including metadata

app.py

@@ -14,7 +14,7 @@ from flair_recognizer import FlairRecognizer
 
 st.title("Anonymise your text!")
 st.markdown(
-    "This mini-app anonymises text using Flair. You can find the code on [GitHub(WIP)](#)"
+    "This mini-app anonymises text using Flair. You can find the code in the Files and versions tab above."
 )
 
 # Configure logger
@@ -88,8 +88,8 @@ def analyze_text():
                 return_decision_process=False,
             )
 
-            # if st.session_state.metadata:
-            #     analyze_results = include_manual_input(analyze_results)
+            if st.session_state.excluded_words:
+                analyze_results = include_manual_input(analyze_results)
 
             if st.session_state.allowed_words:
                 analyze_results = exclude_manual_input(analyze_results)
@@ -101,19 +101,22 @@ def analyze_text():
             )
             
 
-# def include_manual_input(analyze_results):
-#     analyze_results_extended=[]
-#     logging.info(
-#                 f"analyse results before adding extra words: {analyze_results}\n"
-#             )
-#     for word in st.session_state.text.split():
-#         if word in st.session_state.metadata:
-#             [m.start() for m in re.finditer('test', 'test test test test')]
-#             analyze_results_extended.append("type: MANUAL, start: 0, end: 3, score: 1.0")
-#     logging.info(
-#                 f"analyse results after adding allowed words: {analyze_results_extended}\n"
-#             )
-#     return analyze_results
+def include_manual_input(analyze_results):
+    analyze_results_extended=[]
+    logging.info(
+                f"analyse results before adding extra words: {analyze_results}\n"
+            )
+    for word in st.session_state.excluded_words:
+        if word in st.session_state.text:
+            index_entries = [m.start() for m in re.finditer(word, st.session_state.text)]
+            for entry in index_entries:
+                start=entry
+                end=entry+len(word)
+                analyze_results_extended.append("type: MANUAL ADD, start: {}, end: {}, score: 1.0".format(start,end))
+    logging.info(
+                f"analyse results after adding allowed words: {analyze_results_extended}\n"
+            )
+    return analyze_results_extended
 
 def exclude_manual_input(analyze_results):
     analyze_results_fltered=[]
@@ -163,9 +166,9 @@ def clear_results():
     st.session_state.anon_results=""
     st.session_state.analyze_results=""
 
-##############################
-#### Initialize variables ####
-##############################
+#######################################
+#### Initialize "global" variables ####
+#######################################
 
 if "text_error" not in st.session_state:
     st.session_state.text_error = ""
@@ -180,7 +183,7 @@ if "n_requests" not in st.session_state:
 ####### Page arguments #######
 ##############################
 
-# Every widget with a key is automatically added to Session State
+# Every widget with a key is automatically added to Session State as a global variable.
 
 # In Streamlit, interacting with a widget triggers a rerun and variables defined
 # in the code get reinitialized after each rerun.
@@ -198,7 +201,7 @@ st.text_input(
 st.text_input(
     label="Data to be redacted (optional)",
     placeholder="John, Mary, London",
-    key='metadata',
+    key='excluded_words',
     on_change=clear_results
 )
 st.text_input(