Create app.py

app.py

@@ -0,0 +1,143 @@
+import spacy 
+import re
+
+from datetime import datetime
+
+# load the spacy model
+spacy.cli.download("en_core_web_lg")
+nlp = spacy.load("en_core_web_lg")
+
+# Define a function to extract dates from text
+def extract_dates(text):
+    """
+    Identify dates both in numeric and free-text from text, using date regex patterns and NER tag
+    """
+
+    # Define regex patterns for common date formats
+    # Regular expressions that include the \b word boundary character to ensure that the date pattern only matches if it is not part of a longer pattern that has already been matched
+    date_patterns = [
+        r'\b\d{1,2}[-/]\d{1,2}[-/]\d{2,4}\b',  # Matches dates like "01/01/22" or "1-1-2022"
+        r'\b\d{1,2}[-/]\d{1,2}\b(?!\d)',       # Matches dates like "01/01" or "1-1", but not "10-10-2022"
+        r'\b[A-Z][a-z]{2,8} \d{1,2},? \d{2,4}\b',  # Matches dates like "January 1, 2022" or "Feb 28, 22"
+        r'\b\d{1,2} [A-Z][a-z]{2,8} \d{2,4}\b',  # Matches dates like "1 January 2022" or "28 Feb 22"
+        r'\b[A-Z][a-z]{2,8} \d{2,4}\b',        # Matches dates like "January 2022" or "Feb 22"
+    ]
+
+    # Find all matches for date patterns in the text
+    matches = []
+    for pattern in date_patterns:
+        for match in re.findall(pattern, text):
+
+            # Check if the match is part of a longer date pattern that has already been matched
+            if all(match not in m for m in matches):
+                matches.append(match)
+    
+    # Use SpaCy to extract additional dates
+    doc = nlp(text)
+
+    for ent in doc.ents:
+        if ent.label_ == 'DATE':
+            date_str = ent.text
+
+            # Checks each SpaCy date reference against the matches list to ensure that it is not already included
+            if all(date_str not in m for m in matches):
+                matches.append(date_str)
+    
+    # Remove duplicates and return the matches
+    return list(set(matches))
+
+
+def convert_dates(date_list):
+    """
+    Assign to the identified formatted dates the proper date format and then, on the formatted dates, assign the relevant date tags (e.g. specify which is the day, the month, etc)
+    """
+    
+    output_list = []
+    for date_str in date_list:
+        try:
+            dt = datetime.strptime(date_str, '%B %d, %Y')
+            output_list.append(f"day:{dt.day}, month:{dt.month}, year:{dt.year}")
+        except ValueError:
+    
+            try:
+                dt = datetime.strptime(date_str, '%m-%d-%y')
+                output_list.append(f"day:{dt.day}, month:{dt.month}, year:{dt.year}")
+            except ValueError:
+    
+                try:
+                    dt = datetime.strptime(date_str, '%d/%m')
+                    output_list.append(f"day:{dt.day}, month:{dt.month}")
+                except ValueError:
+    
+                    try:
+                        dt = datetime.strptime(date_str, '%B %d')
+                        output_list.append(f"day:{dt.day}, month:{dt.month}")
+                    except ValueError:
+    
+                        try:
+                            dt = datetime.strptime(date_str, '%b %d')
+                            output_list.append(f"day:{dt.day}, month:{dt.month}")
+                        except ValueError:
+    
+                            try:
+                                dt = datetime.strptime(date_str, '%B %Y')
+                                output_list.append(f"month:{dt.month}, year:{dt.year}")
+                            except ValueError:
+    
+                                try:
+                                    dt = datetime.strptime(date_str, '%Y')
+                                    output_list.append(f"year:{dt.year}")
+                                except ValueError:
+                                    
+                                    try:
+                                        dt = datetime.strptime(date_str, '%d/%m/%y')
+                                        output_list.append(f"day:{dt.day}, month:{dt.month}, year:{dt.year}")
+                                    except ValueError:
+                                    
+                                      try:
+                                        dt = datetime.strptime(date_str, '%Y %d %m')
+                                        output_list.append({'day': dt.day, 'month': dt.month, 'year': dt.year})
+                                      except ValueError:
+                                    
+                                        try:
+                                            dt = datetime.strptime(date_str, '%B')
+                                            output_list.append(dt.strftime('month:%-m'))
+                                        except ValueError as e:
+                                            output_list.append(f'INVALID FORMAT: {date_str}')
+    return output_list
+
+
+
+def dates_binding(text):
+
+ identified_dates = extract_dates(text)
+ formatted_dates = convert_dates(identified_dates)
+
+ return formatted_dates
+
+
+from transformers import pipeline
+import gradio as gr
+
+title = "Natural Language module Demo for Dates"
+description = "This is a simple demo just for demonstration purposes, so that Serco team might have the chance to validate the results of the Natural Language module concerning date identification, while in progress"
+
+examples = [
+    ["Earthquake located in Ishkoshim, Tajikistan in 2018 05 09 with magnitude greater than 6.2"],
+    ["Give me all the earthquakes with magnitude above than 6 in the region of Athens for the month of January 2, 2023"],
+    ["Earthquake happened in Rome during 1986 with a magnitude of 6"],
+    ["I want all earthquakes larger than 5.0 that occurred in Rome during 3/5/20"],
+    ["I want all earthquakes happened in Greece 31/12"],
+    ["I want all earthquakes happened in Greece January 5"],
+    ["I want all earthquakes happened in Greece 3-4-20"]
+]
+
+gr.Interface(
+    fn=dates_binding, 
+    inputs="text", 
+    outputs="text",
+    title=title,
+    description=description,
+    examples=examples,
+    enable_queue=True,
+    ).launch()