Target Adapters

A Sluice target adapter loads transformed records to the system that will own the data going forward. The open-source core ships with two built-in target adapters — csv and pg — that cover generic flat-file output and PostgreSQL bulk insert. ERP-specific adapters (bc, ifs, bluecherry) are paid add-ons from Caracal Lynx; their config is documented here so the same YAML works whether or not you have the add-on installed.

Common config

Every target has the same adapter field plus adapter-specific keys. The adapter value is one of:

csv | pg | bc | ifs | bluecherry | rest

For ERP adapters, see Commercial Support.

csv — Generic CSV output (built-in)

The simplest target. Writes the transformed staging table to a CSV file. Use this when you need a portable, human-readable load file or a CI artefact.

target:
  adapter: csv
  output: ./output/customers.csv
  includeHeader: true
  delimiter: ","
  encoding: utf-8
  nullValue: ""

Key	Type	Default	Notes
output	string	—	Path to write.
includeHeader	boolean	true	Write column names on the first row.
delimiter	string	,	Field delimiter.
encoding	string	utf-8	File encoding.
nullValue	string	""	How nulls are rendered.

pg — PostgreSQL bulk insert (built-in)

Loads the staging table into a PostgreSQL table. Supports insert, ignore-on-conflict, and upsert modes.

target:
  adapter: pg
  connection: ${TARGET_PG}
  table: customers
  schema: public
  onConflict: upsert
  upsertKey: [customer_no]

Key	Type	Default	Notes
connection	string	—	PostgreSQL connection string. Resolved from ${ENV_VAR} at runtime.
table	string	—	Target table name.
schema	string	public	Target schema.
onConflict	enum	fail	fail | upsert | ignore.
upsertKey	string[]	—	Required when onConflict: upsert. Column names that form the conflict key.

ifs — IFS ERP CSV import (paid add-on)

IFS imports CSV with no header row, in a column order defined by the IFS entity import contract.

target:
  adapter: ifs
  entity: CustomerInfo
  output: ./output/acme-corp-customers-ifs.csv
  includeHeader: false
  columnOrder:
    - CustomerNo
    - Name
    - Address1
    - ZipCode
    - Country
  dateFormat: YYYY-MM-DD

Key	Type	Default	Notes
entity	string	—	IFS entity name (e.g. CustomerInfo, Supplier). Recorded in the load log.
output	string	—	Path to write.
includeHeader	boolean	false	IFS expects no header.
columnOrder	string[]	—	Forces the column order. Must match transform to: field names.
dateFormat	string	YYYY-MM-DD	dayjs format token.
delimiter	string	,	Field delimiter.
encoding	string	utf-8	File encoding.

Notes:

• Date columns are formatted with dayjs using dateFormat.
• The exact column order is critical — IFS imports by position, not by name.

bluecherry — BlueCherry ERP CSV import (paid add-on)

BlueCherry (CGS) uses fixed-format CSV imports per entity type. The adapter validates required columns at connect() time, before any data is read, so misconfigured pipelines fail fast.

target:
  adapter: bluecherry
  entity: Style
  output: ./output/style-co-styles-bc.csv
  template: default
  includeHeader: true
  dateFormat: MM/DD/YYYY
  nullValue: ""

Key	Type	Default	Notes
entity	enum	—	One of Style, Vendor, PurchaseOrder, PODetail, Season, ColourSize.
output	string	—	Path to write.
template	string	default	default uses built-in column order; or path to a header-only template CSV.
includeHeader	boolean	true	BlueCherry expects headers.
dateFormat	string	MM/DD/YYYY	BlueCherry is US-origin; default is US date format.
nullValue	string	""	How nulls are rendered.

Required columns per entity (validated at connect()):

Entity	Required columns
Style	StyleNo, StyleDesc, Division, Season, CostPrice, RetailPrice, ActiveFlag
Vendor	VendorNo, VendorName, Country, CurrencyCode
PurchaseOrder	PONumber, VendorNo, Season, OrderDate, DeliveryDate
PODetail	PONumber, StyleNo, ColourCode, SizeCode, Quantity, CostPrice
Season	SeasonCode, SeasonDesc, StartDate, EndDate
ColourSize	StyleNo, ColourCode, ColourDesc, SizeCode, SizeDesc

Notes:

• Any column whose name ends with Date (case-insensitive) is automatically formatted using dateFormat via dayjs.
• If your BlueCherry instance uses different column names, point template at a header-only template CSV — the adapter uses that as the definitive column order.
• These column names are conventions; verify against your BlueCherry import documentation before running a live migration.

bc — Business Central REST (paid add-on)

Loads to Microsoft Dynamics 365 Business Central via OData REST. Uses OAuth 2.0 client credentials for authentication and OData $batch for throughput.

target:
  adapter: bc
  baseUrl: ${BC_BASE_URL}
  company: ${BC_COMPANY}
  entity: customers
  apiVersion: v2.0
  onConflict: upsert
  batchEndpoint: true

Key	Type	Default	Notes
baseUrl	string	—	Resolved from ${ENV_VAR}.
company	string	—	BC company GUID or name.
entity	string	—	OData entity name (lowercase, plural — e.g. customers, salesInvoices).
apiVersion	string	v2.0	OData API version.
onConflict	enum	fail	fail | upsert | ignore. With upsert, a 409 triggers a PATCH.
batchEndpoint	boolean	true	Use OData $batch (max 100 ops per batch).

Notes:

• OAuth 2.0 client credentials: POST https://login.microsoftonline.com/{tenantId}/oauth2/v2.0/token. Tokens are cached in memory and refreshed 60 seconds before expiry.
• $batch endpoint: POST {baseUrl}/api/{apiVersion}/companies({company})/$batch with Content-Type: multipart/mixed; boundary=batch_{uuid}.
• 4xx errors (other than 409 on upsert) are logged and rowsFailed is incremented; the run continues if run.onError: continue.

Custom target adapters

For systems Sluice doesn’t cover, implement the TargetAdapter interface as a Tier 2 file plugin or a Tier 3 npm package — see the Plugin System guide.