Spaces:
Runtime error
Runtime error
Merge pull request #2 from ugm2/feature/add_url_indexing
Browse files- interface/components.py +21 -1
- interface/pages.py +2 -0
- interface/utils.py +11 -0
- requirements.txt +6 -5
interface/components.py
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
import streamlit as st
|
2 |
-
from interface.utils import get_pipelines
|
3 |
from interface.draw_pipelines import get_pipeline_graph
|
4 |
|
5 |
|
@@ -64,3 +64,23 @@ def component_text_input(container):
|
|
64 |
{"text": doc["text"], "id": doc_id} for doc_id, doc in enumerate(texts)
|
65 |
]
|
66 |
return corpus
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import streamlit as st
|
2 |
+
from interface.utils import get_pipelines, extract_text_from_url
|
3 |
from interface.draw_pipelines import get_pipeline_graph
|
4 |
|
5 |
|
|
|
64 |
{"text": doc["text"], "id": doc_id} for doc_id, doc in enumerate(texts)
|
65 |
]
|
66 |
return corpus
|
67 |
+
|
68 |
+
|
69 |
+
def component_article_url(container):
|
70 |
+
"""Draw the Article URL widget"""
|
71 |
+
with container:
|
72 |
+
urls = []
|
73 |
+
doc_id = 1
|
74 |
+
with st.expander("Enter URLs"):
|
75 |
+
while True:
|
76 |
+
url = st.text_input(f"URL {doc_id}", key=doc_id)
|
77 |
+
if url != "":
|
78 |
+
urls.append({"text": extract_text_from_url(url)})
|
79 |
+
doc_id += 1
|
80 |
+
st.markdown("---")
|
81 |
+
else:
|
82 |
+
break
|
83 |
+
corpus = [
|
84 |
+
{"text": doc["text"], "id": doc_id} for doc_id, doc in enumerate(urls)
|
85 |
+
]
|
86 |
+
return corpus
|
interface/pages.py
CHANGED
@@ -5,6 +5,7 @@ from interface.components import (
|
|
5 |
component_show_pipeline,
|
6 |
component_show_search_result,
|
7 |
component_text_input,
|
|
|
8 |
)
|
9 |
|
10 |
|
@@ -57,6 +58,7 @@ def page_index(container):
|
|
57 |
|
58 |
input_funcs = {
|
59 |
"Raw Text": (component_text_input, "card-text"),
|
|
|
60 |
}
|
61 |
selected_input = option_menu(
|
62 |
"Input Text",
|
|
|
5 |
component_show_pipeline,
|
6 |
component_show_search_result,
|
7 |
component_text_input,
|
8 |
+
component_article_url,
|
9 |
)
|
10 |
|
11 |
|
|
|
58 |
|
59 |
input_funcs = {
|
60 |
"Raw Text": (component_text_input, "card-text"),
|
61 |
+
"URL": (component_article_url, "card-link"),
|
62 |
}
|
63 |
selected_input = option_menu(
|
64 |
"Input Text",
|
interface/utils.py
CHANGED
@@ -1,5 +1,7 @@
|
|
1 |
import core.pipelines as pipelines_functions
|
2 |
from inspect import getmembers, isfunction
|
|
|
|
|
3 |
|
4 |
|
5 |
def get_pipelines():
|
@@ -10,3 +12,12 @@ def get_pipelines():
|
|
10 |
" ".join([n.capitalize() for n in name.split("_")]) for name in pipeline_names
|
11 |
]
|
12 |
return pipeline_names, pipeline_funcs
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import core.pipelines as pipelines_functions
|
2 |
from inspect import getmembers, isfunction
|
3 |
+
from newspaper import Article
|
4 |
+
import streamlit as st
|
5 |
|
6 |
|
7 |
def get_pipelines():
|
|
|
12 |
" ".join([n.capitalize() for n in name.split("_")]) for name in pipeline_names
|
13 |
]
|
14 |
return pipeline_names, pipeline_funcs
|
15 |
+
|
16 |
+
|
17 |
+
@st.experimental_memo
|
18 |
+
def extract_text_from_url(url: str):
|
19 |
+
article = Article(url)
|
20 |
+
article.download()
|
21 |
+
article.parse()
|
22 |
+
|
23 |
+
return article.text
|
requirements.txt
CHANGED
@@ -1,5 +1,6 @@
|
|
1 |
-
streamlit
|
2 |
-
streamlit_option_menu
|
3 |
-
farm-haystack
|
4 |
-
black
|
5 |
-
plotly
|
|
|
|
1 |
+
streamlit==1.12.2
|
2 |
+
streamlit_option_menu==0.3.2
|
3 |
+
farm-haystack==1.8.0
|
4 |
+
black==22.8.0
|
5 |
+
plotly==5.10.0
|
6 |
+
newspaper3k==0.2.8
|