Skip to contents

Parse and Validate JSON Against a Structure

Source: R/parse_json.R
parse_json.Rd

Parses a JSON string and validates it against a schema created with build_structure(). Returns an R object reflecting the JSON content if valid, or throws a detailed error.

Usage

parse_json(json_string, structure)

Arguments

json_string

A character string containing the JSON input.

structure

A structure definition created with build_structure(), based on s_* functions.

Value

The parsed and validated data as an R object:

  • JSON objects → named lists

  • JSON arrays → unnamed lists

  • Strings, numbers, booleans → scalar R values (character, numeric, integer, logical)

Errors are raised for invalid syntax, type mismatches, missing or extra fields, duplicate keys, and integer overflows.

Details

Validation checks include:

  • Types: JSON values must match the declared structure (e.g., string vs integer).

  • Required fields: All fields in s_map() must be present.

  • No extra fields: Additional fields not in the structure cause errors (unless .ignore_extra_fields = TRUE).

  • Duplicate keys: Disallowed.

  • Integer range: Must fit within R's 32-bit integer limits for s_integer().

  • Homogeneous arrays: Elements in s_vector() must match the defined structure.

Parsing is powered by serde_json in Rust for performance.

Examples

# Define and build a structure
schema <- build_structure(s_map(
  id = s_integer(),
  username = s_string(),
  is_active = s_logical(),
  scores = s_vector(s_double())
))

# Valid input
json <- '{"id":1,"username":"user","is_active":true,"scores":[9.5,8.0]}'
parse_json(json, schema)
#> $id
#> [1] 1
#> 
#> $username
#> [1] "user"
#> 
#> $is_active
#> [1] TRUE
#> 
#> $scores
#> [1] 9.5 8.0
#> 

# Common errors:
# - Invalid JSON
# - Wrong type (e.g., "id": "abc")
# - Missing fields
# - Unexpected extra fields
# - Integer overflow
# - Duplicate keys