st.markdown("---")
text_citation = st.text_input(label="Enter a citation string", value=citation_selected)
parse_citation_clicked = st.button("Parse Citation")
if parse_citation_clicked:
citation_selected = text_citation
response = requests.get(f"http://localhost:8000/parscit/{citation_selected}")
json = response.json()
text = json["text_tokens"]
tags = json["tags"].split()
# tokenize the text using white space
tokenizer = WordTokenizer(tokenizer="spacy-whitespace")
doc = tokenizer.nlp(" ".join(text))
# start index of every token
token_indices = [token.idx for token in doc]
# get start end index of every word
start_end_indices = itertools.zip_longest(
token_indices, token_indices[1:], fillvalue=len(" ".join(text))
)
start_end_indices = list(start_end_indices)
HTML_WRAPPER = """<div style="overflow-x: auto; border: 1px solid #e6e9ef; border-radius: 0.25rem; padding: 1rem; margin-bottom: 2.5rem">{}</div>"""
ents = []
for tag, (start_idx, end_idx) in zip(tags, start_end_indices):
ents.append({"start": start_idx, "end": end_idx, "label": tag})
f"{header} ({normalized})" for header, normalized in zip(
section_headers, normalized_section_headers)
]
st.write("### Sections (Normalized Sections)")
st.write(
HTML_WRAPPER.format("<br />".join(header_normalized_header)),
unsafe_allow_html=True,
)
st.write("### Parsed References. ")
for reference, tags in zip(references, parsed_reference_strings):
HTML_WRAPPER = """<div style="overflow-x: auto; border: 1px solid #e6e9ef; border-radius: 0.25rem; padding: 1rem; margin-bottom: 2.5rem">{}</div>"""
tokenizer = WordTokenizer(tokenizer="spacy-whitespace")
doc = tokenizer.nlp(reference)
# start index of every token
token_indices = [token.idx for token in doc]
# get start end index of every word
start_end_indices = itertools.zip_longest(token_indices,
token_indices[1:],
fillvalue=len(reference))
start_end_indices = list(start_end_indices)
ents = []
for tag, (start_idx, end_idx) in zip(tags.split(),
start_end_indices):
ents.append({"start": start_idx, "end": end_idx, "label": tag})