Update src/streamlit_app.py
Browse files- src/streamlit_app.py +54 -80
src/streamlit_app.py
CHANGED
|
@@ -1,67 +1,48 @@
|
|
| 1 |
import os
|
| 2 |
-
os.environ["HOME"] = "/tmp" #
|
| 3 |
|
| 4 |
import streamlit as st
|
| 5 |
import requests
|
| 6 |
import urllib.parse
|
| 7 |
-
from langdetect import detect
|
| 8 |
|
| 9 |
st.set_page_config(page_title="WikiTrail", layout="wide")
|
| 10 |
|
| 11 |
-
|
| 12 |
-
st.markdown(""
|
| 13 |
-
<style>
|
| 14 |
-
body {
|
| 15 |
-
font-family: 'Segoe UI', sans-serif;
|
| 16 |
-
background-color: #f8f9fa;
|
| 17 |
-
}
|
| 18 |
-
.title {
|
| 19 |
-
font-size: 48px;
|
| 20 |
-
text-align: center;
|
| 21 |
-
color: #2c3e50;
|
| 22 |
-
font-weight: bold;
|
| 23 |
-
margin-bottom: 5px;
|
| 24 |
-
}
|
| 25 |
-
.subtitle {
|
| 26 |
-
text-align: center;
|
| 27 |
-
font-size: 18px;
|
| 28 |
-
color: #555;
|
| 29 |
-
margin-top: 0px;
|
| 30 |
-
}
|
| 31 |
-
.section {
|
| 32 |
-
background-color: #ffffff;
|
| 33 |
-
padding: 20px;
|
| 34 |
-
border-radius: 16px;
|
| 35 |
-
box-shadow: 0 4px 12px rgba(0,0,0,0.08);
|
| 36 |
-
margin-bottom: 25px;
|
| 37 |
-
}
|
| 38 |
-
.emoji-title {
|
| 39 |
-
font-size: 28px;
|
| 40 |
-
margin-bottom: 10px;
|
| 41 |
-
color: #2c3e50;
|
| 42 |
-
}
|
| 43 |
-
</style>
|
| 44 |
-
""", unsafe_allow_html=True)
|
| 45 |
-
|
| 46 |
-
# Title
|
| 47 |
-
st.markdown("<div class='title'>📚 WikiTrail</div>", unsafe_allow_html=True)
|
| 48 |
-
st.markdown("<div class='subtitle'>Explore Wikipedia topics visually and get a summarized journey.</div>", unsafe_allow_html=True)
|
| 49 |
|
| 50 |
-
# Language
|
| 51 |
languages = {
|
| 52 |
"English": "en",
|
| 53 |
"Hindi (हिन्दी)": "hi",
|
| 54 |
"Telugu (తెలుగు)": "te",
|
| 55 |
"Tamil (தமிழ்)": "ta"
|
| 56 |
}
|
| 57 |
-
lang_name = st.selectbox("🌐 Select Language", list(languages.keys())
|
| 58 |
lang_code = languages[lang_name]
|
| 59 |
|
| 60 |
-
# Topic
|
| 61 |
-
topic_input = st.text_input("🔍 Enter a topic (in English)", placeholder="e.g., India, Gandhi,
|
| 62 |
|
| 63 |
-
# Wikipedia
|
| 64 |
-
def
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
safe_title = urllib.parse.quote(title.replace(" ", "_"))
|
| 66 |
url = f"https://{lang}.wikipedia.org/api/rest_v1/page/summary/{safe_title}"
|
| 67 |
res = requests.get(url)
|
|
@@ -74,8 +55,8 @@ def fetch_topic_summary(title, lang):
|
|
| 74 |
}
|
| 75 |
return None
|
| 76 |
|
| 77 |
-
#
|
| 78 |
-
def
|
| 79 |
safe_title = urllib.parse.quote(title.replace(" ", "_"))
|
| 80 |
url = f"https://{lang}.wikipedia.org/w/api.php?action=query&format=json&origin=*&titles={safe_title}&prop=links&pllimit=5"
|
| 81 |
res = requests.get(url)
|
|
@@ -86,43 +67,40 @@ def fetch_related_topics(title, lang):
|
|
| 86 |
return [link['title'] for link in pages[0]['links']]
|
| 87 |
return []
|
| 88 |
|
| 89 |
-
#
|
| 90 |
-
def summarize_bullets(
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
-
top = cleaned[:limit]
|
| 96 |
-
return ["• " + s + "." for s in top] if top else ["No summary available."]
