commit ac6763df1b6987f7b6b39a4d0ddf379f0633919a
Author: Sara Montecino <saramontecino@outlook.com>
Date:   Sun Oct 30 02:20:51 2022 -0700

    Initial commit

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..93526df
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+venv/
+__pycache__/
diff --git a/.vscode/launch.json b/.vscode/launch.json
new file mode 100755
index 0000000..9c519d0
--- /dev/null
+++ b/.vscode/launch.json
@@ -0,0 +1,17 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Main",
+            "type": "python",
+            "request": "launch",
+            "program": "main.py",
+            "console": "integratedTerminal",
+            "justMyCode": true,
+            "args": ["C:\\Users\\saram\\Downloads\\Statement_082022_4653.pdf"]
+        }
+    ]
+}
\ No newline at end of file
diff --git a/budget.db b/budget.db
new file mode 100755
index 0000000..1eeed46
Binary files /dev/null and b/budget.db differ
diff --git a/capital_one.py b/capital_one.py
new file mode 100755
index 0000000..513b28d
--- /dev/null
+++ b/capital_one.py
@@ -0,0 +1,101 @@
+import PyPDF2 as pdf
+import re
+import sys
+import math
+from datetime import datetime
+
+import model
+
+class Regex:
+    def error(self, msg):
+        print(f"ERROR: {self.__class__.__name__} failed: {msg}")
+        sys.exit(-1)
+    
+    def extract(self, text):
+        return None
+
+class Balance(Regex):
+    pattern = r'\nNew\sBalance.*\$([0-9,\.]*)'
+    
+    def extract(self, text):
+        results = re.findall(self.pattern, text)
+        if len(results) == 0:
+            self.error("No matches.")
+
+        if len(results) > 1:
+            self.error(f"Too many matches {len(results)}")
+
+        return float(results[0].replace(',',''))
+
+class Year(Regex):
+    pattern = r'\nAvailable\sCredit\s+\(as\sof\s(.*)\)'    
+
+    def extract(self, text):
+        results = re.findall(self.pattern, text)
+        if len(results) == 0:
+            self.error("No matches.")
+
+        if len(results) > 1:
+            self.error(f"Too many matches {len(results)}")
+
+        date = datetime.strptime(results[0], '%b %d, %Y')
+        return date.year
+
+class Transactions(Regex):
+    pattern = r'\n([a-zA-Z]{3}\s[0-9]{1,2})\s[a-zA-Z]{3}\s[0-9]{1,2}(.*[a-zA-Z])\s*(\-?)\s\$([0-9,\.]*)'
+
+    def extract(self, text):
+        results = re.findall(self.pattern, text)
+        transactions = []
+        for result in results:
+            if len(result) != 4:
+                self.error("ERROR: Invalid result.")
+
+            date = result[0]
+            description = " ".join(result[1].split())
+            is_payment = '-' in result[2]
+            amount = float(result[3].replace(',', ''))
+
+            if is_payment:
+                print(f"Skipping payment: {amount}")
+                continue
+            
+            transactions.append([date, description, amount])
+        
+        return transactions
+    
+class Parser(model.BaseParser):
+    def __init__(self):
+        self.balance = Balance()
+        self.year = Year()
+        self.transactions = Transactions()
+
+    @property
+    def source(self):
+        return model.TransactionSource.CAPITAL_ONE
+
+    def parse(self, file_name):
+        text = ''
+        with open(file_name, 'rb') as pdf_file:
+            reader = pdf.PdfReader(pdf_file)
+            for page in reader.pages:
+                text += page.extract_text()
+
+        balance = self.balance.extract(text)
+        year = self.year.extract(text)
+        transactions = self.transactions.extract(text)
+
+        # Validate transactions match extracted budget
+        total = sum(x[2] for x in transactions)
+        if not math.isclose(total, balance, abs_tol=0.001):
+            print(f"ERROR: Actual {total} != Expected {balance}")
+            sys.exit(-1)
+        
+        # Add year to all the parsed dates.
+        for transaction in transactions:
+            orig_date = transaction[0]
+            orig_date = datetime.strptime(orig_date, "%b %d")
+            new_date = orig_date.replace(year = year)
+            transaction[0] = new_date
+        
+        return transactions
\ No newline at end of file
diff --git a/database.py b/database.py
new file mode 100755
index 0000000..8496df9
--- /dev/null
+++ b/database.py
@@ -0,0 +1,69 @@
+import datetime
+from peewee import *
+
+DATABASE = 'budget.db'
+instance = SqliteDatabase(DATABASE, pragmas=[('foreign_keys', 'on')])
+
+def create_tables():
+    """Helper function to create database tables. Should be called manually."""
+    with instance:
+        instance.create_tables([User, TransactionCategory, Transaction, Source])
+    
+    # Make my user.
