streamlit-octostar-research 0.1.7

Streamlit component that connects to the Octostar Research App

streamlit-octostar-research

Streamlit component that connects to the Octostar Research App

Installation instructions

pip install streamlit-octostar-research

Usage instructions

from streamlit_octostar_research.desktop import searchXperience, result, get_open_workspace_ids, close_workspace, get_current_workspace_id, show_notification, open
from streamlit_octostar_research.ontology import send_query
from streamlit_octostar_research.extras import create_link_chart
import csv
from io import StringIO


st.header("This is a streamlit app")
st.subheader("Connected to the Octostar")

current_workspace_id = get_current_workspace_id();
open_workspaces = get_open_workspace_ids()


if st.button("Search Xperience"):
    searchXperience(key='search1')

search1 = result(key='search1')
if search1 is not None:
    st.header("Search Result")
    st.json(search1, expanded=True)
    if len(search1) > 0:
        show_notification(f"Search result is available: {len(search1)} items")

query = st.text_area('Search', '''
    select count(*) from dtimbr.person where entity_label like '%Robert%'
    ''')

if query and st.button("Run Query"):
    send_query(query, key='search2')

search2 = result(key='search2')
if search2 is not None:
        st.header("Query Result")
        st.json(search2, expanded=True)

chartdata = st.text_area('New Link Chart Data', '''
node|Giovanni|person|Giovanni
node|Robert|person|Robert
node|Simone|person|Simone
node|Octostar Research|company|Octostar Research
node|Investigation Software|product|Investigation Software
node|Investricor|company|Investricor
node|John|person|John
edge|Giovanni|Octostar Research|owns
edge|Robert|Octostar Research|works for
edge|Simone|Octostar Research|works for
edge|John|Investricor|works for
edge|Investricor|Octostar Research|invests in
edge|Octostar Research|Investigation Software|develops

    ''')
linkchart1 = result(key='linkchart1')
if chartdata and st.button("Create Link Chart"):
    column_names = ['type', 'entity_id', 'entity_type', 'entity_label']

    # Create a DictReader to parse the input string and map the column names
    csv_reader = csv.DictReader(StringIO(chartdata), delimiter='|', fieldnames=column_names)

    nodes = [{key: value for key, value in row.items() if key != 'type'}
             for row in csv_reader if row['type'] == "node"]

    column_names = ['type', 'from', 'to', 'label']
    csv_reader = csv.DictReader(StringIO(chartdata), delimiter='|', fieldnames=column_names)
    edges = [{key: value for key, value in row.items() if key != 'type'}
             for row in csv_reader if row['type'] == "edge"]

    create_link_chart(key='linkchart1', nodes=nodes, edges=edges, name="Octostar Research", draft=True, path="linkcharts")

if linkchart1 is not None:
    open(linkchart1)
    st.header("Link Chart")
    st.json(linkchart1, expanded=True)


if st.button("Error Notification"):
    show_notification({"message": "From Streamlit", "description": "Example of an error notification", "level": "error"})


st.header("About the Desktop")
st.write(f"Current workspace id: {current_workspace_id}")

if open_workspaces is not None:
    st.header("Open Workspaces (ids)")
    st.subheader("See what can happen, with the right permissions")
    cols = st.columns((1, 1))
    fields = ['uid', "action"]
    for col, field_name in zip(cols, fields):
        col.write(field_name)

    for ws in open_workspaces:
        col1, col2  = st.columns((1, 1))
        col1.write(ws)
        if col2.button(label='Remove from Desktop',key=f"del-{ws}", disabled=ws == current_workspace_id):
            close_workspace(ws)

else:
    st.write('No open workspaces found')