@protoutil/core

@protoutil/core

A set of utilities for working with well-known protobuf types. These utilities assume you are using protobuf-es to work with messages.

Install

Use your configured package manager to install the @protoutil/core package. i.e. install from npm using npm install @protoutil/core.

Usage

CheckSum

The calculateMessageCheckSum function can be used to calculate a non-cryptographic checksum for a given message. The checksums for any two messages with identical values should be identical. But, the checksums for any two messages with different values should never match. Internally, this is used to create pagination page tokens and ETags in the AIP package. But, it is useful for quickly comparing any two messages:

import { calculateMessageCheckSum } from '@protoutil/core';

const message1 = create(MySchema, { foo: 'bar' });
const checksum1 = calculateMessageCheckSum(MySchema, message1);

const message2 = create(MySchema, { foo: 'bar' });
const checksum2 = calculateMessageCheckSum(MySchema, message2);

const message3 = create(MySchema, { baz: 'quz' });
const checksum3 = calculateMessageCheckSum(MySchema, message3);

checksum1 === checksum2; // true
checksum1 === checksum3; // false
checksum2 === checksum3; // false

Duration

A Duration represents a signed, fixed-length span of time represented as a count of seconds and fractions of seconds at nanosecond resolution. It is independent of any calendar and concepts like "day" or "month". It is related to Timestamp in that the difference between two Timestamp values is a Duration and it can be added or subtracted from a Timestamp. Range is approximately +-10,000 years.

The duration function creates and validates a Duration message:

import { duration } from '@protoutil/core';

duration(1n, 1_000_000); // returns a duration object representing 1.001 seconds
duration(315_576_000_001n); // throws an error as durations have a max length of +/-10,000 years

The assertValidDuration and isValidDuration functions validate Duration objects:

import { assertValidDuration, isValidDuration } from '@protoutil/core';

assertValidDuration(d); // throws an error if `d` is not a valid duration
isValidDuration(d); // returns true if `d` is a valid duration and false otherwise

The durationFromString and durationString functions convert strings to and from Duration protobuf messages:

import { durationFromString, durationString } from '@protoutil/core';

durationFromString('2s'); // returns a `Duration` message representing 2 seconds
durationString(d); // returns a string representation i.e. '1.001s'

The durationFromNanos and durationNanos functions convert nanoseconds to and from Duration protobuf messages (useful for performing math operations on durations and timestamps):

import { durationFromNanos, durationNanos } from '@protoutil/core';

durationFromNanos(1_000_000n); // returns a `Duration` message representing 1 millisecond
durationNanos(d); // returns a BigInt i.e. 1_000_000n

// Math example:
const epochTimestamp = timestampFromNanos(0n);
const epochTimestampNanos = timestampNanos(epochTimestamp);
const oneWeekDuration = durationFromString(`${7 * 24 * 60 * 60}s`);
const oneWeekDurationNanos = durationNanos(oneWeekDuration);
const oneWeekAfterEpoch = timestampFromNanos(epochTimestampNanos + oneWeekDurationNanos);
timestampDateString(oneWeekAfterEpoch); // returns '1970-01-08T00:00:00.000Z'

The clampDuration function can be used to ensure that a Duration is between two specified values. By default, clampDuration uses the minimum and maximum values as defined in the protobuf spec (-315,576,000,000 seconds & +315,576,000,000 seconds).

import { clampDuration } from '@protoutil/core';

let min = duration(5n);
let max = duration(10n);
clampDuration(duration(15n), min, max); // returns the max value
clampDuration(duration(1n), min, max); // returns the min value
clampDuration(duration(7n), min, max); // returns the original value
clampDuration(duration(-315,576,000,001n)); // returns a min (-315,576,000,000 seconds) duration
clampDuration(duration(315,576,000,001n)); // returns a max (+315,576,000,000 seconds) duration

Temporal Functions

Temporal is a Stage 3 TC39 proposal which has begun shipping in experimental releases of browsers. Since support is still experimental, we use the temporal-polyfill. Using Temporal instead of Date means native support for nanosecond resolution and simplified operations when working with calendar dates, time zones, date/time calculations, and more. Read more about the Temporal API here.

