From 9167792b078b25e14dff73000d185370fcc59b19 Mon Sep 17 00:00:00 2001
From: Patrick Nguyen <me@patrickphat.com>
Date: Wed, 2 Sep 2020 23:39:10 +0700
Subject: [PATCH] Remove recursive parser (#12)

* remove recursive parser
* fix :bug:
---
 setup.py                          |  2 +-
 urbamt/translator.py              | 35 ++++++++++++++++++++-----------
 urbamt/utils/tree_manipulation.py |  1 -
 3 files changed, 24 insertions(+), 14 deletions(-)

diff --git a/setup.py b/setup.py
index bf4e40a..41481d9 100644
--- a/setup.py
+++ b/setup.py
@@ -5,7 +5,7 @@
 
 setuptools.setup(
      name='urbamt',  
-     version='0.0.1-b1',
+     version='0.0.1-b2',
      author="Patrick Phat Nguyen",
      author_email="me@patrickphat.com",
      description="Universal Rule-based Machine Translation Toolkit (URBaMT)",
diff --git a/urbamt/translator.py b/urbamt/translator.py
index 5320baa..3ba9186 100644
--- a/urbamt/translator.py
+++ b/urbamt/translator.py
@@ -2,7 +2,7 @@
 from .utils.tree_manipulation import translate_tree_grammar
 from .utils.misc import remove_trailing_space
 import nltk 
-from nltk import RecursiveDescentParser as Parser
+from nltk.parse.chart import BottomUpLeftCornerChartParser as Parser
 
 class URBAMT_Translator:
     """"""
@@ -11,7 +11,6 @@ def __init__(self,
                  src_to_tgt_grammar: Dict, 
                  src_to_tgt_dictionary: Dict):
         """Initialize the translator
-
         Args:
             src_grammar (str): source language grammar written in nltk style
             E.g: src_grammar = \"""
@@ -28,14 +27,10 @@ def __init__(self,
                                 JJ -> 'good'
                                 NN -> 'school'
                                \"""
-
             src_to_tgt_grammar (Dict): Transition between source grammar and target grammar as a dict
             E.g: src2target_grammar =  {
                                     "NP1 -> JJ NN": "NP1 -> NN JJ" 
                                         }
-
-
-
             src_to_tgt_dictionary (Dict): Dictionary of word-by-word transition from src language to target language
             E.g: en_to_vi_dict = {
                                 "I":"tôi",
@@ -54,24 +49,28 @@ def __init__(self,
     def __process_text_input(txt):
         return remove_trailing_space(txt)
 
-    def translate(self, sentences: List[str] or str):
+    def translate(self, sentences: List[str] or str, allow_multiple_translation = False):
         """Translate a list of sentences
-
         Args:
             sentences (List[str]): A list of str-typed sentences
-
         Returns:
             List[str]: A list of translated sentences
         """
         if isinstance(sentences,str):
             sentences = [sentences]
 
-        translated_sentence = []
+        translated_sentences = []
+        failed_sentences = []
+        
         for sentence in sentences:
             sentence = self.__process_text_input(sentence)
             trees = self.parser.parse(sentence.split())
 
+            # Flag to check if there are trees in generator (grammar matched)
+            translated = False
+
             for t in trees:
+                translated = True
 
                 # Translate grammar
                 trans_gram_sentence = translate_tree_grammar(t,self.src_to_tgt_grammar)
@@ -79,6 +78,18 @@ def translate(self, sentences: List[str] or str):
                 # Translate words
                 trans_lang_sentence = ' '.join([self.src_to_tgt_dictionary.get(word,word) for word in trans_gram_sentence.split()])
                 
-                translated_sentence.append(trans_lang_sentence)
+                translated_sentences.append(trans_lang_sentence)
+
+                # Get 1 sentence only, will support multi sentence
+                break
+
+            if translated == False:
+                failed_sentences.append(sentence)
+
+        # String to display failed sentence
+        failed_sentences = '\n'.join(failed_sentences)
+
+        if len(failed_sentences) > 0:
+            raise ValueError(f"Please check your grammar again, failed to translated these sentence \n {failed_sentences}")
 
-        return translated_sentence
\ No newline at end of file
+        return translated_sentences
\ No newline at end of file
diff --git a/urbamt/utils/tree_manipulation.py b/urbamt/utils/tree_manipulation.py
index 85a2ec7..0e8c3d4 100644
--- a/urbamt/utils/tree_manipulation.py
+++ b/urbamt/utils/tree_manipulation.py
@@ -75,7 +75,6 @@ def calculate_displacement(src_grammar, tgt_grammar):
     src_grammar_lst = src_grammar.split()
     tgt_grammar_lst = tgt_grammar.split()
     
-    print(src_grammar_lst.index("->"))
     src_grammar_lst = src_grammar_lst[src_grammar_lst.index("->")+1:]
     tgt_grammar_lst = tgt_grammar_lst[tgt_grammar_lst.index("->")+1:]
     displacement = []