Canonical Schema

Every extracted Unit is normalized into a 40+ field canonical schema. Fields are grouped into five categories.

Core Unit Fields (9)

Field	Type	Description
`property_name`	string	Property/community name
`property_address`	string \| null	Property street address
`unit_id`	string	Unit identifier (required)
`floor_plan`	string \| null	Floor plan name/code
`sqft`	integer \| null	Unit square footage
`beds`	integer \| null	Bedroom count
`baths`	float \| null	Bathroom count (supports half-baths)
`unit_status`	string	Normalized status (see below)
`tenant_name`	string \| null	Current tenant name

Status Normalization

Raw status values are normalized to one of 7 canonical values:

Canonical	Common Raw Values
`Occupied`	OCC, CURRENT, O, RENEWED
`Vacant`	VAC, VACANT, V, AVAILABLE, READY
`Notice`	NTV, NOTICE, N, NTC, ON NOTICE
`Model`	MODEL, MDL, M
`Down`	DOWN, DWN, D, OFFLINE, MAKE READY
`Applicant`	APP, APPLICANT, APPLIED, PENDING
`Unknown`	Anything unrecognized

Lease Fields (6)

Field	Type	Description
`move_in`	YYYY-MM-DD \| null	Move-in date
`move_out`	YYYY-MM-DD \| null	Move-out date
`lease_start`	YYYY-MM-DD \| null	Current lease start date
`lease_end`	YYYY-MM-DD \| null	Current lease end date
`lease_term_months`	integer \| null	Lease term in months
`is_mtm`	boolean \| null	Month-to-month flag

All dates are normalized to ISO YYYY-MM-DD format. The parser handles:

MM/DD/YYYY and M/D/YYYY
YYYY-MM-DD (passthrough)
DD-Mon-YY (e.g., 15-Jan-24)
Excel serial numbers (days since 1900-01-01)
Various separator characters (/, -, .)

Rent (1)

Field	Type	Description
`market_rent`	number \| null	Market/asking rent amount

Charge Categories (20)

All charge fields are number | null. Currency symbols and formatting are stripped. Parenthesized negatives like (1,250.00) are converted to -1250.

Field	Description
`charge_base_rent`	Monthly base rent
`charge_pet`	Pet rent/fee
`charge_parking`	Parking charges
`charge_storage`	Storage unit fees
`charge_utilities`	Utility charges (RUBS, flat fee)
`charge_trash`	Trash/valet trash
`charge_water_sewer`	Water and sewer charges
`charge_cable_internet`	Cable/internet fees
`charge_insurance`	Renter’s insurance
`charge_amenity`	Amenity/club fees
`charge_concession`	Concessions (usually negative)
`charge_employee_discount`	Employee discount
`charge_mtm_premium`	Month-to-month premium
`charge_furniture`	Furnished unit fee
`charge_garage`	Garage charges
`charge_laundry`	Laundry fees
`charge_late_fee`	Late payment fees
`charge_other`	Catch-all for unclassified charges
`charge_other_2`	Second catch-all
`charge_other_3`	Third catch-all

Balances & Totals (4)

Field	Type	Description
`total_charges`	number \| null	Sum of all charge fields
`balance_current`	number \| null	Current balance owed
`balance_prepaid`	number \| null	Prepaid amount
`deposit_amount`	number \| null	Security deposit amount

Output File Format

The Clean Output contains two Sheets:

Rent Roll — one row per Unit, columns in the order listed above
Validation Report — one row per validation finding (errors and warnings)

All extraction is literal — no inference, no imputation. If a field isn’t present in the source data, it’s null in the output.