights=$cicada/samples/scfg/weights \
- --operation output:file=-,kbest=10,weights=$cicada/samples/scfg/weights
+ --operation apply:prune=true,size=100,weights=$exdir/scfg/weights \
+ --operation output:file=-,kbest=10,weights=$exdir/scfg/weights
--- samples/t2s/sample.sh.orig 2013-08-19 16:18:37.000000000 +0900
+++ samples/t2s/sample.sh 2013-10-27 20:08:28.000000000 +0900
@@ -1,6 +1,7 @@
#!/bin/sh
-cicada=../..
+bindir=@PREFIX@/bin
+exdir=@PREFIX@/share/examples/cicada
##
## 1. pentreebank is converted into hypergraph
@@ -20,20 +21,20 @@
## apply features
## output kbests
-$cicada/progs/cicada_filter_penntreebank \
- --input $cicada/samples/t2s/input.txt \
+$bindir/cicada_filter_penntreebank \
+ --input $exdir/t2s/input.txt \
--normalize \
| \
-$cicada/progs/cicada \
+$bindir/cicada \
--input - \
--input-forest \
--goal '[ROOT]' \
- --tree-grammar $cicada/samples/t2s/grammar.bin \
+ --tree-grammar $exdir/t2s/grammar.bin \
--tree-grammar fallback \
- --feature-function "ngram:file=$cicada/samples/scfg/ngram.bin" \
+ --feature-function "ngram:file=$exdir/scfg/ngram.bin" \
--feature-function word-penalty \
--feature-function rule-penalty \
--operation binarize:direction=cyk,order=1 \
--operation compose-tree \
- --operation apply:prune=true,size=1000,weights=$cicada/samples/t2s/weights \
- --operation output:file=-,kbest=10,weights=$cicada/samples/t2s/weights
+ --operation apply:prune=true,size=1000,weights=$exdir/t2s/weights \
+ --operation output:file=-,kbest=10,weights=$exdir/t2s/weights
Reference in New Issue View Git Blame Copy Permalink