Imports
Path: /{workspaceSlug}/import. The entry point for every new data source.
Supported formats
| Source | What it is | Plans |
|---|---|---|
| File | Local upload (CSV, XLSX, JSONL) | all |
| URL feed | URL of an XML / CSV / JSONL feed — Sytalog downloads it and processes it | all |
| Scrape | URL of a product page / e-shop category — Sytalog runs a Playwright scraper | Beta · Enterprise only |
| XML availability | Heureka-compatible availability XML feed | paid plans |
| XML images | Heureka-compatible images XML feed | paid plans |
Import flow
- Pick source — file / URL / scrape.
- Preview columns — Sytalog auto-detects columns; tick the ones to import. UI shows "Selected X of Y columns".
- Name the dataset — placeholder "Stocks Q1".
- Start import — the job joins the queue (RQ + Redis). The page switches to a live log.
- Watch progress: discovered URLs (scrape), detail queue / processed / saved / skipped / errors / blocked (403/429/challenge), estimated total, job ID.
- On finish:
success→ "Open created dataset" deep link;finished_empty→ "Scraping completed, but no products were found" + hint. - Optional: download JSON report with full statistics.
File size limits
| Format | Náhled | Katalog | Provoz | Enterprise |
|---|---|---|---|---|
| CSV / XML / JSONL | 25 MB | 100 MB | 500 MB | 500 MB |
| XLSX | 10 MB | 25 MB | 50 MB | 50 MB |
Files larger than the cap are rejected during upload, not truncated. Sytalog does not limit row count — only file size matters.
XML feed import
- Bulk Import → pick URL feed → enter the XML feed URL → Sytalog detects the structure, you pick columns, you start.
- For availability XML (SKU → stock / price):
/{workspaceSlug}/import/xml→ XML availability - For image XML (SKU → image URL):
/{workspaceSlug}/import/xml→ XML images
The XML must be publicly accessible (no auth) and return a valid XML response.
Scraping (Enterprise · Beta)
Bulk Import → Scrape → product-page / category URL → advanced settings (max pages, max products) → run.
Sytalog uses Playwright for JS rendering. Some e-shops have anti-bot protection that may cause a finished_empty status. Contact support — for Enterprise we can add a custom selector.
Duplicate guard
- Uploading the same file / URL within the last 15 minutes → modal "Source was uploaded recently. Yes, import again / Cancel".
- A dataset with the same name → modal "Dataset already exists. Choose: Overwrite / Rename".
Column mapping
Sytalog auto-detects columns at import time — from headers, sample values, and known feed formats (Heureka, etc.). The UI shows "Detected N columns" and you uncheck the ones you don't want.
For a dataset update (refreshing an existing dataset with a new file), auto-mapping is even smarter — Sytalog suggests "dataset column ↔ feed column" pairs and you can adjust them manually.