insta: a snapshot testing library for Rust

Snapshots tests (also sometimes called approval tests) are tests that assert values against a reference value (the snapshot). This is similar to how assert_eq! lets you compare a value against a reference value but unlike simple string assertions, snapshot tests let you test against complex values and come with comprehensive tools to review changes.

Snapshot tests are particularly useful if your reference values are very large or change often.

#[test]
fn test_hello_world() {
    insta::assert_debug_snapshot!(vec![1, 2, 3]);
}

Curious? There is a screencast that shows the entire workflow: watch the insta introduction screencast. Or if you're not into videos, read the one minute introduction.

Install insta:

Recommended way if you have cargo-edit installed:

$ cargo add --dev insta

Alternatively edit your Cargo.toml manually and add insta as manual dependency.

And for an improved review experience also install cargo-insta:

$ cargo install cargo-insta

use insta::assert_debug_snapshot;

#[test]
fn test_snapshots() {
    assert_debug_snapshot!(vec![1, 2, 3]);
}

The recommended flow is to run the tests once, have them fail and check if the result is okay. By default the new snapshots are stored next to the old ones with the extra .new extension. Once you are satisifed move the new files over. To simplify this workflow you can use cargo insta review which will let you interactively review them:

$ cargo test
$ cargo insta review

For more information on updating see Snapshot Updating.

This crate exports multiple macros for snapshot testing:

• assert_snapshot! for comparing basic string snapshots.
• assert_debug_snapshot! for comparing Debug outputs of values.
• assert_display_snapshot! for comparing Display outputs of values.
• assert_csv_snapshot! for comparing CSV serialized output of types implementing serde::Serialize. (requires the csv feature)
• assert_toml_snapshot! for comparing TOML serialized output of types implementing serde::Serialize. (requires the toml feature)
• assert_yaml_snapshot! for comparing YAML serialized output of types implementing serde::Serialize.
• assert_ron_snapshot! for comparing RON serialized output of types implementing serde::Serialize. (requires the ron feature)
• assert_json_snapshot! for comparing JSON serialized output of types implementing serde::Serialize.

Snapshots are stored in the snapshots folder right next to the test file where this is used. The name of the file is <module>__<name>.snap where the name of the snapshot. Snapshots can either be explicitly named or the name is derived from the test name.

Additionally snapshots can also be stored inline. In that case the cargo-insta tool is necessary. See inline snapshots for more information.

For macros that work with serde::Serialize this crate also permits redacting of partial values. See redactions for more information.

The committed snapshot files will have a header with some meta information that can make debugging easier and the snapshot:

---
expression: "vec![1, 2, 3]"
source: tests/test_basic.rs
---
[
    1,
    2,
    3
]

During test runs snapshots will be updated according to the INSTA_UPDATE environment variable. The default is auto which will write all new snapshots into .snap.new files if no CI is detected so that cargo-insta can pick them up. Normally you don't have to change this variable.

INSTA_UPDATE modes:

• auto: the default. no for CI environments or new otherwise
• always: overwrites old snapshot files with new ones unasked
• unseen: behaves like always for new snapshots and new for others
• new: write new snapshots into .snap.new files
• no: does not update snapshot files at all (just runs tests)

When new or auto is used as mode the cargo-insta command can be used to review the snapshots conveniently:

$ cargo install cargo-insta
$ cargo test
$ cargo insta review

"enter" or "a" accepts a new snapshot, "escape" or "r" rejects, "space" or "s" skips the snapshot for now.

For more information invoke cargo insta --help.

By default the tests will fail when the snapshot assertion fails. However if a test produces more than one snapshot it can be useful to force a test to pass so that all new snapshots are created in one go.

This can be enabled by setting INSTA_FORCE_PASS to 1:

$ INSTA_FORCE_PASS=1 cargo test --no-fail-fast

A better way to do this is to run cargo insta test --review which will run all tests with force pass and then bring up the review tool:

$ cargo insta test --review

All snapshot assertion functions let you leave out the snapshot name in which case the snapshot name is derived from the test name (with an optional leading test_ prefix removed.

This works because the rust test runner names the thread by the test name and the name is taken from the thread name. In case your test spawns additional threads this will not work and you will need to provide a name explicitly. There are some situations in which rust test does not name or use threads. In these cases insta will panic with an error. The backtrace feature can be enabled in which case insta will attempt to recover the test name from the backtrace.

Explicit snapshot naming can also otherwise be useful to be more explicit when multiple snapshots are tested within one function as the default behavior would be to just count up the snapshot names.

