# ShopDB CSV Import Guide

Short version: one CSV per table, a header row of exact column names, one record
per row. The importer matches on natural keys, so re-importing updates instead of
duplicating.

## Rules for hand-built native CSVs

1. First row is the header. Use the exact column names below, lowercase, no spaces.
2. Include every REQUIRED column. Other columns are optional; drop the ones you
   do not use.
3. Foreign keys are entered by NAME, not id. Put the human name and the importer
   resolves it: `assettypeid=machine`, `statusid=In Use`, `locationid=A1`,
   `vendorid=Haas Automation`. The referenced record must already exist (import
   lookup tables first, see order below).
4. Booleans (`isactive`, etc.) are `1`/`0` or `true`/`false`. Leave blank for the
   default.
5. Dates are `YYYY-MM-DD`. Leave blank if unknown.
6. Leave a cell blank to skip it. Do not type `NULL` or `N/A`.
7. Save as UTF-8 CSV, comma-separated. ASCII only (no smart quotes, no em-dashes).

## Import order (dependencies first)

assetstatuses, assettypes, locationtypes, modeltypes, computertypes,
machinetypes, businessunits, locations, vendors, models, operatingsystems, assets,
computers, machines.

An asset row must exist before its `computers`/`machines` subtype row: the subtype
links back with `assetid` (the asset's `assetnumber`).

## Columns (REQUIRED in bold)

- assetstatuses: **status**, description, color, isactive
- locationtypes: **locationtype**, description, color, isactive
- modeltypes: **modeltype**, category, description, icon, isactive
- computertypes: **computertype**, description, icon, color, isactive
- machinetypes: **machinetype**, description, icon, color, isactive
- businessunits: **businessunit**, code, description, parentid, isactive
- locations: **locationname**, building, floor, room, description, locationtypeid,
  parentlocationid, isactive
- vendors: **vendor**, description, website, supportphone, supportemail, notes,
  isactive
- models: **modelnumber**, modeltypeid, vendorid, description, notes, isactive
- operatingsystems: **osname**, osversion, architecture, endoflife, isactive
- assets: **assetnumber**, **assettypeid**, name, serialnumber, gaugelabreference,
  maintenancereference, statusid, locationid, businessunitid, notes, isactive
- computers: **assetid**, computertypeid, hostname, osid, vendorid, modelnumberid,
  loggedinuser, isactive
- machines: **assetid**, machinetypeid, vendorid, modelnumberid, requiresmanualconfig,
  islocationonly, lastmaintenancedate, nextmaintenancedate, maintenanceintervaldays,
  controllervendorid, controllermodelid, isactive

Measuring tools are assets: add an `assets` row with `assettypeid=measuring_tool`.

## Example (assets.csv)

	assetnumber,name,serialnumber,assettypeid,statusid,locationid
	MILL-01,Haas VF-2 Mill,SN1001,machine,In Use,A1
	PC-01,Shopfloor PC 01,SN2001,computer,In Use,B1
	MIC-01,Mitutoyo Micrometer,SN3001,measuring_tool,In Use,Gage Lab

Then machines.csv links the machine subtype:

	assetid,machinetypeid,vendorid
	MILL-01,Mill,Haas Automation

## Azure and Thor exports (do NOT edit)

For Azure (Entra device-group) and Thor (PC inventory) files, upload the export
AS-IS. The importer auto-detects them by their header and extracts the fields it
needs. Do not rename headers, reorder columns, add/remove rows, or reformat cells.
Editing the export breaks auto-detection.