|
| 97 |
|
| 98 |
# Main logic
|
| 99 |
if topic_input:
|
| 100 |
with st.spinner("🔍 Searching Wikipedia..."):
|
| 101 |
summaries = []
|
| 102 |
|
| 103 |
-
|
| 104 |
-
|
|
|
|
|
|
|
| 105 |
|
| 106 |
-
|
| 107 |
-
|
|
|
|
|
|
|
| 108 |
summaries.append(main["summary"])
|
| 109 |
st.markdown(f"### {main['title']}")
|
| 110 |
st.write(main["summary"])
|
| 111 |
st.markdown(f"[Read More →]({main['link']})", unsafe_allow_html=True)
|
| 112 |
else:
|
| 113 |
-
st.warning(
|
| 114 |
-
st.markdown("</div>", unsafe_allow_html=True)
|
| 115 |
-
st.stop()
|
| 116 |
-
|
| 117 |
-
st.markdown("</div>", unsafe_allow_html=True)
|
| 118 |
|
| 119 |
-
# Related
|
| 120 |
-
st.
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
| 124 |
-
|
| 125 |
-
data = fetch_topic_summary(rel, lang_code)
|
| 126 |
if data and data["summary"] not in summaries:
|
| 127 |
summaries.append(data["summary"])
|
| 128 |
with st.expander(data["title"]):
|
|
@@ -130,12 +108,8 @@ if topic_input:
|
|
| 130 |
st.markdown(f"[Read More →]({data['link']})", unsafe_allow_html=True)
|
| 131 |
else:
|
| 132 |
st.info("No related topics found.")
|
| 133 |
-
st.markdown("</div>", unsafe_allow_html=True)
|
| 134 |
|
| 135 |
-
#
|
| 136 |
-
st.
|
| 137 |
-
|
| 138 |
-
|
| 139 |
-
for b in bullets:
|
| 140 |
-
st.markdown(b)
|
| 141 |
-
st.markdown("</div>", unsafe_allow_html=True)
|
|
|
|
| 1 |
import os
|
| 2 |
+
os.environ["HOME"] = "/tmp" # Hugging Face fix
|
| 3 |
|
| 4 |
import streamlit as st
|
| 5 |
import requests
|
| 6 |
import urllib.parse
|
|
|
|
| 7 |
|
| 8 |
st.set_page_config(page_title="WikiTrail", layout="wide")
|
| 9 |
|
| 10 |
+
st.title("📚 WikiTrail")
|
| 11 |
+
st.markdown("Explore Wikipedia topics visually and get a summarized journey.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
|
| 13 |
+
# Language selector
|
| 14 |
languages = {
|
| 15 |
"English": "en",
|
| 16 |
"Hindi (हिन्दी)": "hi",
|
| 17 |
"Telugu (తెలుగు)": "te",
|
| 18 |
"Tamil (தமிழ்)": "ta"
|
| 19 |
}
|
| 20 |
+
lang_name = st.selectbox("🌐 Select Language", list(languages.keys()))
|
| 21 |
lang_code = languages[lang_name]
|
| 22 |
|
| 23 |
+
# Topic input
|
| 24 |
+
topic_input = st.text_input("🔍 Enter a topic (in English)", placeholder="e.g., India, Gandhi, Telangana")
|
| 25 |
|
| 26 |
+
# Translate topic using Wikipedia search
|
| 27 |
+
def get_translated_title(query, lang):
|
| 28 |
+
search_url = f"https://{lang}.wikipedia.org/w/api.php"
|
| 29 |
+
params = {
|
| 30 |
+
"action": "query",
|
| 31 |
+
"list": "search",
|
| 32 |
+
"srsearch": query,
|
| 33 |
+
"format": "json",
|
| 34 |
+
"origin": "*"
|
| 35 |
+
}
|
| 36 |
+
res = requests.get(search_url, params=params)
|
| 37 |
+
if res.status_code == 200:
|
| 38 |
+
results = res.json()
|
| 39 |
+
search_list = results.get("query", {}).get("search", [])
