hosting
POST /v1/site-imports
Restore a site from a web archive, or migrate one in.
Authentifizierung
Senden Sie einen API-Schlüssel als Bearer-Token. Der Schlüssel muss über die Berechtigung hosting.import.manage verfügen; ein Schlüssel ohne diese wird mit 403 statt 404 abgelehnt.
Dieser Endpunkt erfordert keine Organisations-ID. Ihr Schlüssel identifiziert bereits die zugehörige Organisation, und die Antwort ist entsprechend eingeschränkt.
Ausprobieren
Ersetzen Sie alles in spitzen Klammern durch Ihre eigenen Werte und den Platzhalter für den Schlüssel durch einen Schlüssel aus Ihrem Dashboard.
curl -X POST https://api.zinndigital.com/v1/site-imports \
-H "Authorization: Bearer zdk_live_…" \
-H "Content-Type: application/json" \
-d '{ "source": <SiteImportSource>, "output": <SiteImportOutput>, "source_domain": <string> }'Angemeldet? Die API-Konsole in Ihrem Dashboard trägt automatisch Ihre echte Organisations-ID sowie Ihren eigenen Schlüssel ein und führt die Anfrage gegen die Live-API aus, sodass Sie die tatsächliche Antwort sehen können. Öffnen Sie diesen Endpunkt in der API-Konsole
Details
Starts a durable, resumable import workflow and returns the new job state, never a bare acknowledgement. A Wayback restore **must** name `snapshot_day`: the most recent capture of an expired domain is usually a parking page, so there is no safe default (422 without it). The output is either a static HTML document root or a WordPress import — both are supported and neither is a default. A site with an import already running returns 409. ⚠️ `ownership_attested` is recorded, not enforced: the platform's obligation here is to know who requested what. Requires `hosting.import.manage`.
Parameter
| Name | Typ | Erforderlich | Was es ist |
|---|---|---|---|
Idempotency-Key (header) | string | Nein | Client-generated key that makes an unsafe request replay-safe: the server stores the first response and returns it verbatim for repeats. |
Anfragekörper
| Name | Typ | Erforderlich | Was es ist |
|---|---|---|---|
source | SiteImportSource | Ja | Where the content comes from. ⛔ Deliberately not the same enum as a site's `source`: that one records what a site was built from, for ever, while this selects a fetch strategy a… |
output | SiteImportOutput | Ja | What the import produces. Both are supported; neither is a default. |
source_domain | string | Ja | The domain whose archive to restore, or — for `archive_url` — the URL of the archive to migrate in. |
target_domain | string | Nein | The domain the result is served on. Defaults to `source_domain`. |
site_id | Uuid | Nein | Restore into an existing site. Omit to produce a bundle and choose the target later. |
snapshot_day | string | Nein | Which day's snapshot to restore. **Required for `wayback`** — the most recent capture of an expired domain is usually a parking page, so there is no safe default. |
crawl_depth | integer | Nein | — |
request_budget | integer | Nein | The hard ceiling on requests this job may make against the source. Bounded per job so one 400,000-capture domain cannot consume the platform's whole archive allowance. |
ownership_attested | boolean | Nein | ⚠️ The customer's assertion that they control the domain, recorded because it cannot be reconstructed afterwards. It is **not** a legal control and is not presented as one. |
Antwort
| Name | Typ | Erforderlich | Was es ist |
|---|---|---|---|
id | Uuid | Ja | UUIDv7 identifier — sortable by creation time (docs/02 §8). |
site_id | string | Nein | The site this import lands on, or empty. Empty is a real outcome, not an error: an aged-domain restore produces a downloadable bundle and the customer picks the target afterwards. |
source | SiteImportSource | Ja | Where the content comes from. ⛔ Deliberately not the same enum as a site's `source`: that one records what a site was built from, for ever, while this selects a fetch strategy a… |
output | SiteImportOutput | Ja | What the import produces. Both are supported; neither is a default. |
status | SiteImportStatus | Ja | `fetching` and `building` are separate states rather than one `running` because they fail for opposite reasons and the customer's next action differs — a stall in `fetching` is… |
status_reason | string | Nein | A machine code the dashboard renders through its own catalogue — never an English sentence composed by the engine, which could not be shown to the other 57 locales. |
source_domain | string | Ja | — |
target_domain | string | Nein | — |
snapshot_day | string | Nein | — |
crawl_depth | integer | Nein | — |
urls_total | integer | Ja | — |
urls_done | integer | Ja | — |
urls_failed | integer | Ja | — |
progress_percent | integer | Ja | For a progress bar only. The per-URL counters are the primary display, because a percentage cannot say *which* step. |
requests_spent | integer | Ja | — |
request_budget | integer | Ja | — |
budget_exhausted | boolean | Ja | True when the job stopped because it ran out of requests. Surfaced deliberately — it otherwise looks identical to a job that finished, and the customer's next action depends on… |
bundle_bytes | integer | Nein | — |
created_at | string | Ja | — |
started_at | string | Nein | — |
finished_at | string | Nein | — |
can_cancel | boolean | Ja | — |
Fehler, die dieser Endpunkt zurückgeben kann
401 · 403 · 404 · 409 · 422 · 429 · 503