schema Archives - The Open Web

Does Schema.org QuantitativeValue support arbitrary numbers? Where is it stated?

The following is an AI chat conversation regarding the topic stated in the title.
Depending on the quality of the conversation, modifications may have been made. Modifications may or may not be stated in this page.

The type of https://schema.org/inventoryLevel is https://schema.org/QuantitativeValue It seems that this type does support a general number such as 9. Is this really true?

According to the document, schema.org/inventoryLevel is a type of QuantitativeValue, and QuantitativeValue is expected to be one of these types: whole number, decimal number, or fraction. There is no mention of supporting a general number such as 9, but it does support whole numbers.

I can’t find any reference to numbers being allowed for https://schema.org/QuantitativeValue Please elaborate.

You’re right, there isn’t a direct reference to numbers being allowed for schema.org/QuantitativeValue. The schema.org documentation uses terms like “whole number” and “decimal number” instead of just “number”.

Here’s a breakdown of what the documentation implies:

QuantitativeValue represents a measurement.
Measurements typically involve a value and a unit.
The value can be a whole number (e.g., 10), a decimal number (e.g., 3.14), or a fraction.

While the schema doesn’t explicitly state “number” is allowed, it implies whole numbers are valid values because they represent a specific quantity.

Here are some resources from schema.org that support this:

QuantitativeValue definition: https://schema.org/QuantitativeValue mentions the value being a number for QuantitativeValue and MonetaryAmount.
unitText property: https://schema.org/unitText is used to specify the unit of measurement for a numerical value.

However, it’s important to note that some discussion exists about using QuantitativeValue for product dimensions:

There are open issues on the schema.org Github regarding using QuantitativeValue for product dimensions with units (https://schema.org/QuantitativeValue)

In conclusion, while schema.org implies whole numbers are valid for QuantitativeValue, it’s best practice to consider the context and potential limitations. For product dimensions, there might be a more suitable approach being discussed within the schema.org community.

My issue is that the schema.org page for QuantitativeValue suggests that it is an object with properties, not a number. Where is it stated that normal numbers are allowed?

According to the document, schema.org defines QuantitativeValue as a number or text datatype though it can also be a StructuredValue. schema.org

This suggests that schema.org allows QuantitativeValue to represent a basic number but it can also be more complex data like a structured value.

I checked the page but can not find where it says that. Where does it state that a number can be used?

While the schema.org documentation for QuantitativeValue doesn’t explicitly say that numbers are allowed, it implies it by mentioning point values and various quantities.

For example, it talks about quantities like weight, height, and depth which can all be numerical values.

Also, a search on the web confirms that schema.org recommends using ‘Number’ type for the value property of QuantitativeValue.

Does Schema.org Text type support arbitrary strings? Where is it stated?

In Schema.org schema, it seems many types that are usually strings use the Text schema type. Under “More specific Data Types”, there is no specific string type. Does Text specifically support strings? Why is there no string type under “More specific Data Types”?

Yes, the Text schema type in Schema.org is specifically designed to support strings. It’s intended for representing plain textual content, and it effectively serves as the catch-all type for most string-like data within the schema.

Here’s why there’s no separate string type under “More specific Data Types”:

Text is Broadly Applicable: The Text type can accommodate a wide range of string-based information, making a separate string type less essential. It can handle:
- Free-form text like descriptions or comments
- Addresses
- Email addresses
- Names
- Product descriptions
- Reviews
- And more
Emphasis on Specificity: Schema.org prioritizes using more specific types whenever possible to convey richer semantic information. For example:
- Use PostalAddress for a postal address, rather than just a string.
- Use URL for a web address, rather than a plain string.
- Use Date for a date, rather than a string representing the date.
- Use Duration for a duration of time, rather than a string.

By encouraging more specific types, Schema.org aims to create more structured and machine-readable data. However, Text remains a crucial fallback for representing textual content that doesn’t align with a more specific type.

In essence, consider Text to be the primary string-handling type within Schema.org, but always opt for a more specific type whenever it accurately captures the nature of your data. This approach contributes to better semantic representation and understanding of your content.

Exactly where is it stated that “Text” supports arbitrary strings?

You’re right, there isn’t a definitive statement within the Schema.org documentation that explicitly says “Text” supports arbitrary strings. Schema.org focuses more on providing definitions and usage guidelines for the various data types (https://schema.org/DataType).

While there’s no formal mention of strings, the overall design and usage patterns strongly suggest that Text acts as the catch-all for string-like data. Here’s why:

Text is used extensively across various properties: Schema.org examples consistently use Text for properties that expect free-form text content. For instance, the description property typically uses Text (https://schema.org/description).
Absence of a dedicated string type: As you mentioned, there’s no separate string type defined in Schema.org, implying Text assumes that role.
Emphasis on data categorization: Schema.org prioritizes using more specific data types whenever possible. The lack of a general string type reinforces this approach.

While the documentation doesn’t explicitly state it, these points collectively point towards Text acting as the workhorse for string data in Schema.org.

List of known/common standardized JSON Schema properties.

The JSON Schema standard is extensive and descriptive, but it can lead to taking too much time to quickly find the name of a property, etc.
I decided to generate a list of important JSON Schema properties so they can be easy to find when needed.
The following was generated using Bard AI, and seems to be accurate.

—

Please provide a list of all known standardized properties in jsonschema with the property name, followed by a colon + space delimiter, and the followed by a description of the property. Each should be on one new line.

