Run tree-sitter queries on a JSON file or string

See https://tree-sitter.github.io/tree-sitter/ on writing tree-sitter queries.

Usage

query_json(file = NULL, text = NULL, query, ranges = NULL)

Arguments

file: Path of a JSON file. Use either file or text.
text: JSON string. Use either file or text.
query: Character string, the tree-sitter query to run.
ranges: Can be used to parse part(s) of the input. It must be a data frame with integer columns start_row, start_col, end_row, end_col, start_byte, end_byte, in this order.

Value

A list with entries patterns and matched_captures. patterns contains information about all patterns in the queries and it is a data frame with columns: id, name, pattern, match_count. matched_captures contains information about all matches, and it has columns id, pattern, match, start_byte, end_byte, start_row, start_column, end_row, end_column, name, code. The pattern column of matched_captured refers to the id column of patterns.

Examples

# A very simple JSON document
txt <- "{ \"a\": 1, \"b\": \"foo\", \"c\": 20 }"

# Take a look at it
load_json(text = txt) |> format_selected()
#> # json (5 lines)
#> 1 | {
#> 2 |     "a": 1,
#> 3 |     "b": "foo",
#> 4 |     "c": 20
#> 5 | }

# Select all pairs where the value is a number
query_json(text = txt, query = "((pair value: (number) @num))")
#> $patterns
#> # A data frame: 1 × 4
#>      id name  pattern                           match_count
#>   <int> <chr> <chr>                                   <int>
#> 1     1 NA    "((pair value: (number) @num))\n"           2
#> 
#> $matched_captures
#> # A data frame: 2 × 11
#>      id pattern match start_byte end_byte start_row start_column
#>   <int>   <int> <int>      <int>    <int>     <int>        <int>
#> 1     1       1     1          8        8         1            8
#> 2     1       1     2         28       29         1           28
#> # ℹ 4 more variables: end_row <int>, end_column <int>, name <chr>,
#> #   code <chr>
#>