def malformed_query_warning_html(bad_search_txt):
warning_header_txt = f'Oops, we didn\'t understand that search.'
warning_body_txt = \
f'\n Your search was: "{bad_search_txt}"\n. Try the format entity1: ' \
f'value1, entity2: value2. For example: "material: PbTe, ' \
f'property: thermoelectric"'
return common_warning_html(warning_header_txt, warning_body_txt)
def malformed_query_warning_html(bad_search_txt):
"""
The html block when a query is malformed and searched.
Args:
bad_search_txt (str): The search text.
Returns:
(dash_html_components.Div): The html block when a malformed query is
entered and searched.
"""
warning_header_txt = f"Oops, we didn't understand that search."
warning_body_txt = (
f'\n Your search was: "{bad_search_txt}"\n. Try the format entity1: '
f'value1, entity2: value2. For example: "material: PbTe, '
f'property: thermoelectric"')
return common_warning_html(warning_header_txt, warning_body_txt, "is-half")
def abstracts_entities_results_html(text, normalize):
# Extract highlighted
try:
result = rester.get_ner_tags(text, concatenate=True, normalize=normalize)
except MatScholarRestError:
rester_error_txt = \
"Our server is having trouble with that abstract. We are likely " \
"undergoing maintenance, check back soon!"
return common_rester_error_html(rester_error_txt)
tagged_doc = result["tags"]
relevance = result["relevance"]
highlighted = highlight_entities(tagged_doc)
# Add the warning
if not relevance:
warning_header_txt = "Warning! Abstract not relevant."
warning_body_txt = \
"Our classifier has flagged this document as not relevant to " \
"inorganic materials science. Expect lower than optimum " \
"performance."
warning = common_warning_html(warning_header_txt, warning_body_txt)
else:
warning = html.Div("")
# Update download link
doc = {"sentences": []}
for sent in tagged_doc:
new_sent = []
for token, tag in sent:
new_sent.append({"token": token, "tag": tag})
doc["sentences"].append(new_sent)
json_string = json.dumps(doc)
json_string = "data:text/csv;charset=utf-8," + \
urllib.parse.quote(json_string)
download_link = html.A(
"Download entities as json",
id="entity-download-link",
href=json_string,
download="tagged_docs.json",
target="_blank"
)
download_container = html.Div(
download_link,
className="has-text-size-4 has-margin-top 10"
)
label = html.Label("Extracted Entity Tags:")
label_container = html.Div(label,
className="is-size-4 has-margin-top-30")
highlighted_container = html.Div(highlighted)
label_label = html.Label("Labels:")
label_label_container = html.Div(label_label,
className="is-size-4 has-margin-top-30")
entity_colormap_key = copy.deepcopy(entity_color_map_extended)
entities_keys = []
for e, color in entity_colormap_key.items():
# don't need the "other" label
if e == "other":
continue
entity_key = html.Div(e, className=f"is-size-4 msweb-is-{color}-txt has-text-weight-bold")
entity_key_container = html.Div(
entity_key,
className="flex-column is-narrow has-margin-5 box"
)
entities_keys.append(entity_key_container)
entity_key_container = html.Div(
entities_keys,
className="columns is-multiline has-margin-5"
)
results = html.Div(
[
warning,
label_container,
highlighted_container,
label_label_container,
entity_key_container,
download_container
]
)
return results
def extract_entities_results_html(text, normalize):
"""
Get an html block of the results for an entity extraction.
Args:
text (str): The abstract text to extract entities from.
normalize (bool): Whether to normalize the entities or not.
Returns:
(dash_html_components.Div): The html block for the entity extraction
results.
"""
try:
result = rester.get_ner_tags(text,
concatenate=True,
normalize=normalize)
except MatScholarRestError:
rester_error_txt = RESTER_ERROR_TEXT
return common_rester_error_html(rester_error_txt)
tagged_doc = result["tags"]
relevance = result["relevance"]
highlighted = highlight_entities_html(tagged_doc)
# Add the warning
if not relevance:
warning_header_txt = "Warning! Abstract not relevant."
warning_body_txt = (
"Our classifier has flagged this document as not relevant to "
"inorganic materials science. Expect lower than optimum "
"performance.")
warning = common_warning_html(warning_header_txt, warning_body_txt,
"is-fullwidth")
else:
warning = html.Div("")
# Update download link
doc = {"sentences": []}
for sent in tagged_doc:
new_sent = []
for token, tag in sent:
new_sent.append({"token": token, "tag": tag})
doc["sentences"].append(new_sent)
json_string = json.dumps(doc)
json_string = "data:text/csv;charset=utf-8," + urllib.parse.quote(
json_string)
download_link = html.A(
"Download entities as json",
id="entity-download-link",
href=json_string,
download="tagged_docs.json",
target="_blank",
)
download_container = html.Div(
download_link, className="has-text-size-4 has-margin-top 10")
label = html.Label("Extracted Entity Tags:")
label_container = html.Div(label, className="is-size-4 has-margin-top-30")
highlighted_container = html.Div(highlighted)
label_label = html.Label("Labels:")
label_label_container = html.Div(label_label,
className="is-size-4 has-margin-top-30")
entity_colormap_key = copy.deepcopy(entity_color_map_extended)
entities_keys = []
for e, color in entity_colormap_key.items():
# don't need the "other" label
if e == "other":
continue
entity_key = html.Div(
e,
className=f"is-size-4 msweb-is-{color}-txt has-text-weight-bold")
entity_key_container = html.Div(
entity_key, className="flex-column is-narrow has-margin-5 box")
entities_keys.append(entity_key_container)
entity_key_container = html.Div(
entities_keys, className="columns is-multiline has-margin-5")
results = html.Div([
warning,
label_container,
highlighted_container,
label_label_container,
entity_key_container,
download_container,
])
return results