first step many corpora

.gitignore

@@ -0,0 +1 @@
+*.code-workspace

app.py

@@ -66,6 +66,14 @@ st.sidebar.markdown(
     unsafe_allow_html=True,
 )
 
+corpus = st.sidebar.selectbox(
+    "Corpus",
+    (
+        "LAION",
+        "C4",
+    ),
+    index=3,
+)
 query = st.sidebar.text_input(label="Search query", value="")
 language = st.sidebar.selectbox(
     "Language",
@@ -116,18 +124,20 @@ footer = """
 st.sidebar.markdown(footer, unsafe_allow_html=True)
 
 
-def scisearch(query, language, num_results=10):
+def scisearch(query, corpus, language, num_results=10):
     try:
         query = query.strip()
         if query == "" or query is None:
             return
 
+        corpus = corpus.strip()
+        address = os.environ.get("address") if corpus == "LAION" else os.environ.get("address")
         post_data = {"query": query, "k": num_results}
         if language != "detect_language":
             post_data["lang"] = language
 
         output = requests.post(
-            os.environ.get("address"),
+            address, # os.environ.get("address"),
             headers={"Content-type": "application/json"},
             data=json.dumps(post_data),
             timeout=60,
@@ -143,7 +153,6 @@ def scisearch(query, language, num_results=10):
                     Detected language <b>{detected_lang}</b> is not supported.<br>
                     Please choose a language from the dropdown or type another query.
                     </p><br><hr><br>"""
-
         results = payload["results"]
         highlight_terms = payload["highlight_terms"]
     except Exception as e:
@@ -206,7 +215,7 @@ def process_results(hits: list, highlight_terms: list) -> str:
 
 
 if st.sidebar.button("Search"):
-    hits, highlight_terms = scisearch(query, LANG_MAPPING[language], max_results)
+    hits, highlight_terms = scisearch(query, corpus, LANG_MAPPING[language], max_results)
     html_results = process_results(hits, highlight_terms)
     rendered_results = f"""
         <div id="searchresultsarea">