The durationFromTemporal and durationTemporal functions convert Temporal.Duration objects to and from Duration protobuf messages:

import { durationFromTemporal, durationTemporal } from '@protoutil/core';

durationFromTemporal(duration); // returns a `Duration` object representing the `Temporal.Duration`
durationTemporal(message); // returns a `Temporal.Duration` object representing the `Duration`

Fields

The getField function will get a field from a message given the field descriptor. The setField function will set a field on a message given the field descriptor and a value. Both getField and setField respect oneof values. Note that setField does not validate the type before setting the field (PRs are welcome):

import { getField, setField } from '@protobuf/core';

getField(message, fieldDescriptor); // Gets the value of the field (or returns undefined);
setField(message, fieldDescriptor, value); // Sets the value of the field

FieldMask

FieldMask represents a set of symbolic field paths, for example:

paths: "f.a"
paths: "f.b.d"

Here f represents a field in some root message, a and b fields in the message found in f, and d a field found in the message in f.b.

Field masks are used to specify a subset of fields that should be returned by a get operation or modified by an update operation.

The FieldMask spec does not allow for wildcards and repeated or map fields must be the last part of the path. The final argument for all relevant FieldMask functions is strict, which defaults to true. If strict is true, the function will only allow field masks that are valid according to the spec. However, the AIP Guidelines allow for wildcards in field masks. So, if you want to allow wildcards, you can set strict to false. This will allow for field masks with standalone wildcards or wildcards in repeated or map fields (i.e. '*', 'foo.*', 'foo.*.bar', etc.).

The fieldMask function creates a FieldMask message and asserts that it is valid for the given schema:

import { fieldMask } from '@protoutil/core';

fieldMask(MyMessageSchema 'my_path', 'my_other_path');
// will throw if no fields named 'my_path' or 'my_other_path' are defined on `MyMessageSchema`

The assertValidFieldMask and isValidFieldMask functions validate FieldMask objects for the given schema:

import { assertValidFieldMask, isValidFieldMask } from '@protoutil/core';

assertValidFieldMask(MySchema, fm); // throws if `fm` is not valid for `MySchema`
isValidFieldMask(MySchema, fm); // return true if `fm` is valid for `MySchema` or false otherwise

The fieldMaskHasPath function returns true if a FieldMask matches the given path:

import { fieldMask, fieldMaskHasPath } from '@protoutil/core';

const fm = fieldMask(MySchema, 'a');
fieldMaskHasPath(fm, 'a'); // true
fieldMaskHasPath(fm, 'b'); // false
fieldMaskHasPath(fm, 'a.b'); // true since 'a' was in the field mask so nested fields will be, too.

The applyFieldMask function will apply a field mask to a message. If inverse is true, all fields will be returned EXCEPT the ones in the mask.

import { fieldMask, applyFieldMask } from '@protoutil/core';

const fm = fieldMask(MySchema, 'a');
const message = create(MySchema, { ... });
const updated = applyFieldMask(MySchema, message, fm); // returns a message where only the 'a' field is populated
const inverse = applyFieldMask(MySchema, message, fm, true); // returns a message where the 'a' field is NOT populated

applyFieldMask does not mutate the original message.

The mergeFieldMasks function accepts an arbitrary number of FieldMask objects and merges their paths. When combining, if one FieldMask has a parent field and another has one of its children, only the parent will be returned since it will apply to both:

import { fieldMask, mergeFieldMasks } from '@protoutil/core';

const one = fieldMask(MySchema, 'a');
const two = fieldMask(MySchema, 'a.b', 'c');
mergeFieldMasks(one, two); // returns ['a', 'c'] as paths in the `FieldMask`

The intersectFieldMasks function accepts an arbitrary number of FieldMask objects and returns the intersection of their paths. When combining, if one FieldMask has a parent field and another has one of its children, only the child will be returned since it is the only path that intersects both:

import { fieldMask, intersectFieldMasks } from '@protoutil/core';

const one = fieldMask(MySchema, 'a');
const two = fieldMask(MySchema, 'a.b', 'c');
intersectFieldMasks(one, two); // returns ['a.b'] as paths in the `FieldMask`

Int32

The assertValidInt32 and isValidInt32 functions validate that number values are 32-bit integers:

import { assertValidInt32, isValidInt32 } from '@protoutil/wkt';

assertValidInt32(num); // throws if `num` is not a 32 bit integer
isValidInt32(num); // return true if `num` is a 32-bit integer or false otherwise

Int64

The assertValidInt64 and isValidInt64 functions validate that BigInt values are 64-bit integers:

import { assertValidInt64, isValidInt64 } from '@protoutil/wkt';

assertValidInt64(num); // throws if `num` is not a 64 bit integer
isValidInt64(num); // return true if `num` is a 64-bit integer or false otherwise

Timestamp

A Timestamp represents a point in time independent of any time zone or local calendar, encoded as a count of seconds and fractions of seconds at nanosecond resolution. The count is relative to an epoch at UTC midnight on January 1, 1970, in the proleptic Gregorian calendar which extends the Gregorian calendar backwards to year one.

