30 terms
Glossary
The vocabulary this site uses, defined in one sentence each. The one-sentence rule is the point: a definition that needs the paragraph around it to make sense is no use to a reader skimming, and no use to a model quoting it.
A
- Actor
- A packaged, runnable program on the Apify platform that takes a JSON input, does work on the web, and writes structured results to a dataset.
C
- Cache age
- How long ago a row's data was actually read from the target site, as distinct from when your run produced the row.
- CCPA
- The California Consumer Privacy Act, which grants California residents rights over their personal information and whose treatment of publicly available data is narrower than the phrase suggests.
- Concurrency
- How many requests a scraper has in flight at once — the single setting that most often decides whether a run finishes or spends its time in penalty cooldowns.
D
- Dataset
- Append-only tabular storage on the Apify platform where an Actor writes its results, exportable as JSON, CSV or Excel.
- DGT environmental badge
- Spain's official vehicle emissions classification, issued by the Dirección General de Tráfico as the codes `0`, `ECO`, `C` and `B`, which govern access to urban low-emission zones.
E
- Error dataset
- A separate dataset holding the rows that failed, with the reason for each, so a partial run is auditable rather than silently incomplete.
- Explicit null
- A field that is present in every row but carries `null` when it has no value, instead of being omitted from the row entirely.
F
- Facet expansion
- A discovery strategy that reaches a large inventory by combining filters — make, then region, then price band — instead of paginating deep into one unfiltered list.
- Fill rate
- The share of rows in which a given field actually contains a value, measured over a real sample rather than asserted.
G
- GDPR Article 14
- The provision of Regulation (EU) 2016/679 that obliges you to inform people when you obtain their personal data from a source other than themselves, as a rule within one month.
H
- Headless browser
- A real browser engine driven programmatically with no visible window, used when a site only assembles its data after JavaScript runs.
I
- Input schema
- A JSON Schema document that declares every option an Actor accepts, with its type, default and description, and from which the platform generates the run form.
K
- Km 0
- A nearly-new car registered by the dealer rather than by a customer, sold as used despite having almost no mileage.
L
- llms.txt
- A plain-text file at a site's root that describes the site, its scope and its canonical URLs for the benefit of language models rather than search crawlers.
N
- Nextdoor city slug
- Nextdoor's identifier for a city, formatted `city-name--state`: lowercase, single hyphens inside the name, a double hyphen, then the two-letter state code.
O
- Output view
- A declared projection of a dataset that selects and renames fields, letting one set of rows be presented as several different tables.
P
- Pagination depth
- How many pages deep into a result list a crawler goes before it stops, often capped both by the site's robots.txt and by the point at which the site stops returning results.
- Pay-per-event pricing
- A pricing model in which the user is charged for each result an Actor delivers rather than for the compute time the run consumed.
- Personal data
- Any information relating to an identifiable living person, including combinations that identify somebody only when taken together.
- Price rank indicator
- Coches.net's own placement of a listing against its own estimate of what that model is worth on the Spanish market.
R
- Rate limiting
- A server-side control that rejects requests once a client exceeds a threshold, usually with HTTP 429 and a cooldown before normal service resumes.
- Residential proxy
- A proxy that routes requests through IP addresses assigned to home internet connections, so traffic originates from the country and network type a site expects of its visitors.
- robots.txt
- A plain-text file at a site's root in which the site declares, in machine-readable form, which URL patterns automated clients may and may not request.
S
- Server-side rendering
- A site architecture in which the server sends fully populated HTML, so the data is present in the first response and no browser is needed to see it.
- Service request
- A post in which a neighbor asks for a provider rather than reviewing one — "anyone know a good plumber?" — which sits in the same Nextdoor feed as genuine recommendations.
- Session cookie
- A credential a site issues to a logged-in browser, which a scraper can borrow to reach private content — and which expires, taking the scraper with it.
- Signed-out surface
- Everything a site serves to a visitor with no account and no session — the only part of a site a scraper can read without borrowing somebody's credentials.
- Structured data
- Machine-readable markup, usually JSON-LD following schema.org, that states explicitly what a page is about instead of leaving a crawler to infer it.
T
- Trade area
- The catchment a business actually serves, which crosses municipal boundaries — and which is what a distance-ordered directory returns when you ask it for a city.