sid570 commited on
Commit
2fb7158
·
verified ·
1 Parent(s): 6dada24

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +9 -56
app.py CHANGED
@@ -77,7 +77,6 @@ def detect_entities(text):
77
  while i < len(tokens):
78
  token = tokens[i]
79
 
80
- # Sentence boundary
81
  if token in [".", "!", "?"]:
82
  sentence_start = True
83
  i += 1
@@ -87,74 +86,31 @@ def detect_entities(text):
87
  i += 1
88
  continue
89
 
90
- # ====================================================
91
- # RULE 1 — FIRST WORD OF SENTENCE (FIXED)
92
- # ====================================================
93
  if sentence_start:
94
- # 1️⃣ Dictionary has highest priority
95
  match = find_dictionary_match(tokens, i)
96
  if match:
97
- entities.append({
98
- "text": match["text"],
99
- "type": match["type"]
100
- })
101
  i += match["length"]
102
- sentence_start = False
103
- continue
104
-
105
- # 2️⃣ Capital phrase grouping at sentence start
106
- if token[0].isupper():
107
- phrase = [token]
108
- j = i + 1
109
-
110
- while (
111
- j < len(tokens)
112
- and tokens[j].isalpha()
113
- and tokens[j][0].isupper()
114
- ):
115
- phrase.append(tokens[j])
116
- j += 1
117
-
118
- phrase_text = " ".join(phrase)
119
-
120
- # Capitalized first phrase not in vocab → UNKNOWN
121
- if phrase_text.lower() not in VOCAB:
122
- entities.append({
123
- "text": phrase_text,
124
- "type": "UNKNOWN"
125
- })
126
-
127
- i = j
128
- sentence_start = False
129
- continue
130
 
131
- # 3️⃣ Lowercase first word → leave untouched
132
- i += 1
133
  sentence_start = False
134
  continue
135
 
136
- # ====================================================
137
- # RULE 2 — SECOND WORD ONWARDS (UNCHANGED)
138
- # ====================================================
139
  match = find_dictionary_match(tokens, i)
140
  if match:
141
- entities.append({
142
- "text": match["text"],
143
- "type": match["type"]
144
- })
145
  i += match["length"]
146
  continue
147
 
148
- # Capital phrase grouping (e.g. Shopping Mall)
149
  if token[0].isupper():
150
  phrase = [token]
151
  j = i + 1
152
-
153
- while (
154
- j < len(tokens)
155
- and tokens[j].isalpha()
156
- and tokens[j][0].isupper()
157
- ):
158
  phrase.append(tokens[j])
159
  j += 1
160
 
@@ -162,14 +118,11 @@ def detect_entities(text):
162
  "text": " ".join(phrase),
163
  "type": "UNKNOWN"
164
  })
165
-
166
  i = j
167
  continue
168
 
169
- # Normal word
170
  i += 1
171
 
172
- return entities
173
 
174
 
175
  # ====================================================
 
77
  while i < len(tokens):
78
  token = tokens[i]
79
 
 
80
  if token in [".", "!", "?"]:
81
  sentence_start = True
82
  i += 1
 
86
  i += 1
87
  continue
88
 
89
+ # RULE 1 — FIRST WORD
 
 
90
  if sentence_start:
 
91
  match = find_dictionary_match(tokens, i)
92
  if match:
93
+ entities.append({"text": match["text"], "type": match["type"]})
 
 
 
94
  i += match["length"]
95
+ else:
96
+ if token.lower() not in VOCAB:
97
+ entities.append({"text": token, "type": "UNKNOWN"})
98
+ i += 1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
99
 
 
 
100
  sentence_start = False
101
  continue
102
 
103
+ # RULE 2 — SECOND WORD ONWARDS
 
 
104
  match = find_dictionary_match(tokens, i)
105
  if match:
106
+ entities.append({"text": match["text"], "type": match["type"]})
 
 
 
107
  i += match["length"]
108
  continue
109
 
 
110
  if token[0].isupper():
111
  phrase = [token]
112
  j = i + 1
113
+ while j < len(tokens) and tokens[j].isalpha() and tokens[j][0].isupper():
 
 
 
 
 
114
  phrase.append(tokens[j])
115
  j += 1
116
 
 
118
  "text": " ".join(phrase),
119
  "type": "UNKNOWN"
120
  })
 
121
  i = j
122
  continue
123
 
 
124
  i += 1
125
 
 
126
 
127
 
128
  # ====================================================