#!/usr/bin/env python3 import cgi, sys sys.path.insert(0, '/opt/ngon/apps') from managers.auth_manager import AuthManager, generate_login_page_html _form = cgi.FieldStorage() _auth = AuthManager('gen_service') _auth_required, _should_exit, _headers = _auth.require_auth(_form) if _should_exit: print("Content-Type: application/json") if _headers: print(_headers) print("") if _auth_required: print('{"success": false, "error": "Authentication required"}') else: print('{"success": true}') sys.exit(0) if _auth_required: print("Content-Type: text/html") print("Cache-Control: no-cache, no-store, must-revalidate") print("Expires: 0") print("") print(generate_login_page_html("Generator Service History")) sys.exit(0) import cgi import cgitb import sys sys.path.insert(0, '/var/www/html/ngon') from links import generate_dropdown_html, generate_dropdown_css, generate_dropdown_js cgitb.enable() _current_user = ''.join(c for c in (AuthManager.get_current_user() or '') if c.isalnum() or c in '_-.') # Service records change under the reader (imports, estimate status, QBO export # stamps), and with no directive a browser applies heuristic caching — WebKit # especially — so someone can be shown records that have since moved on. print("Content-Type: text/html") print("Cache-Control: no-cache, no-store, must-revalidate") print("Expires: 0") print("") print(""" NGON Generator Service History

NGON Mining - Generator Service History

""" + generate_dropdown_html(AuthManager.get_user_access()) + """
Browse / Search
Import
Analysis
PM / SC
Parts / Categories
QBO Export
Bills
Estimates
Query Builder
Invoice Date Gen ID Site Description Type Where * Category Group Amount Actions
+
Drop an Excel file here, or click to browse
Expected columns (any order): Invoice, Order Date, Unit Number, Description, $ Amount

Add Category

Category Rules

PriorityPatternMatch Type CategoryGrouped CategoryTypeActions

Cost by Site

Cost by Category

Monthly Trend

Top 20 Generators by Cost

Monthly Cost by Site

Spend by Work Type — shop pulled out of SC; partly derived

Mesa's export has no service_location column, so this split is inferred: an invoice that billed mileage was a field visit; one that billed transport, or labor with no travel, was shop work; one with neither is parts-only. Transport-billed is the only hard shop evidence — the “labor, no mileage” case also catches a field trip where travel was waived. Treat as a strong indicator, not ground truth.

Share of Spend

Monthly Spend — PM / SC / Shop / Parts-only

Work TypeInvoicesGens Total% of Spend
Site PM (field) SC (field) Shop Parts-only Total Shop Share

PM / SC / Shop Visits by Site per Month — shop is its own category, no longer counted as SC

SiteMonthPM VisitsSC VisitsShop Visits PM CostSC Cost Shop CostTotal Cost

Average PM / SC Cost per Generator by Site — field visits only; shop excluded

Avg Cost per Gen by Site

Avg Visits per Gen by Site

SitePM GensPM VisitsPM TotalPM Avg/Gen SC GensSC VisitsSC TotalSC Avg/Gen

Top 15 Categories by Cost

Category by Site

Top Category Trends by Month

2026 Spend by Group per Month — click a group to expand its sub-categories

Category Detail

CategoryGroupOccurrencesGens Affected Total CostAvg Cost/OccurrenceTop Site
+
Drop a Mesa repair-quote PDF here, or click to browse
Parsed on the server — you review every estimate before it is saved.
Mesa never puts the quote number on the invoice, so matches are inferred from parts overlap, total, and date on the same generator. High-confidence matches are marked completed; weaker ones are linked but left pending for you to confirm.
Pending
Completed
Declined
Superseded
All
Quote NoQuote DateGen IDSite ModelGrand Total StatusMatched InvoiceAgeActions

Paste a QuickBooks bill export to see service spend split by paid vs unpaid. Upload the file directly (.xlsx, .xls or .csv) or paste the rows — either way the columns are detected from the header, so both layouts work. The paid export carries a Bill No. and matches exactly; the unpaid export has none, so those can only be matched on amount, which is not always unique.

+
Drop an .xlsx, .xls or .csv export here, or click to browse
Paid vs unpaid is worked out from the columns — override it above if needed.

Spend by Work Type — Paid vs Unpaid

Work Type Paid billsPaid $ Unpaid billsUnpaid $ Outstanding Total

Monthly Spend by Category

Month comes from the matched invoice, not the bill — the unpaid export has no bill date, and its due dates ran 57–128 days past the work. Click a row to drill into that month.

Month ShopSC (field) PM (field)Parts-only Not matched PaidUnpaid Total

Bills

StatusMonthBill No.Bill DateDue AmountInvoiceGen Work TypeMatchDifference

Build a spreadsheet of generator data — one row per generator, columns pulled from any source. Pick columns, add filters, run, then export to CSV.

""")