All The Places Data Format
The output of the periodic run of all spiders posted on https://www.alltheplaces.xyz/ is a single GeoJSON
FeatureCollection where each
Feature contains the data for a single scraped item. The file is gzipped to save space on disk.
Each GeoJSON feature has an
id field. The ID is a hash based on the
@spider fields and should be consistent between builds. You might use this to determine if new objects show up or disappear between builds.
In most cases, the feature will include a
geometry field following the GeoJSON spec. There are some spiders that aren't able to recover a position from the venue's website. In those cases, the geometry is set to
null and only the properties are included.
Although it's not supported at the time of this writing, we hope to include a geocoding step in the pipeline so that these feature will get a position added.
Each GeoJSON feature will have a
properties object with the following keys:
||Yes||A unique identifier for this feature inside this spider. The code that generates the output will remove duplicates based on the value of this key.|
||Yes||The name of the spider that produced this feature. It is specified in each spider, so it doesn't necessarily related to the file name of the spider.|
||No||The name of the feature. This is usually extracted from the venue's page, so it will probably be different per feature. This is often the location specific part of a chain location's name, like the name of the mall it's in, without the chain name included.|
|Brand||Information about the brand that operates or owns the venue|
||No||The brand or chain name of the feature. This will generally be the same for most features outputted by a scraper. Some scrapers will output for companies that own multiple brands, like Duane Reade and Walgreens for the Walgreens scraper.|
||No||The Wikidata item ID for the brand of the feature. This is a machine-readible identifier counterpart for the human-readible
|Address||Information about the address of the venue|
||No||The full address for the venue in one line of text. Usually this follows the format of street, city, province, postcode address. This field might exist instead of the other address-related fields, especially if the spider can't reliably extract the individual parts of the address.|
||No||The house number part of the address.|
||No||The street name.|
||No||The city part of the address.|
||No||The state or province part of the address.|
||No||The postcode part of the address.|
||No||The country part of the address.|
|Contact||Contact information for the venue|
||No||The telephone number for the venue. Note that this is usually pulled from a website assuming local visitors, so it probably doesn't include the country code.|
||No||The website for the venue. We try to make this a URL specific to the venue and not a generic URL for the brand that is operating the venue.|
||No||The opening hours for the venue. When we can, the format for this field follows OpenStreetMap's
Spiders can also include extra fields that will show up but aren't necessarily documented outside their spider source code. If enough spiders find interesting things to include in an extra property, it might be included here in the documentation in the future.