Updates `categorize_data` to work with the `Transaction` class, which is now
used by the database.
Vacation categorizing will no longer replace previous category assigned.
`DBManager` gained new method to retrieve transactions between two dates
without any category.
Add method to `Transaction` to prepare for category update in database.
Transaction's returns from database queries are checked and assigned in
`if`s with the walrus (`:=`) operator.
Fix parser func arguments. The database should be passed as an argument,
not used before parsing.
Reordered and added logging to `DBManager` for better recording.
`parse_data` from parser.py now takes a `DBManager` and runs
`insert_transactions` after parsing.
`DBManager` now takes `Transaction` type and returns the same on select
queries.
`DBManager` refactored with extensive type hinting.
`Transaction` updated to reflect use in `DBManager`.
`categorize_data` is the new entry for data categorization and receives
a DBManager. Contains the categorizing logic.
Categorizer configuration now done solely from categories.yaml file.
Ancilliary database methods added to DBManager required for categorizing
transactions.
Adds categorize to command line options.
Removes obsolete restart options and method from runnable.py.
Fixes parse and categorize method, now take cmd line arguments and
`DBManager`.
Removes obsolete tools.py, all functions already rewritten in relevant
modules.
Updated categories.yaml with new keys.
parsers.py redid with single default parser that takes configurating
parameters from a yaml file.
Additional parsing configuration can be achieved with the additional_parser
attribute on yaml and rewriting the func and parser method on child
classes of Parser.
func will be called after each transaction is created and the parser
should call the parent parser method or rewrite the entire parser
process.
The parse_data function is now called from the runnable and the parsing
process is now called from there. The parse command can take an optional
bank before is tries to extract it from the filename and multiple paths,
either files or directories.
The Transaction __init__ was fixed to take inputs from previously
initiated Transaction.
Also adds utils.py with helper functions.
Removed init and restart functions and writes to state file.
Moved create tables to init method of DBManager and launch from init
command.
Restart removed.
Rearranged order in runnable.py according to sections.
Only init command working since rest of runnable function not updated
at the moment.
Moves argparser functions to runnable.py and PFState to state.py
Main function will now only call run from runnable.py. Should make it
easier to run additional functions. Program can now be run from main.py,
or imported from pfbudget.
Added multiple Path.mkdir() when .pfstate, backup, raw and data
directories weren't created.
Adds try-except clause around functions called in init, so that errors
in the early stages are catched to remove already initialized state
files.
Moves filename definitions to tools using global vars.
Removed state.filename = p from main.py introduced when the state
filename was changed.
`self._save()` removed from `PFState` `__init__` as it's not needed. All
setattr already call it.
Two new graphs created, discrete, where all discretionary expenses are
plotted, and average, for the yearly average. This later one will be
moved to report form.
Health expenses separated from required. Commute moved from fixed to
required expenses.
New by_month_and_category functions added to transactions.py.
Moves all graph and matplotlib dependencies to graph.py.
Moves Transactions and Categories methods outside the class for better
visibility.
Variable expenses renamed to required.
transactions.py now has by_year, by_month, by_categories in global scope
and slightly improved perfomance.
Add new tools.py to handle subcommands.
Most workflow moved to subcommands handlers, which in turn calls
tools.py functions.
Moves internal state (dict which gets parsed into PFState) to folder
.pfstate (sort of like git) and renamed to state.
New PFState class and factory pfstate function added.
*.pickle changed to .pfstate on .gitignore. Everything local stored in
.pfstate.
Adds vacation, status and backup commands. Modifies init, restart and
parse to receive PFState and call tools functions.
Vacation can add, list or remove vacations date intervals.
Backup saves either a full transaction list or the entire data folder
into .pfstate/backup folder.
Categorization functions moves to tools, deleted old one on
categories.py.
Moves parse_data and load_transactions/save_trasactions outside of
respective class to global scope.
Travel category class moved to the end of file, since it uses others in
its search method.
Removes unused initializer.py file.
The initialize function has been updated with a new possible option for
a new raw file, where the Last Accessed date changed the content hadn't.
Also removes sort() when using extend() on a list as it doesn't work
correctly.
Categorization code moved to if False statement to add posterior toggle
behaviour.
New transaction print added (.desc()) which prints the description of a
transaction instead of its category to help on the manual
categorization.
Raw files now on raw dir instead of .raw
It now also accepts empty lines and # at the start for comments on data
files. TODO these are overwritten whenever the data files are updated.
It separates the monthly transactions by categories and tracks income
and spending along the defined time range.
Now prints the total income/spending and the net result for the define
period of time.
Plot code is surrounded with an if False for future toggle on/off.
New `Transactions` class derived from `list` to add auxilliary methods to a
transaction list.
New get_transactions_by_(year,month,category) methods return dict sorted
by (year,month,category).
Also adds sort_by_bank inplace.
Get methods added for income, fixed expenses, variable expenses and
discretionary expenses.
Search has regex check to detect placeholder category. Removes the need
for an additional search.
reader.py is removed, with main.py becoming the main function where all
others are called from.
All categories added on categories.py, each with its own search function
and regex. The categorize function is also defined on the base class
The parsers have been cleaned to a more strimmed version. Each parser has
its own parse method and encoding and separator attributes.
The `Transaction` class has been incremented with all the comparator
methods. It also has the read and write from file methods.
The transactions.pickle is no longer used, since changes can be done
directly to the parsed data in the data_dir, making the file unused. A
manual categorization function has been created to help fill in the
gaps, interchangeable with directly editing the parsed .csv files. The
master record of data are the .csv present in the data_dir.
Working csv reader, which reorders the transactions on transaction.csv
based on date. Then it aggregates the information per month based on the
used categories, and prints that information.
