accounting thoughts

Accounting Program plan and pseucdocode anf flow


FLOW AND BRAINSTORM:

Initial thoughts:
Program opens a .txt document containing standard formatted set of business transactions and payroll information
It parses and ID's what they say and acts on them to create a set of financial documents and information for the desired circumstances,
	could be done as either XLS document or HTML wil links and coolness
	XLS:
		requires added package
		could be clearer
		language I know
		if borders possible, could look better
	HTML:
		links
		need to learn HTML
		probably wont be fancy at all
		probably better
Finally, it creates an encoded document with all the needed information to picj up where that cycle left off (acct. balences, etc)

More specific thoughts:

	THE GRAMMAR:
		my original thoughts as follows,
		initially flow indicators (don't do payroll, etc) as %:
			%payroll off
			%prior /prior.txt ##(This indicates prior doctument)
		then variable assignments as @
			@overtime 40hrs
			@overtimerate 2.0
			## These simply replace system variables immidiatly after creating them, changes in table of constants returned by parser
			 --- This means the parser should return a table of constants
		then business transations in following forms (as examples):
			month:July
			4:purchased $400.74 of trains (Transportation Equiptment):check 123
			5:sold $300 of kitchen appliances (Cooking Equiptment):receipt 561385
			6:purchased $19.99 of widgets (Widgets) on account (Egmire Equiptment):invoice 713
			10:paid $400.00 for water (Water Expense):check 94
			14:paid $300.99 to Egmire Equiptment for account:check 843

			Anexample of the regex for these:

			b'(\d+)\:[Pp]urchased\s+\$(\d+\.\d{2})\s+of?[\s\w]+\(([\s\w]+)\)\:([\w\s\d]+)' ##the first one

			DO as variable in loop, ex

			loop = [[regex 1,needed vars],
					[regex 2,vars],
					[regex 3,vars]]

			for line in (lines) - or can use file object:
				for i in loop:
					reobj = search(i[0],line)
					if reobj:
						blahblahblah
						break
				else:
					continue (or something along these lines)
		And thats what I have there

	NEXT: PSEUDO / FLOW
		program should first open and return text of docukment specified by first argument
		function opentext(takes first argument as argument):
			return text of document
		function parse(takes text as argument)
			/*This function wioll do the folowing;
				it will go through program as specified about and compile a list of list containing all the needed information about the transcations
				and return that as a tuple with all the necessary information pertinant to payroll
			*/

			CODE for subfunction (unless re.group has a way to output all as array / tuple)
				reobj = regex object
				index = 1
				output = []
				excep = False
				while excep = False:
					try:
						output.append(reobj.group(index))
					except IndexError:
						excep = True
					index += 1
				return output

			return transcations_p,payroll_p
		function accounts(takes transactions array as argument):
			/*This function will go through the transactions and break them up int o accounts and track the account balences;
				it will output in the following form;
					{account name:[[date,dr,cr],[date,dr,cr],...,account name:[date,dr,cr],...}
					### --- (unless changed)
				this is ripe for the general ledger
			*/
		#Now thought - I need to be able to translate individua l transactions to the leder direct fro the journal - perhaps do this differentyl
		#perhaps hash the transaction arrays as an ID and push them through like that
		#or perhaps make a transation class and make them objects that can stre dates etc ---I like this