Spaces:
Build error
Build error
Commit
·
6977cda
1
Parent(s):
1a9aed5
Fix the extract summary display
Browse files- Summarizer.py +10 -8
- app.py +11 -13
Summarizer.py
CHANGED
|
@@ -27,20 +27,22 @@ class Summarizer:
|
|
| 27 |
summarized_list.append(sentence._text)
|
| 28 |
return summarized_list
|
| 29 |
|
| 30 |
-
def __extractive_summary(self, parser: DocumentParser, sentences_count):
|
| 31 |
summarized_sentences = self.lsa_summarizer(parser.document, sentences_count)
|
| 32 |
summarized_list = Summarizer.sentence_list(summarized_sentences)
|
| 33 |
-
|
| 34 |
-
return all_sentences_list, summarized_list
|
| 35 |
|
| 36 |
-
def extractive_summary_from_text(self, text: str, sentences_count: int) ->
|
| 37 |
parser = PlaintextParser.from_string(text, Tokenizer(Summarizer.DEFAULT_LANGUAGE))
|
| 38 |
return self.__extractive_summary(parser, sentences_count)
|
| 39 |
|
| 40 |
-
def extractive_summary_from_url(self, url: str, sentences_count: int) ->
|
| 41 |
parser = HtmlParser.from_url(url, Tokenizer(Summarizer.DEFAULT_LANGUAGE))
|
| 42 |
return self.__extractive_summary(parser, sentences_count)
|
| 43 |
|
| 44 |
-
def abstractive_summary(self,
|
| 45 |
-
|
| 46 |
-
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
summarized_list.append(sentence._text)
|
| 28 |
return summarized_list
|
| 29 |
|
| 30 |
+
def __extractive_summary(self, parser: DocumentParser, sentences_count) -> list:
|
| 31 |
summarized_sentences = self.lsa_summarizer(parser.document, sentences_count)
|
| 32 |
summarized_list = Summarizer.sentence_list(summarized_sentences)
|
| 33 |
+
return summarized_list
|
|
|
|
| 34 |
|
| 35 |
+
def extractive_summary_from_text(self, text: str, sentences_count: int) -> list:
|
| 36 |
parser = PlaintextParser.from_string(text, Tokenizer(Summarizer.DEFAULT_LANGUAGE))
|
| 37 |
return self.__extractive_summary(parser, sentences_count)
|
| 38 |
|
| 39 |
+
def extractive_summary_from_url(self, url: str, sentences_count: int) -> list:
|
| 40 |
parser = HtmlParser.from_url(url, Tokenizer(Summarizer.DEFAULT_LANGUAGE))
|
| 41 |
return self.__extractive_summary(parser, sentences_count)
|
| 42 |
|
| 43 |
+
def abstractive_summary(self, extract_summary_sentences: list) -> list:
|
| 44 |
+
extract_summary = " ".join([sentence for sentence in extract_summary_sentences])
|
| 45 |
+
abstractive_summary_list = []
|
| 46 |
+
for result in self.pipeline(wrap(extract_summary, 512)):
|
| 47 |
+
abstractive_summary_list.append(result['summary_text'])
|
| 48 |
+
return abstractive_summary_list
|
app.py
CHANGED
|
@@ -33,19 +33,20 @@ def main() -> None:
|
|
| 33 |
)
|
| 34 |
return terms_and_conditions_pipeline
|
| 35 |
|
| 36 |
-
def display_abstractive_summary(
|
| 37 |
st.subheader("Abstractive Summary")
|
| 38 |
st.markdown('#####')
|
| 39 |
-
|
|
|
|
| 40 |
|
| 41 |
-
def display_extractive_summary(
|
| 42 |
st.subheader("Extractive Summary")
|
| 43 |
st.markdown('#####')
|
| 44 |
-
|
| 45 |
-
replaced_text = terms_and_conditions
|
| 46 |
for sentence in summary_sentences:
|
| 47 |
replaced_text = replaced_text.replace(sentence,
|
| 48 |
f"<span style='background-color: #FFFF00'>{sentence}</span>")
|
|
|
|
| 49 |
st.write(replaced_text, unsafe_allow_html=True)
|
| 50 |
|
| 51 |
def is_valid_url(url: str) -> bool:
|
|
@@ -82,17 +83,14 @@ def main() -> None:
|
|
| 82 |
if submit_button:
|
| 83 |
|
| 84 |
if is_valid_url(tc_text_input):
|
| 85 |
-
|
| 86 |
-
sentences_length_input)
|
| 87 |
else:
|
| 88 |
-
|
| 89 |
-
sentences_length_input)
|
| 90 |
|
| 91 |
-
|
| 92 |
-
abstract_summary = summarizer.abstractive_summary(extract_summary)
|
| 93 |
|
| 94 |
-
|
| 95 |
-
|
| 96 |
|
| 97 |
|
| 98 |
if __name__ == "__main__":
|
|
|
|
| 33 |
)
|
| 34 |
return terms_and_conditions_pipeline
|
| 35 |
|
| 36 |
+
def display_abstractive_summary(summary_sentences: list) -> None:
|
| 37 |
st.subheader("Abstractive Summary")
|
| 38 |
st.markdown('#####')
|
| 39 |
+
for sentence in summary_sentences:
|
| 40 |
+
st.markdown(f"- {sentence}", unsafe_allow_html=True)
|
| 41 |
|
| 42 |
+
def display_extractive_summary(terms_and_conditions_text: list, summary_sentences: list) -> None:
|
| 43 |
st.subheader("Extractive Summary")
|
| 44 |
st.markdown('#####')
|
| 45 |
+
replaced_text = terms_and_conditions_text
|
|
|
|
| 46 |
for sentence in summary_sentences:
|
| 47 |
replaced_text = replaced_text.replace(sentence,
|
| 48 |
f"<span style='background-color: #FFFF00'>{sentence}</span>")
|
| 49 |
+
replaced_text = replaced_text.replace('/\r?\n/g', '<br/>')
|
| 50 |
st.write(replaced_text, unsafe_allow_html=True)
|
| 51 |
|
| 52 |
def is_valid_url(url: str) -> bool:
|
|
|
|
| 83 |
if submit_button:
|
| 84 |
|
| 85 |
if is_valid_url(tc_text_input):
|
| 86 |
+
extract_summary_sentences = summarizer.extractive_summary_from_url(tc_text_input, sentences_length_input)
|
|
|
|
| 87 |
else:
|
| 88 |
+
extract_summary_sentences = summarizer.extractive_summary_from_text(tc_text_input, sentences_length_input)
|
|
|
|
| 89 |
|
| 90 |
+
abstract_summary_list = summarizer.abstractive_summary(extract_summary_sentences)
|
|
|
|
| 91 |
|
| 92 |
+
display_abstractive_summary(abstract_summary_list)
|
| 93 |
+
display_extractive_summary(tc_text_input, extract_summary_sentences)
|
| 94 |
|
| 95 |
|
| 96 |
if __name__ == "__main__":
|