diff options
| -rw-r--r-- | README.md | 7 | ||||
| -rw-r--r-- | example/3/cdec.ini | 5 | ||||
| -rw-r--r-- | example/3/in (renamed from example/3/3.in) | 0 | ||||
| -rw-r--r-- | example/toy/cdec.ini | 2 | ||||
| -rw-r--r-- | example/toy/grammar-test | 24 | ||||
| -rw-r--r-- | example/toy/toy-cdec.json (renamed from example/toy/toy1.json) | 0 | ||||
| -rwxr-xr-x | util/cdec2json.py (renamed from util/cdec_hg_to_json.py) | 2 |
7 files changed, 20 insertions, 20 deletions
@@ -1,11 +1,12 @@ -nothing to see here +not much to see here, yet (SCFG machine translation decoder in ruby, currently implements CKY+ parsing and hypergraph viterbi) helpful stuff * https://github.com/jweese/thrax/wiki/Glue-grammar + * http://aclweb.org/aclwiki/index.php?title=Hypergraph_Format todo - * animate? - * to json +==== + * animate parsing? * integrate with HG diff --git a/example/3/cdec.ini b/example/3/cdec.ini index 23c6ad3..ee65b4e 100644 --- a/example/3/cdec.ini +++ b/example/3/cdec.ini @@ -1,6 +1,5 @@ formalism=scfg -grammar=grammars/grammar.3.gz intersection_strategy=full -#grammar=grammar-test +grammar=grammars/grammar.3.gz #add_pass_through_rules=true -#weights=weights + diff --git a/example/3/3.in b/example/3/in index ee3509c..ee3509c 100644 --- a/example/3/3.in +++ b/example/3/in diff --git a/example/toy/cdec.ini b/example/toy/cdec.ini index 8276d9b..f12de54 100644 --- a/example/toy/cdec.ini +++ b/example/toy/cdec.ini @@ -1,5 +1,5 @@ formalism=scfg +intersection_strategy=full grammar=grammar #grammar=grammar-test #add_pass_through_rules=true -#weights=weights diff --git a/example/toy/grammar-test b/example/toy/grammar-test index 4027c71..18900ae 100644 --- a/example/toy/grammar-test +++ b/example/toy/grammar-test @@ -1,22 +1,22 @@ -[S] ||| [B,1] ||| [1] ||| logp=0 -[S] ||| ich [V,1] ein [JJ,2] haus ||| i [1] a [2] ||| logp=0 +[S] ||| [B,1] ||| [B,1] ||| logp=0 +[S] ||| ich [V,1] ein [JJ,2] haus ||| i [V,1] a [JJ,2] ||| logp=0 [S] ||| ich sah ein kleines haus ||| i saw a small house ||| logp=0 -[S] ||| ich sah ein [JJ,1] ||| i saw a [1] ||| logp=0 -[B] ||| [C,1] ||| [1] ||| logp=0 -[C] ||| [Q,1] ||| [1] ||| logp=0 -[Q] ||| [NP,1] [VP,2] ||| [1] [2] ||| logp=0 +[S] ||| ich sah ein [JJ,1] ||| i saw a [JJ,1] ||| logp=0 +[B] ||| [C,1] ||| [C,1] ||| logp=0 +[C] ||| [Q,1] ||| [Q,1] ||| logp=0 +[Q] ||| [NP,1] [VP,2] ||| [NP,1] [VP,2] ||| logp=0 [NP] ||| ich ||| i ||| logp=-0.5 use_i=1.0 -[NP] ||| ein [NN,1] ||| a [1] ||| logp=0 use_a=1.0 -[NN] ||| [JJ,1] haus ||| [1] house ||| logp=0 use_house=1 -[NN] ||| [JJ,1] haus ||| [1] shell ||| logp=0 use_shell=1 -[JJ] ||| [H,1] ||| [1] ||| logp=0 +[NP] ||| ein [NN,1] ||| a [NN,1] ||| logp=0 use_a=1.0 +[NN] ||| [JJ,1] haus ||| [JJ,1] house ||| logp=0 use_house=1 +[NN] ||| [JJ,1] haus ||| [JJ,1] shell ||| logp=0 use_shell=1 +[JJ] ||| [H,1] ||| [H,1] ||| logp=0 [H] ||| kleines [Z,1] ||| small [1] ||| logp=0 -[Z] ||| [I,1] ||| [1] ||| logp=0 +[Z] ||| [I,1] ||| [I,1] ||| logp=0 [I] ||| haus ||| house ||| logp=0 [JJ] ||| kleines ||| small ||| logp=0 use_small=1.0 [JJ] ||| kleines ||| little ||| logp=0 use_little=1.0 [JJ] ||| grosses ||| big ||| logp=0 [JJ] ||| grosses ||| large ||| logp=0 -[VP] ||| [V,1] [NP,2] ||| [1] [2] ||| logp=0 +[VP] ||| [V,1] [NP,2] ||| [V,1] [NP,2] ||| logp=0 [V] ||| sah ||| saw ||| logp=-0.25 use_saw=1.0 [V] ||| fand ||| found ||| logp=0 diff --git a/example/toy/toy1.json b/example/toy/toy-cdec.json index c58c249..c58c249 100644 --- a/example/toy/toy1.json +++ b/example/toy/toy-cdec.json diff --git a/util/cdec_hg_to_json.py b/util/cdec2json.py index 2fcc409..76e2cd4 100755 --- a/util/cdec_hg_to_json.py +++ b/util/cdec2json.py @@ -3,10 +3,10 @@ import cdec import sys, argparse + def hg2json(hg, weights): """ output a JSON representation of a cdec hypegraph - (see http://aclweb.org/aclwiki/index.php?title=Hypergraph_Format ) """ res = '' res += "{\n" |