jq-rs

Overview

Prior to v0.4.0 this crate was named json-query.

This rust crate provides access to jq 1.6 via the libjq C API (rather than "shelling out").

By leveraging jq we can extract data from json strings using jq's dsl.

This crate requires Rust 1.32 or above.

Usage

The interface provided by this crate is very basic. You supply a jq program string and a string to run the program over.

use jq_rs;
// ...

let res = jq_rs::run(".name", r#"{"name": "test"}"#);
assert_eq!(res.unwrap(), "\"test\"\n".to_string());

In addition to running one-off programs with jq_rs::run(), you can also use jq_rs::compile() to compile a jq program and reuse it with different inputs.

use jq_rs;

let tv_shows = r#"[
 {"title": "Twilight Zone"},
 {"title": "X-Files"},
 {"title": "The Outer Limits"}
]"#;

let movies = r#"[
 {"title": "The Omen"},
 {"title": "Amityville Horror"},
 {"title": "The Thing"}
]"#;

let mut program = jq_rs::compile("[.[].title] | sort").unwrap();

assert_eq!(
 &program.run(tv_shows).unwrap(),
 "[\"The Outer Limits\",\"Twilight Zone\",\"X-Files\"]\n"
);

assert_eq!(
 &program.run(movies).unwrap(),
 "[\"Amityville Horror\",\"The Omen\",\"The Thing\"]\n",
);

A Note on Performance

While the benchmarks are far from exhaustive, they indicate that much of the runtime of a simple jq program goes to the compilation. In fact, the compilation is quite expensive.

run one off             time:   [48.594 ms 48.689 ms 48.800 ms]
Found 6 outliers among 100 measurements (6.00%)
3 (3.00%) high mild
3 (3.00%) high severe

run pre-compiled        time:   [4.0351 us 4.0708 us 4.1223 us]
Found 15 outliers among 100 measurements (15.00%)
6 (6.00%) high mild
9 (9.00%) high severe

If you have a need to run the same jq program multiple times it is highly recommended to retain a pre-compiled JqProgram and reuse it.

Handling Output

The return values from jq are strings since there is no certainty that the output will be valid json. As such the output will need to be parsed if you want to work with the actual data types being represented.

In such cases you may want to pair this crate with serde_json or similar.

For example, here we want to extract the numbers from a set of objects:

use jq_rs;
use serde_json::{self, json};

// ...

let data = json!({
 "movies": [
     { "title": "Coraline", "year": 2009 },
     { "title": "ParaNorman", "year": 2012 },
     { "title": "Boxtrolls", "year": 2014 },
     { "title": "Kubo and the Two Strings", "year": 2016 },
     { "title": "Missing Link", "year": 2019 }
 ]
});

let query = "[.movies[].year]";
// program output as a json string...
let output = jq_rs::run(query, &data.to_string()).unwrap();
// ... parse via serde
let parsed: Vec<i64> = serde_json::from_str(&output).unwrap();

assert_eq!(vec![2009, 2012, 2014, 2016, 2019], parsed);

Barely any of the options or flags available from the jq cli are exposed currently. Literally all that is provided is the ability to execute a jq program on a blob of json. Please pardon my dust as I sort out the details.

Linking to libjq

This crate requires access to libjq at build and/or runtime depending on the your choice.

When the bundled feature is enabled (off by default) libjq is provided and linked statically to your crate by jq-sys and jq-src. Using this feature requires having autotools and gcc in PATH in order for the to build to work.

Without the bundled feature, you will need to ensure your crate can link to libjq in order for the bindings to work.

You can choose to compile libjq yourself, or perhaps install it via your system's package manager. See the jq-sys building docs for details on how to share hints with the jq-sys crate on how to link.