|
| 40 |
+
if search_list:
|
| 41 |
+
return search_list[0]["title"]
|
| 42 |
+
return None
|
| 43 |
+
|
| 44 |
+
# Fetch summary
|
| 45 |
+
def fetch_summary(title, lang):
|
| 46 |
safe_title = urllib.parse.quote(title.replace(" ", "_"))
|
| 47 |
url = f"https://{lang}.wikipedia.org/api/rest_v1/page/summary/{safe_title}"
|
| 48 |
res = requests.get(url)
|
|
|
|
| 55 |
}
|
| 56 |
return None
|
| 57 |
|
| 58 |
+
# Fetch related
|
| 59 |
+
def fetch_related(title, lang):
|
| 60 |
safe_title = urllib.parse.quote(title.replace(" ", "_"))
|
| 61 |
url = f"https://{lang}.wikipedia.org/w/api.php?action=query&format=json&origin=*&titles={safe_title}&prop=links&pllimit=5"
|
| 62 |
res = requests.get(url)
|
|
|
|
| 67 |
return [link['title'] for link in pages[0]['links']]
|
| 68 |
return []
|
| 69 |
|
| 70 |
+
# Summarizer
|
| 71 |
+
def summarize_bullets(texts, limit=3):
|
| 72 |
+
full_text = ' '.join(set(texts))
|
| 73 |
+
sentences = full_text.replace('।', '.').replace('?', '.').replace('!', '.').split('.')
|
| 74 |
+
clean = [s.strip() for s in sentences if s.strip()]
|
| 75 |
+
return ["• " + s + "." for s in clean[:limit]] if clean else ["No summary available."]
|
|
|
|
|
|
|
| 76 |
|
| 77 |
# Main logic
|
| 78 |
if topic_input:
|
| 79 |
with st.spinner("🔍 Searching Wikipedia..."):
|
| 80 |
summaries = []
|
| 81 |
|
| 82 |
+
translated_title = get_translated_title(topic_input, lang_code)
|
| 83 |
+
if not translated_title:
|
| 84 |
+
st.error(f"No matching page found in {lang_name} for '{topic_input}'")
|
| 85 |
+
st.stop()
|
| 86 |
|
| 87 |
+
# Main topic
|
| 88 |
+
st.subheader("🔷 Main Topic")
|
| 89 |
+
main = fetch_summary(translated_title, lang_code)
|
| 90 |
+
if main:
|
| 91 |
summaries.append(main["summary"])
|
| 92 |
st.markdown(f"### {main['title']}")
|
| 93 |
st.write(main["summary"])
|
| 94 |
st.markdown(f"[Read More →]({main['link']})", unsafe_allow_html=True)
|
| 95 |
else:
|
| 96 |
+
st.warning("Couldn't fetch main topic summary.")
|
|
|
|
|
|
|
|
|
|
|
|
|
| 97 |
|
| 98 |
+
# Related
|
| 99 |
+
st.subheader("🔗 Related Topics")
|
| 100 |
+
related_titles = fetch_related(translated_title, lang_code)
|
| 101 |
+
if related_titles:
|
| 102 |
+
for title in related_titles:
|
| 103 |
+
data = fetch_summary(title, lang_code)
|
|
|
|
| 104 |
if data and data["summary"] not in summaries:
|
| 105 |
summaries.append(data["summary"])
|
| 106 |
with st.expander(data["title"]):
|
|
|
|
| 108 |
st.markdown(f"[Read More →]({data['link']})", unsafe_allow_html=True)
|
| 109 |
else:
|
| 110 |
st.info("No related topics found.")
|
|
|
|
| 111 |
|
| 112 |
+
# Bullet summary
|
| 113 |
+
st.subheader("🧠 Combined Summary")
|
| 114 |
+
for bullet in summarize_bullets(summaries):
|
| 115 |
+
st.markdown(bullet)
|
|
|
|
|
|
|
|
|