Map a source to a canonical model

Line up a source’s fields with a shared model (Contact, Company, …) so data from many sources lines up. About 4 minutes.

You need:

• An API key in PYROSYNC_API_KEY and the deployment URL in PYROSYNC_URL (for example https://app.example.com).
• A canonical model id in MODEL_ID (from the catalog in the dashboard) and a source id in SOURCE_ID (for example from the CSV quickstart).

1. Ask for a proposal

POST /api/mappings/propose

curl -sS -X POST "$PYROSYNC_URL/api/mappings/propose" \
  -H "X-API-Key: $PYROSYNC_API_KEY" \
  -H "Content-Type: application/json" \
  -d "$(cat <<JSON
{
  "canonicalModelId": "${MODEL_ID}",
  "sourceId": "${SOURCE_ID}",
  "sourceFields": [
    {
      "name": "Email Address",
      "primitive": "string",
      "sampleValues": [
        "ada@example.com"
      ]
    },
    {
      "name": "Full Name",
      "primitive": "string",
      "sampleValues": [
        "Ada Lovelace"
      ]
    }
  ]
}
JSON
)"

Nothing is written. The response gives field matches, a score for each, and whether it reused an earlier mapping.

2. Save it as a draft mapping

POST /api/mappings

MAPPING_ID=$(curl -sS -X POST "$PYROSYNC_URL/api/mappings" \
  -H "X-API-Key: $PYROSYNC_API_KEY" \
  -H "Content-Type: application/json" \
  -d "$(cat <<JSON
{
  "key": "contacts_csv",
  "title": "Contacts CSV",
  "sourceId": "${SOURCE_ID}",
  "sourceKind": "file",
  "canonicalModelId": "${MODEL_ID}",
  "sourceFingerprint": "<fingerprint from the proposal>",
  "spec": "<spec from the proposal>"
}
JSON
)" | jq -r '.data.id')
echo "$MAPPING_ID"

3. Preview it against real rows

POST /api/mappings/{id}/preview

curl -sS -X POST "$PYROSYNC_URL/api/mappings/$MAPPING_ID/preview" \
  -H "X-API-Key: $PYROSYNC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "records": [
    {
      "Email Address": "ada@example.com",
      "Full Name": "Ada Lovelace"
    }
  ]
}'

Result: A draft mapping with a preview of the mapped records. Publishing stays a separate, reviewed step.