Syntax¶

RCL is a superset of json. Any json document is a valid RCL expression which evaluates to itself as json. RCL furthermore features the following constructs.

Comments¶

Comments start with // and run until the end of the line. Comments in RCL are slightly unusual in that there are some locations where comments are not allowed.¹ Generally, prefer to put comments on their own line, before the item they comment on.

// Comment like this.
let answer = 42;
let question = "unknown"; // The formatter would move this to the next line.
{ question: answer }

At the start of the document, a line that starts with #! is allowed, in order to support executable files. For example:

#!/usr/bin/env -S rcl eval
"This document prints this string when executed."

Booleans and null¶

The booleans are written true and false, null is written null.

Strings and f-strings¶

Strings are quoted with " and support the same escape sequences as json.

"This is a string."

Multi-line strings can be quoted with """. In both cases, adding an f in front turns the string into a format string, which can have one or more holes delimited by {}, to interpolate content into it:

f"""
The answer to the ultimate question is {2 * 3 * 7}.
"""

See the chapter on strings for the full details.

Identifiers¶

Names of variables, and dict fields that use record syntax, are identifiers. Identifiers must start with an underscore or ASCII letter, and can furthermore contain ASCII digits, and -, a hyphen.

Allowing the hyphen in identifiers makes some types of configuration dicts cleaner to write, but the downside is that it can cause confusion with the - operator. The expression a-b is an identifier, not the binary operator - applied to a and b. To subtract b from a, add spaces around the operator: a - b.

Lists¶

Lists are surrounded by []. The list separator is , and a trailing comma is allowed but not required.

[
  ["Apple", "Banana", "Pear"],
  ["Eggplant", "Pepper", "Zuccini"],
]

Dictionaries¶

Dictionaries, dicts for short, are surrounded by {}. Dicts can be written in json form, where the left-hand side is an expression. Then the key and value are separated by :. The element separator is ,. A trailing comma is optional.

{
  "name": "apple",
  "flavor": "sweet",
}

The left-hand side does not have to be a string, although using other types than strings precludes serialization to json.

{
  1: "I",
  5: "V",
  5 + 5: "X",
}

Alternatively, dicts can be written in record form, where the left-hand side is an identifier. Then the key and value are separated by =. A trailing comma is optional. The following value is identical to the first one above.

{
  name = "apple",
  flavor = "sweet",
}

Note, without type annotations, the empty collection {} is a dict, not a set.

Sets¶

Sets are surrounded by {} and work otherwise the same as lists. The following list contains two identical sets:

[
  {"Apple", "Pear"},
  {"Apple", "Pear", "Apple"},
]

Note, without type annotations, the empty collection {} is a dict, not a set. To produce an empty set, we can use std.empty_set. A type annotation will also force {} to be a set:

let empty_set: Set[Int] = {};

Let bindings¶

Values can be bound to names with a let-binding.

let name = "apple";
let flavor = "sweet";
[name, flavor]

A let-binding is an expression, not an assignment statement. The expression evaluates to the expression after ;.

Let bindings can optionally contain a type annotation:

let answer: Int = 42;

List indexing¶

Brackets are used to index into lists. Indices must be integers and are 0-based. Negative indices index from the back of the list.

let xs = ["Deckard", "Rachael", "Tyrell"];
// Evaluates to "Deckard".
xs[0]
// Evaluates to "Tyrell".
xs[-1]

Dictionary indexing¶

Brackets are also used to look up a key in a dictionary.

let replicants = {
  "NEXUS-7 N7FAA52318": "Rachael",
  "NEXUS-6 N6MAA10816": "Roy Batty",
  "NEXUS-6 N6MAC41717": "Leon Kowalski",
};
// Evaluates to "Rachael".
replicants["NEXUS-7 N7FAA52318"]

Looking up a key that does not occur in the dictionary causes evaluation to abort with an error. To handle optional keys gracefully, use the Dict.get method.

Field access¶

The . can be used to access methods on values.