+    instance.connect()
+    User.create(username='ciphercules')
+    instance.close()
+
+def add_default_categories():
+    categories = [
+        {"food": ["snacks", "fast_food", "groceries", "restaurant"]},
+        {"clothing": []},
+        {"event": ["event_food", "birthday", "movie_theater"]},
+        {"finance": ["interest"]},
+        {"hobby": ["gym", "game_development", "projects", "education"]},
+        {"home_improvement": []},
+        {"pet": ["health", "dog_food"]},
+        {"media": ["book", "music", "television", "video_game"]},
+        {"health": ["medicine"]},
+        {"transit": ["car_insurance", "car_registration", "gas", "parking", "taxi", "car_maintenance"]},
+        {"utilities": ["electricity", "gas", "laundry", "cell_phone", "trash", "water"]},
+        {"rent": []},
+    ]
+
+    instance.connect()
+    with instance.atomic():
+        for parent, children in categories.items():
+            parent_db = TransactionCategory.create(name=parent)
+            for child in children:
+                TransactionCategory.create(name=child, parent=parent_db)
+
+class BaseModel(Model):
+    class Meta:
+        database = instance
+
+class User(BaseModel):
+    username = CharField(unique=True, primary_key=True)
+
+class TransactionCategory(BaseModel):
+    primary_key = AutoField(primary_key=True)
+    name = CharField(unique=True)
+    parent = ForeignKeyField('self', null=True, backref='children')
+
+class Source(BaseModel):
+    filename=CharField(unique=True, primary_key=True)
+    type = IntegerField()
+    created_date = DateTimeField(default=datetime.datetime.now)
+    user = ForeignKeyField(User, backref='transactions')
+
+class Transaction(BaseModel):
+    # Metadata
+    primary_key = AutoField(primary_key=True)
+    source = ForeignKeyField(Source, backref='transactions')
+    created_date = DateTimeField(default=datetime.datetime.now)
+    user = ForeignKeyField(User, backref='transactions')
+
+    # Real data
+    transaction_date = DateTimeField()
+    description = CharField()
+    amount = FloatField()
+    subcategory = ForeignKeyField(TransactionCategory, backref='+', null=True)
\ No newline at end of file
diff --git a/main.py b/main.py
new file mode 100755
index 0000000..9b28ebf
--- /dev/null
+++ b/main.py
@@ -0,0 +1,45 @@
+import sys
+import argparse
+import os
+
+import capital_one
+import database
+
+parser = argparse.ArgumentParser(prog="BudgetBear", description="Calculate a budget fit for a bear.")
+parser.add_argument('files', type=str, nargs='+', help='File to parse transactions from.')
+args = parser.parse_args()
+
+username = 'ciphercules'
+
+file_parsers = [capital_one.Parser()]
+for f in args.files:
+    for file_parser in file_parsers:
+        # Use the first successful parser.
+        transactions = file_parser.parse(f)
+        if not transactions:
+            continue
+
+        # Add to database
+        database.instance.connect()
+        with database.instance.atomic():
+            # Add source file first.
+            source = database.Source.create(
+                filename=os.path.basename(f),
+                type=file_parser.source,
+                user=username
+            )
+
+            # Add each transaction
+            for transaction in transactions:
+                date, description, amount = transaction
+                database.Transaction.create(
+                    user=username,
+                    transaction_date=date,
+                    description=description,
+                    amount=amount,
+                    source=source
+                )
+        database.instance.close()
+
+        # We successfully updated database with this parser, exit loop.
+        break
\ No newline at end of file
diff --git a/model.py b/model.py
new file mode 100755
index 0000000..fbf73f5
--- /dev/null
+++ b/model.py
@@ -0,0 +1,14 @@
+class BaseParser:
+    def __init__(self):
+        self.butt = None
+
+    def parse(self):
+        return None
+
+    @property
+    def source(self):
+        return -1
+
+class TransactionSource:
+    """Enum of possible transaction sources"""
+    CAPITAL_ONE = 1
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
new file mode 100755
index 0000000..b982e94
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,3 @@
+peewee==3.15.3
+PyPDF2==2.11.1
+typing_extensions==4.4.0
\ No newline at end of file
diff --git a/todo.txt b/todo.txt
new file mode 100755
index 0000000..940509e
--- /dev/null
+++ b/todo.txt
@@ -0,0 +1,14 @@
+# Goal: Command line tool for easily doing my budget
+Every week, I need to:
+  1. Categorize my purchases
+  2. Update SplitWise for shared purchases
+
+# Tasks
+
+## Parse capital one statement
+x Given a PDF bank statement from capital one, extract the transaction date, description, and amount from each transaction
+Store the data in a SQL lite database
+   x Create database schema
+   x Automatically add capital one transactions to database
+Add web page to add categories
+Add web page to categorize transactions
\ No newline at end of file