Fix command line

README.md

@@ -19,12 +19,12 @@ export PATH="${PATH}:$(pwd)/tools/bin"
 
 テキストを直接与えて実行する場合
 ```bash
-echo 'ごはんを食べるつもりです。' | python -m trf.cmdline
+echo 'ごはんを食べるつもりです。' | python -m trf
 ```
 
 テキストファイルを指定して実行する場合
 ```bash
-python -m trf.cmdline -f FILENAME
+python -m trf -f FILENAME
 ```
 
 ## TRFがサポートする指標一覧

setup.py

@@ -14,6 +14,5 @@
           "numpy",
           "janome",
           "sqlalchemy"
-      ],
-      entry_points={"console_scripts": ["trf = trf.cmdline:main"]})
+      ])
 #      package_data={'', ['']},

trf/__init__.py

@@ -0,0 +1,156 @@
+import sys
+import argparse
+from typing import List
+
+from trf.analyser import Analyser
+from trf.acceptability import Acceptability
+from trf.util import check_executable
+
+
+def translate(en: str):
+
+    if en == 'n_sentences':
+        return '文数'
+    elif en == 'mean_n_mrphs':
+        return '平均文長'
+    elif en == 'n_tokens':
+        return 'トークン数'
+    elif en == 'n_types':
+        return 'タイプ数'
+    elif en == 'mean_tree_depths':
+        return '係り受け木の深さ'
+    elif en == 'r_conditional':
+        return '仮定節'
+    elif en == 'log_prob':
+        return '容認度 (LogProb)'
+    elif en == 'mean_lp':
+        return '容認度 (Mean LP)'
+    elif en == 'norm_lp_div':
+        return '容認度 (Norm LP (Div))'
+    elif en == 'norm_lp_sub':
+        return '容認度 (Norm LP (Sub))'
+    elif en == 'slor':
+        return '容認度 (SLOR)'
+    else:
+        return en
+
+
+class Metric:
+    def __init__(self, name: str, val: str):
+
+        self.name = name
+        self.val = val
+        self.name_ja = translate(name)
+
+    def __str__(self):
+        return '\t'.join([self.name_ja, self.val])
+
+
+class Section:
+
+    def __init__(self, cat: str, metrics: List[Metric]):
+
+        self.cat = cat
+        self.metrics = metrics
+        if cat == 'basic':
+            self.cat_ja = '基本指標'
+        elif cat == 'vocabulary':
+            self.cat_ja = '語彙に基づく指標'
+        elif cat == 'syntax':
+            self.cat_ja = '統語情報に基づく指標'
+        elif cat == 'language_model':
+            self.cat_ja = '言語モデルに基づく指標'
+        else:
+            self.cat_ja = ''
+
+    def show(self, lang: str='ja'):
+        if lang == 'ja':
+            print('[{}]'.format(self.cat_ja))
+            for metric in self.metrics:
+                print('{}={}'.format(metric.name_ja, metric.val))
+        else:
+            print('Unsupported language')
+            sys.exit(1)
+
+
+def _f(score: float) -> str:
+    return 'None' if score is None else '{:.2f}'.format(score)
+
+
+def main():
+
+    executables = ['juman', 'knp', 'rnnlm']
+    for e in executables:
+        check_executable(e)
+
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument('-f',
+                        '--filename',
+                        type=str,
+                        help='target text')
+
+    parser.add_argument('--delimiter',
+                        type=str,
+                        default='\n',
+                        help='features to calculate')
+
+    parser.add_argument('-m',
+                        '--rnnlm-model-path',
+                        type=str,
+                        default='data/jawiki-20160818-100M-words',
+                        help='RNNLM model path')
+
+    parser.add_argument('--output-lang',
+                        type=str,
+                        default='ja',
+                        help='ja')
+
+    args = parser.parse_args()
+
+    text = ''
+    if args.filename is not None:
+        with open(args.filename, mode='r') as f:
+            text = f.read().replace('\n', '')
+    else:
+        text = sys.stdin.read()
+
+    analyser = Analyser(text, delimiter=args.delimiter)
+
+    metrics = []
+    metrics.append(Metric('n_sentences', analyser.n_sentences))
+    metrics.append(Metric('mean_n_mrphs', analyser.mean_n_mrphs))
+    metrics.append(Metric('n_tokens', analyser.n_chunks))
+    metrics.append(Metric('n_types', analyser.n_types))
+    Section('basic', metrics).show()
+
+    metrics = []
+    for k, v in analyser.rs_pos.items():
+        metrics.append(Metric('品詞：{}'.format(k),
+                              '{:.2f}'.format(v)))
+    Section('vocabulary', metrics).show()
+
+    metrics = []
+    metrics.append(Metric('mean_tree_depths',
+                          '{:.2f}'.format(analyser.mean_tree_depths)))
+    metrics.append(Metric('r_conditional',
+                          '{:.2f}'.format(analyser.r_conditional)))
+    for k, v in analyser.rs_modality.items():
+        metrics.append(Metric('モダリティ：{}'.format(k),
+                              '{:.2f}'.format(v)))
+    Section('syntax', metrics).show()
+
+    metrics = []
+    a = Acceptability(text,
+                      args.delimiter,
+                      args.rnnlm_model_path)
+    metrics.append(Metric('log_prob', _f(a.log_prob)))
+    metrics.append(Metric('mean_lp', _f(a.mean_lp)))
+    metrics.append(Metric('norm_lp_div', _f(a.norm_lp_div)))
+    metrics.append(Metric('norm_lp_sub', _f(a.norm_lp_sub)))
+    metrics.append(Metric('slor', _f(a.slor)))
+    Section('language_model', metrics).show()
+
+
+if __name__ == '__main__':
+    main()

trf/__main__.py

@@ -0,0 +1,4 @@
+import sys
+from trf import main
+
+sys.exit(main())