// Evaluates to 3.
"abc".len()

// Evaluates to false.
{1, 2, 3}.contains(4)

The . can also be used to access fields of dictionaries. In most cases this can be used as a more readable alternative to indexing notation.

let replicant = { name = "Zhora Salome", model = "NEXUS-6 N6FAB61216" };
// Evaluates to "Zhora Salome".
replicant.name

When a dictionary contains a key with the same name as a built-in method, the method takes precedence.

let confusing = { len = 100 };

// Evaluates to builtin method Dict.len, not to the integer 100.
confusing.len

To access the value, use indexing notation instead. The same applies to keys that are not valid identifiers, such as keys with spaces or non-ASCII letters.

let confusing = { len = 100 };
// Evaluates to 100.
confusing["len"]

let populations = {
  "Amsterdam": 1_459_402,
  "Düsseldorf": 1_220_000,
  "New York": 19_426_449,
};
// Evaluates to [1459402, 1220000, 19426449].
[populations.Amsterdam, populations["Düsseldorf"], populations["New York"]]

Conditionals¶

An if-else expression evaluates to the then or else part depending on the condition:

let log_level = if flags.contains("--verbose"): 5 else: 1;

let rustc_codegen_opts =
  if config.is_debug:
    { opt-level = 0, debuginfo = 2 }
  else:
    { opt-level = 2, target-cpu = "native" };

Because an if-else expression is an expression, the else part is mandatory.

Operators¶

The following operators are supported. Most of them are similar to Python.

Unary operators that operate to the left of an expression, e.g. not x:

Binary operators that operate between two expressions, e.g. x and y:

Unlike most other languages (but like Pony), RCL does not have different precedence levels. To avoid confusing combinations of operators, you have to use parentheses:

// Invalid: Unclear whether this means (X and Y) or Z, or X and (Y or Z).
let should_log_verbose =
  settings.contains("log") and settings.log_level >= 2
  or settings.contains("debug");

// Disambiguate with parens:
let should_log_verbose =
  (settings.contains("log") and (settings.log_level >= 2))
  or settings.contains("debug");

Comprehensions¶

Inside collection literals (lists, dicts, and sets), aside from single elements, it is possible to use comprehensions. There are three supported constructs: for, if, and let.

let dict = {"name": "pear", "flavor": "sweet"};
[for key, value in dict: value]
// Evaluates to:
["pear", "sweet"]

[if log_level >= 2: "Verbose message"]
// When log_level < 2, evaluates to:
[]
// When log_level >= 2, evaluates to:
["Verbose message"]

{let x = 10; "value": x}
// Evaluates to:
{"value": 10}

These can be combined arbitrarily:

let labels = {
  for server in servers:
  let all_server_labels = server_labels[server] | default_labels;
  for label in all_server_labels:
  if not excluded_labels.contains(label):
  label
};

An if inside a comprehension controls the loop, it is not part of an if-else expression. To use an if-else expression inside a comprehension, enclose it in parentheses:

let target_os = {
  for server in servers:
  // This 'if' excludes servers from before 2021 from the resulting dict.
  if server.year_acquired >= 2021:
  server.name:
  // This 'if' is part of an if-else expression.
  (if server.year_acquired >= 2023: "ubuntu:22.04" else: "ubuntu:20.04")
};

There can be multiple loops per collection, and they can be mixed with single elements:

let small_numbers = [1, 2, 3];
let large_numbers = [100, 200, 300];
[
  for n in small_numbers: n,
  10,
  for n in large_numbers: n,
]
// Evaluates to:
[1, 2, 3, 10, 100, 200, 300]

Assertions¶

You can use assertions in expressions and inside comprehensions:

// Expression form:
assert condition, "Message for when the assertion fails.";
body

// Comprehension form:
[
  for widget in widgets:
  assert widget.is_valid(), f"Widget {widget.id} is invalid.";
  widget
]

The message is mandatory (unlike in Python). When the assertion fails, evaluation aborts with the given message. The message does not have to be a string, it can be an arbitrary value. When the assertion succeeds, the message does not get evaluated at all.

Debug tracing¶

In larger programs it can sometimes be useful to print what is going on during evaluation. However, RCL is a purely functional language without side effects; the only output it can produce is the final value. To still aid debugging, trace acts as an escape hatch: it has the side effect of printing a value to stderr during evaluation.

Like assertions, you can use trace in expressions and inside comprehensions:

// Expression form:
trace "Value that gets printed just before we evaluate `body`.";
body

// Comprehension form:
let widget_ids = [for widget in widgets: trace widget; widget.id];

The message does not have to be a string, it can be an arbitrary value.

Imports¶

An import expression evaluates to the contents of another RCL document.

let inventory = import "inventory.rcl";
[for server in inventory: server.name]

Import paths are relative to the location of the document itself, but there are some restrictions on whether imports are allowed. See the imports chapter for full details.

Functions¶

A => arrow introduces a function.

let double_input = x => x * 2;
let add = (x, y) => x + y;
// Evaluates to 42.
add(double_input(11), 20)

See the chapter on functions for more details.