From 5864ba5ee4ccf078a2d15f8f9cfb375f4c47e4a6 Mon Sep 17 00:00:00 2001
From: Steffen Knapp <knapps@cl.uni-heidelberg.de>
Date: Thu, 25 Jan 2018 15:52:09 +0100
Subject: [PATCH] added wacky html quotation marks
---
punctuation_feature.py | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/punctuation_feature.py b/punctuation_feature.py
index 904eae1..c86a0b5 100644
--- a/punctuation_feature.py
+++ b/punctuation_feature.py
@@ -7,6 +7,7 @@ def extract(corpus_instance):
#"?!", "!?", "???", "!!!" are no lemmas
allcaps = 0
excessive_punctuation = re.compile('[!?][!?]+')
+ html_quotes = re.compile('"')
review_tokens = (word_tokenize(corpus_instance["TITLE"])) + corpus_instance["TOKENS"]
review_lemmas = ((corpus_instance["TITLE"] + " " + corpus_instance["REVIEW"]))
@@ -22,7 +23,7 @@ def extract(corpus_instance):
corpus_instance_vector.append(len(re.findall(excessive_punctuation, review_lemmas))/len(review_lemmas))
corpus_instance_vector.append(allcaps/len(review_tokens))
-
+ corpus_instance_vector.append(len(re.findall(html_quotes, review_lemmas))/len(review_tokens))
return corpus_instance_vector
@@ -33,5 +34,6 @@ if __name__ == '__main__':
"""
pass
#corpus = read_corpus("minicorpus.csv")
- #corpus_instance = corpus[3]
+ #corpus_instance = corpus[0]
+ #print(corpus_instance["REVIEW"])
#print(extract(corpus_instance))
--
GitLab