diff --git a/TPC4/alunos.csv b/TPC4/alunos.csv
new file mode 100644
index 0000000..e88adf6
--- /dev/null
+++ b/TPC4/alunos.csv
@@ -0,0 +1,4 @@
+Número,Nome,Curso,Notas{3,5}::sum,,,,,
+3162,Cândido Faísca,Teatro,12,13,14,,
+7777,Cristiano Ronaldo,Desporto,17,12,20,11,12
+264,Marcelo Sousa,Ciência Política,18,19,19,20,
diff --git a/TPC4/alunos.json b/TPC4/alunos.json
new file mode 100644
index 0000000..894748b
--- /dev/null
+++ b/TPC4/alunos.json
@@ -0,0 +1,20 @@
+[
+  {
+    "Número": "3162",
+    "Nome": "Cândido Faísca",
+    "Curso": "Teatro",
+    "Notas_sum": 39
+  },
+  {
+    "Número": "7777",
+    "Nome": "Cristiano Ronaldo",
+    "Curso": "Desporto",
+    "Notas_sum": 72
+  },
+  {
+    "Número": "264",
+    "Nome": "Marcelo Sousa",
+    "Curso": "Ciência Política",
+    "Notas_sum": 76
+  }
+]
\ No newline at end of file
diff --git a/TPC4/main.py b/TPC4/main.py
new file mode 100644
index 0000000..9c4b10d
--- /dev/null
+++ b/TPC4/main.py
@@ -0,0 +1,134 @@
+from dataclasses import dataclass
+import json
+import re
+
+
+@dataclass
+class Field:
+    name: str
+
+
+@dataclass
+class ListField(Field):
+    min_size: int
+    max_size: int
+    associated_function: str | None
+
+
+valid_functions = ["sum", "media", "min", "max"]
+
+
+def main():
+    headers = read_header("alunos.csv")
+    read_to_json("alunos.csv", headers)
+
+# Create a function that reads the first line of the csv file and returns a list of fields
+def read_header(file: str) -> list[Field]:
+    with open(file, "r",encoding="utf-8") as f:
+        header = f.readline()
+        # Split the header into fields by comma but ignore commas inside curly brackets (list fields)
+        fields = re.split(r",(?![^{]*})", header)
+        # Regex to parse the fields
+        regex = r"(\w+)(?:{(\d+)(?:,(\d+))?}(?:::(\w+))?)?"
+        # List to store the parsed fields
+        parsed_headers: list[Field] = []
+        for field in fields:
+            # Remove whitespace
+            field = field.strip()
+            # Check if field is empty
+            if field == "":
+                continue
+            # apply regex to field
+            match = re.match(regex, field)
+            if match is None:
+                raise ValueError(f"Invalid field: {field}")
+            name = match.group(1)
+            min_size = match.group(2)
+            max_size = match.group(3)
+            function = match.group(4)
+            # First check if the field is a list
+            if min_size is None:
+                parsed_headers.append(Field(name))
+            else:
+                # Check if the list has a variable size
+                if max_size is None:
+                    max_size = min_size
+                # Check if the list has a valid function
+                if function not in valid_functions:
+                    raise ValueError(f"Invalid function: {function}")
+                parsed_headers.append(
+                    ListField(name, int(min_size), int(max_size), function)
+                )
+        return parsed_headers
+
+
+def read_to_json(file: str, headers: list[Field]):
+    with open(file, "r",encoding="utf-8") as f:
+        # Read the lines of the file
+        lines = f.readlines()
+        # Remove the header
+        lines.pop(0)
+        # Create the output Json List
+        output = []
+        # Iterate over the lines
+        for line in lines:
+            # Iterate over the header fields
+            values = line.split(",")
+            output_line = {}
+            for header in headers:
+                # If the field represents a list , we will have N values separated by comma, if the list is a variable size we will have at least the minimum size with the rest being empty
+                # If the field represents a list with a function we should apply the function to the list and add it to the json
+                # If the field does not represent a list we should just add it to the json
+                # Check header type and read the next x values depending on the header type
+                # Dont forget to remove the values from the line after using them
+                if isinstance(header, ListField):
+                    # Read the next max_size values
+                    list_values = values[: header.max_size]
+                    # Remove the empty or new line strings from list
+                    list_values = [x.strip() for x in list_values if x.strip() != ""]
+                    # Remove the values from the line
+                    values = values[header.max_size :]
+                    # Check if the list has the minimum size
+                    if len(list_values) < header.min_size:
+                        raise ValueError(f"Invalid list size: {header.name}")
+                    # Check if the list has a function
+                    if header.associated_function is not None:
+                        # Apply the function to the list
+                        function_result = apply_function(
+                            header.associated_function, list_values
+                        )
+                        # Add the function result to the json
+                        output_line[
+                            header.name + "_" + header.associated_function
+                        ] = function_result
+                    else:
+                        # Add the list to the json
+                        output_line[header.name] = list_values
+                else:
+                    # Read the next value
+                    value = values.pop(0)
+                    # Add the value to the json
+                    output_line[header.name] = value
+            # Add the line to the json
+            output.append(output_line)
+    #Write pretty json to file
+    with open("alunos.json", "w",encoding="utf-8") as f:
+        json.dump(output, f, indent=2,ensure_ascii=False)
+
+def apply_function(fn: str, values: list[str]):
+    # convert values to int
+    vals = [int(value) for value in values]
+    # match case to apply the correct function
+    match fn:
+        case "sum":
+            return sum(vals)
+        case "media":
+            return sum(vals) / len(vals)
+        case "min":
+            return min(vals)
+        case "max":
+            return max(vals)
+
+
+if __name__ == "__main__":
+    main()