Concepts
Datasets, items, changes, products, categories and sites, and how they fit together.
Dataset
A dataset is a topic (laptops, tyres, drugstore …) collected from several websites. Your plan includes a number of
datasets; you choose them in the portal. Every request can be narrowed to one with dataset=<id>; without it, it covers
all of yours. /datasets lists them with their size and freshness.
Site and source
A site is a website the data comes from, a source one listing of it (a category page, say). Pricana reads every
source regularly and repairs it on its own when the website changes. Items name both (site, source). Data from open
sources carries its licence in site.attribution: pass it on with the data.
Item
An item is one listing: a product offered by one shop, with its title, price, currency, link, image, category and the
shop's other values (attributes, such as availability or shipping). It keeps its id for as long as the shop lists it.
When the shop stops listing it, the item is removed (removedAt is set) but stays readable with its history; if the
shop lists it again, it is relisted.
Change
Whenever something about an item changes, Pricana records a change:
| kind | means |
|---|---|
NEW | the item appeared |
CHANGED | a value changed (changes says which: old and new); a price change also has previousPrice |
REMOVED | the shop no longer lists it |
RELISTED | it came back |
Changes have increasing ids. The change feed lists them in order, webhooks push them to you, and an item's history is its changes over time.
Product
A product groups the items of different shops that sell the same thing: by GTIN (EAN/UPC), else by brand and manufacturer part number (MPN). Products give you the price range and the number of shops per product.
Category
Pricana sorts items into one taxonomy (category.slug, category.path), whatever the shop calls its categories.
category.source says who chose it: AI or MANUAL (a person).
Freshness and plan
How current the data is, and how far back its history goes, depends on your plan: the Free plan shows the sample data as of 24 hours ago with seven days of history; paid plans are current. Limits and quotas has the details.