Since all current discounts have been implemented as processors,

the `DiscountInvoice` class is now removed. No invoice class will
now class `_prepare()` or `_process()`.

`BUSubsidyProcessor`, which handles processing for the BU subsidy,
sets `IS_DISCOUNT_BY_NERC` to `False` because the
subsidy is not provided by NERC. Because of this, `BU Balance` indicates
the money which BU (not the PI they are subsidizing) owes to the
MGHPCC.

The test cases for the BU Subsidy has been refactored to be
more robust and readable.
The `Project` field has been added to `invoice.py`

process_report/invoices/bu_internal_invoice.py

@@ -1,12 +1,10 @@
 from dataclasses import dataclass
-from decimal import Decimal
 
 import process_report.invoices.invoice as invoice
-import process_report.invoices.discount_invoice as discount_invoice
 
 
 @dataclass
-class BUInternalInvoice(discount_invoice.DiscountInvoice):
+class BUInternalInvoice(invoice.Invoice):
     """
     This invoice operates on data processed by these Processors:
     - ValidateBillablePIsProcessor
@@ -20,39 +18,32 @@ class BUInternalInvoice(discount_invoice.DiscountInvoice):
         invoice.COST_FIELD,
         invoice.CREDIT_FIELD,
         invoice.SUBSIDY_FIELD,
-        invoice.BALANCE_FIELD,
+        invoice.PI_BALANCE_FIELD,
     ]
 
-    subsidy_amount: int
-
-    def _prepare(self):
-        def get_project(row):
-            project_alloc = row[invoice.PROJECT_FIELD]
-            if project_alloc.rfind("-") == -1:
-                return project_alloc
-            else:
-                return project_alloc[: project_alloc.rfind("-")]
+    exported_columns_map = {invoice.PI_BALANCE_FIELD: "Balance"}
 
+    def _prepare_export(self):
         self.data = self.data[
             self.data[invoice.IS_BILLABLE_FIELD] & ~self.data[invoice.MISSING_PI_FIELD]
         ]
         self.data = self.data[
             self.data[invoice.INSTITUTION_FIELD] == "Boston University"
-        ].copy()
-        self.data["Project"] = self.data.apply(get_project, axis=1)
-        self.data[invoice.SUBSIDY_FIELD] = Decimal(0)
-
-    def _process(self):
-        data_summed_projects = self._sum_project_allocations(self.data)
-        self.data = self._apply_subsidy(data_summed_projects, self.subsidy_amount)
+        ]
+        self.data = self._sum_project_allocations(self.data)
 
     def _sum_project_allocations(self, dataframe):
         """A project may have multiple allocations, and therefore multiple rows
         in the raw invoices. For BU-Internal invoice, we only want 1 row for
         each unique project, summing up its allocations' costs"""
         project_list = dataframe["Project"].unique()
         data_no_dup = dataframe.drop_duplicates("Project", inplace=False)
-        sum_fields = [invoice.COST_FIELD, invoice.CREDIT_FIELD, invoice.BALANCE_FIELD]
+        sum_fields = [
+            invoice.COST_FIELD,
+            invoice.CREDIT_FIELD,
+            invoice.SUBSIDY_FIELD,
+            invoice.PI_BALANCE_FIELD,
+        ]
         for project in project_list:
             project_mask = dataframe["Project"] == project
             no_dup_project_mask = data_no_dup["Project"] == project
@@ -61,18 +52,3 @@ def _sum_project_allocations(self, dataframe):
             data_no_dup.loc[no_dup_project_mask, sum_fields] = sum_fields_sums
 
         return data_no_dup
-
-    def _apply_subsidy(self, dataframe, subsidy_amount):
-        pi_list = dataframe[invoice.PI_FIELD].unique()
-
-        for pi in pi_list:
-            pi_projects = dataframe[dataframe[invoice.PI_FIELD] == pi]
-            self.apply_flat_discount(
-                dataframe,
-                pi_projects,
-                subsidy_amount,
-                invoice.SUBSIDY_FIELD,
-                invoice.BALANCE_FIELD,
-            )
-
-        return dataframe

process_report/invoices/invoice.py

@@ -37,6 +37,7 @@
 IS_BILLABLE_FIELD = "Is Billable"
 MISSING_PI_FIELD = "Missing PI"
 PI_BALANCE_FIELD = "PI Balance"
+PROJECT_NAME_FIELD = "Project"
 ###
 