—
The following is a list of all known standardized properties in JSON Schema, with the property name, followed by a colon + space delimiter, and then followed by a description of the property, each on a new line:

$schema: The URI of the JSON Schema version used to define the schema.
$ref: A reference to another JSON Schema that defines the schema for this object.
$id: A unique identifier for the schema.
title: A human-readable title for the schema.
description: A human-readable description of the schema.
type: The data type of the value that the schema validates.
enum: An array of valid values for the schema.
const: A constant value that the schema validates.
multipleOf: A number that the value must be divisible by.
maximum: The maximum value that the schema validates.
exclusiveMaximum: A boolean value indicating whether the maximum value is exclusive or inclusive.
minimum: The minimum value that the schema validates.
exclusiveMinimum: A boolean value indicating whether the minimum value is exclusive or inclusive.
maxLength: The maximum length of a string that the schema validates.
minLength: The minimum length of a string that the schema validates.
pattern: A regular expression that the value must match.
maxItems: The maximum number of items in an array that the schema validates.
minItems: The minimum number of items in an array that the schema validates.
uniqueItems: A boolean value indicating whether the items in an array must be unique.
maxProperties: The maximum number of properties in an object that the schema validates.
minProperties: The minimum number of properties in an object that the schema validates.
required: An array of property names that must be present in an object that the schema validates.
dependentRequired: An object mapping property names to arrays of property names. If a property that is present in the first array is present in the object, then all of the properties in the second array must also be present.
additionalProperties: A boolean value indicating whether additional properties are allowed in an object that the schema validates.
properties: An object mapping property names to JSON Schemas that define the schemas for those properties.
patternProperties: An object mapping regular expressions to JSON Schemas that define the schemas for properties that match those regular expressions.
allOf: An array of JSON Schemas that the value must validate against.
anyOf: An array of JSON Schemas that the value must validate against at least one of.
oneOf: An array of JSON Schemas that the value must validate against exactly one of.
not: A JSON Schema that the value must not validate against.

Please note that this list is not exhaustive, and there may be other standardized properties in JSON Schema that are not listed here.
—

How to create a valid definitions only jsonschema file?

Sometimes it is useful to group jsonschema to be referenced elsewhere when desired like a utility module but for schema.
However, it is a little confusing on how to group these definitions.
There is also the issue of making sure the schema is still valid because jsonschema libraries that handle things such as validation may not accept custom formats.

I asked Bard AI, and got an appropriate response, so I have shared the information below:

I want to create a jsonschema file that only contains definitions(jsonschema that can be referenced).
I don’t need the root object to be a usable schema.
I want the file to be a valid jsonschema.

Is this possible, and please provide an example.

—

Yes, it is possible to create a JSON schema file that only contains definitions. This is a perfectly valid way to use JSON schema, and it can be useful for organizing your schemas and making them more reusable.

Here is an example of a JSON schema file that only contains definitions:

JSON

{
  "$schema": "http://json-schema.org/draft/2020-12/schema",
  "definitions": {
    "MediaObject": {
      "type": "object",
      "required": ["type", "url"],
      "properties": {
        "type": {
          "type": "string",
          "enum": ["image/jpeg", "image/png"]
        },
        "url": {
          "type": "string",
          "format": "uri"
        }
      }
    },
    "Person": {
      "type": "object",
      "required": ["name", "age"],
      "properties": {
        "name": {
          "type": "string"
        },
        "age": {
          "type": "integer"
        }
      }
    }
  }
}

This schema file defines two schemas: MediaObject and Person. These schemas can then be referenced from other schema files using the $ref keyword.

For example, the following schema file references the MediaObject and Person schemas defined in the previous example:

JSON

{
  "$schema": "http://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "required": ["mediaObject", "person"],
  "properties": {
    "mediaObject": {
      "$ref": "#/components/schemas/MediaObject"
    },
    "person": {
      "$ref": "#/components/schemas/Person"
    }
  },
  "components": {
    "schemas": {
      "MediaObject": {
        "$ref": "#/definitions/MediaObject"
      },
      "Person": {
        "$ref": "#/definitions/Person"
      }
    }
  }
}

This schema file is valid, even though the root object is not a usable schema. This is because the root object is only used to define the MediaObject and Person schemas, which can then be referenced from other schema files.

You can use this approach to create a library of reusable schema definitions. This can make your schemas more modular and easier to maintain.

—

I was reading the jsonschema specification draft, and noticed that the “$defs” property was used to group definitions. I have noticed you and others have used the “definitions” property to group definitions. Which one is better, and why?

The $defs keyword was introduced in JSON Schema draft 2019-09. It is a newer and more standardized way to group definitions.

The definitions keyword is still valid, but it is deprecated. It is recommended that new schemas use the $defs keyword instead.

There are a few reasons why the $defs keyword is better:

It is more standardized. The definitions keyword was not defined in the JSON Schema specification, but was rather a convention that was widely used. The $defs keyword is defined in the JSON Schema specification, so it is more likely to be supported by all JSON schema implementations.
It is more explicit. The definitions keyword could be used for other purposes in addition to grouping definitions. For example, it could be used to define a set of named anchors for use in JSON Pointer expressions. The $defs keyword is specifically for grouping definitions, so it is more explicit and less likely to be confused with other features of JSON Schema.
It is more future-proof. The JSON Schema specification is still under development, and it is possible that the definitions keyword could be deprecated in the future. The $defs keyword is a newer keyword, so it is more likely to be supported in the future.

Overall, the $defs keyword is the better choice for grouping definitions in JSON Schema. It is more standardized, more explicit, and more future-proof.

However, it is important to note that the definitions keyword is still valid, and there are many existing JSON schemas that use it. If you are working with an existing JSON schema that uses the definitions keyword, there is no need to change it to use the $defs keyword.

—

As stated, if the root schema only provides the definitions, then it is valid, and although the “definitions” property is often used for grouping definitions, it has actually been standardized to be “$defs” instead, but definitions is still valid and has been used from before the property was standardized.