To provide an explicit name provide the name of the snapshot as first argument to the macro:

#[test]
fn test_something() {
    assert_snapshot!("first_snapshot", "first value");
    assert_snapshot!("second_snapshot", "second value");
}

This will create two snapshots: first_snapshot for the first value and second_snapshot for the second value. Without explicit naming the snapshots would be called something and something-2.

Insta by default will output quite a lot of information as tests run. For instance it will print out all the diffs. This can be controlled by setting the INSTA_OUTPUT environment variable. The following values are possible:

• diff (default): prints the diffs
• summary: prints only summaries (name of snapshot files etc.)
• minimal: like summary but more minimal
• none: insta will not output any extra information

Feature: redactions

For all snapshots created based on serde::Serialize output insta supports redactions. This permits replacing values with hardcoded other values to make snapshots stable when otherwise random or otherwise changing values are involved. Redactions became an optional feature in insta 0.11 and can be enabled with the redactions feature.

Redactions can be defined as the third argument to those macros with the syntax { selector => replacement_value }.

The following selectors exist:

• .key: selects the given key
• ["key"]: alternative syntax for keys
• [index]: selects the given index in an array
• []: selects all items on an array
• [:end]: selects all items up to end (excluding, supports negative indexing)
• [start:]: selects all items starting with start
• [start:end]: selects all items from start to end (end excluding, supports negative indexing).
• .*: selects all keys on that depth
• .**: performs a deep match (zero or more items). Can only be used once.

Example usage:

#[derive(Serialize)]
pub struct User {
    id: Uuid,
    username: String,
    extra: HashMap<String, String>,
}

assert_yaml_snapshot!(&User {
    id: Uuid::new_v4(),
    username: "john_doe".to_string(),
    extra: {
        let mut map = HashMap::new();
        map.insert("ssn".to_string(), "123-123-123".to_string());
        map
    },
}, {
    ".id" => "[uuid]",
    ".extra.ssn" => "[ssn]"
});

It's also possible to execute a callback that can produce a new value instead of hardcoding a replacement value by using the dynamic_redaction function:

assert_yaml_snapshot!(&User {
    id: Uuid::new_v4(),
    username: "john_doe".to_string(),
}, {
    ".id" => dynamic_redaction(|value, _| {
        // assert that the value looks like a uuid here
        "[uuid]"
    }),
});

Feature: glob

Sometimes it can be useful to run code against multiple input files. The easiest way to accomplish this is to use the glob! macro which runs a closure for each input path that matches. Before the closure is executed the settings are updated to set a reference to the input file and the appropriate snapshot suffix.

Example:

use std::fs;

glob!("inputs/*.txt", |path| {
    let input = fs::read_to_string(path).unwrap();
    assert_json_snapshot!(input.to_uppercase());
});

The path to the glob macro is relative to the location of the test file. It uses the globwalk crate for actual glob operations.

Additionally snapshots can also be stored inline. In that case the format for the snapshot macros is assert_snapshot!(reference_value, @"snapshot"). The leading at sign (@) indicates that the following string is the reference value. cargo-insta will then update that string with the new value on review.

Example:

#[derive(Serialize)]
pub struct User {
    username: String,
}

assert_yaml_snapshot!(User {
    username: "john_doe".to_string(),
}, @"");

After the initial test failure you can run cargo insta review to accept the change. The file will then be updated automatically.

The following features exist:

• csv: enables CSV support (assert_csv_snapshot!)
• ron: enables RON support (assert_ron_snapshot!)
• toml: enables TOML support (assert_toml_snapshot!)
• redactions: enables support for redactions
• glob: enables support for globbing (glob!)
• colors: enables color output (enabled by default)

There are some settings that can be changed on a per-thread (and thus per-test) basis. For more information see settings.

With insta 0.11 the snapshot format was improved for inline snapshots. The old snapshot format will continue to be available but if you want to upgrade them make sure the tests pass first and then run the following command to force a rewrite of them all:

$ cargo insta test --accept --force-update-snapshots

Insta cannot detect unused snapshot files. The reason for this is that insta does not control the execution of the entirety of the tests so it cannot spot which files are actually unreferenced. However you can use the INSTA_SNAPSHOT_REFERENCES_FILE environment variable to instruct insta to append all referenced files into a list. This can then be used to delete all files not referenced. For instance one could use ripgrep like this:

export INSTA_SNAPSHOT_REFERENCES_FILE="$(mktemp)"
cargo test
rg --files -lg '*.snap' "$(pwd)" | grep -vFf "$INSTA_SNAPSHOT_REFERENCES_FILE" | xargs rm
rm -f $INSTA_SNAPSHOT_REFERENCES_FILE

License: Apache-2.0