 

process_report/process_report.py

@@ -21,6 +21,7 @@
     lenovo_processor,
     validate_billable_pi_processor,
     new_pi_credit_processor,
+    bu_subsidy_processor,
 )
 
 ### PI file field names
@@ -242,7 +243,12 @@ def main():
     )
     new_pi_credit_proc.process()
 
-    processed_data = new_pi_credit_proc.data
+    bu_subsidy_proc = bu_subsidy_processor.BUSubsidyProcessor(
+        "", invoice_month, new_pi_credit_proc.data.copy(), args.BU_subsidy_amount
+    )
+    bu_subsidy_proc.process()
+
+    processed_data = bu_subsidy_proc.data
 
     ### Initialize invoices
 
@@ -280,7 +286,6 @@ def main():
         name=args.BU_invoice_file,
         invoice_month=invoice_month,
         data=processed_data.copy(),
-        subsidy_amount=args.BU_subsidy_amount,
     )
 
     pi_inv = pi_specific_invoice.PIInvoice(

process_report/processors/bu_subsidy_processor.py

@@ -7,6 +7,8 @@
 
 @dataclass
 class BUSubsidyProcessor(discount_processor.DiscountProcessor):
+    IS_DISCOUNT_BY_NERC = False
+
     subsidy_amount: int
 
     def _prepare(self):
@@ -17,43 +19,35 @@ def get_project(row):
             else:
                 return project_alloc[: project_alloc.rfind("-")]
 
-        self.data = self.data[
-            self.data[invoice.IS_BILLABLE_FIELD] & ~self.data[invoice.MISSING_PI_FIELD]
-        ]
-        self.data = self.data[
-            self.data[invoice.INSTITUTION_FIELD] == "Boston University"
-        ].copy()
-        self.data["Project"] = self.data.apply(get_project, axis=1)
+        self.data[invoice.PROJECT_NAME_FIELD] = self.data.apply(get_project, axis=1)
         self.data[invoice.SUBSIDY_FIELD] = Decimal(0)
 
     def _process(self):
-        data_summed_projects = self._sum_project_allocations(self.data)
-        self.data = self._apply_subsidy(data_summed_projects, self.subsidy_amount)
+        self.data = self._apply_subsidy(self.data, self.subsidy_amount)
 
-    def _sum_project_allocations(self, dataframe):
-        """A project may have multiple allocations, and therefore multiple rows
-        in the raw invoices. For BU-Internal invoice, we only want 1 row for
-        each unique project, summing up its allocations' costs"""
-        project_list = dataframe["Project"].unique()
-        data_no_dup = dataframe.drop_duplicates("Project", inplace=False)
-        sum_fields = [invoice.COST_FIELD, invoice.CREDIT_FIELD, invoice.BALANCE_FIELD]
-        for project in project_list:
-            project_mask = dataframe["Project"] == project
-            no_dup_project_mask = data_no_dup["Project"] == project
-
-            sum_fields_sums = dataframe[project_mask][sum_fields].sum().values
-            data_no_dup.loc[no_dup_project_mask, sum_fields] = sum_fields_sums
+    @staticmethod
+    def _get_subsidy_eligible_projects(data):
+        filtered_data = data[
+            data[invoice.IS_BILLABLE_FIELD] & ~data[invoice.MISSING_PI_FIELD]
+        ]
+        filtered_data = filtered_data[
+            filtered_data[invoice.INSTITUTION_FIELD] == "Boston University"
+        ].copy()
 
-        return data_no_dup
+        return filtered_data
 
     def _apply_subsidy(self, dataframe, subsidy_amount):
-        pi_list = dataframe[invoice.PI_FIELD].unique()
+        subsidy_eligible_projects = self._get_subsidy_eligible_projects(dataframe)
+        pi_list = subsidy_eligible_projects[invoice.PI_FIELD].unique()
 
         for pi in pi_list:
-            pi_projects = dataframe[dataframe[invoice.PI_FIELD] == pi]
+            pi_projects = subsidy_eligible_projects[
+                subsidy_eligible_projects[invoice.PI_FIELD] == pi
+            ]
             self.apply_flat_discount(
                 dataframe,
                 pi_projects,
+                invoice.PI_BALANCE_FIELD,
                 subsidy_amount,
                 invoice.SUBSIDY_FIELD,
                 invoice.BALANCE_FIELD,