All minutes are 60 seconds long. Leap seconds are "smeared" so that no leap second table is needed for interpretation, using a 24-hour linear smear.

The range is from 0001-01-01T00:00:00Z to 9999-12-31T23:59:59.999999999Z. By restricting to that range, we ensure that we can convert to and from RFC 3339 date strings.

The timestamp function creates and validates a Timestamp message:

import { timestamp } from '@protoutil/core';

timestamp(1n, 1_000_000); // returns a timestamp object representing 1.001 seconds after the unix epoch
timestamp(253402300800n); // throws an error as timestamps have a max value of 9999-12-31T23:59:59.999999999Z

The assertValidTimestamp and isValidTimestamp functions validate Timestamp objects:

import { assertValidTimestamp, isValidTimestamp } from '@protoutil/core';

assertValidTimestamp(ts); // throws an error if `ts` is not a valid timestamp
isValidTimestamp(d); // returns true if `ts` is a valid timestamp and false otherwise

The timestampFromDateString and timestampDateString functions convert strings to and from Timestamp protobuf messages:

import { timestampFromDateString, timestampDateString } from '@protoutil/core';

timestampFromDateString('1970-01-01T02:07:34.000000321+07:00'); // returns a `Timestamp` message representing the unix epoch
timestampDateString(ts); // returns a string representation i.e. '1970-01-01T00:00:00.000000000Z'

The timestampFromNanos and timestampNanos functions convert nanoseconds to and from Timestamp protobuf messages (useful for performing math operations on durations and timestamps):

import { timestampFromNanos, timestampNanos } from '@protoutil/core';

timestampFromNanos(1_000_000n); // returns a `Timestamp` message representing 1 millisecond after Jan 1, 1970
timestampNanos(d); // returns a BigInt i.e. 1_000_000n

// Math example:
const epochTimestamp = timestampFromNanos(0n);
const epochTimestampNanos = timestampNanos(epochTimestamp);
const oneWeekDuration = durationFromString(`${7 * 24 * 60 * 60}s`);
const oneWeekDurationNanos = durationNanos(oneWeekDuration);
const oneWeekAfterEpoch = timestampFromNanos(epochTimestampNanos + oneWeekDurationNanos);
timestampDateString(oneWeekAfterEpoch); // returns '1970-01-08T00:00:00.000Z'

The roundTimestampNanos function is a helper to make sure that the nanos parameter of a Timestamp is an integer. This can be helpful if you need to perform calculations then validate a Timestamp.

import { roundTimestampNanos, assertValidTimestamp } from '@protoutil/core';

let ts = create(TimestampSchema, { nanos: 3 / 2 });
ts = roundTimestampNanos(ts);
assertValidTimestamp(ts); // should not throw

The clampTimestamp function can be used to ensure that a Timestamp is between two specified values. By default, clampTimestamp uses the minimum and maximum values as defined in the protobuf spec (0001-01-01T00:00:00Z & 9999-12-31T23:59:59.999999999Z).

import { clampTimestamp } from '@protoutil/core';

let min = timestamp(5n);
let max = timestamp(10n);
clampTimestamp(timestamp(15n), min, max); // returns the max value
clampTimestamp(timestamp(1n), min, max); // returns the min value
clampTimestamp(timestamp(7n), min, max); // returns the original value
clampTimestamp(timestamp(-62135596801n)); // returns a min (0001-01-01T00:00:00Z) timestamp
clampTimestamp(timestamp(253402300800n)); // returns a max (9999-12-31T23:59:59.999999999Z) timestamp

Temporal Functions

Temporal is a Stage 3 TC39 proposal which has begun shipping in experimental releases of browsers. Since support is still experimental, we use the temporal-polyfill. Using Temporal instead of Date means native support for nanosecond resolution and simplified operations when working with calendar dates, time zones, date/time calculations, and more. Read more about the Temporal API here.

The temporalTimestampNow function returns a Timestamp object representing the current time using the Temporal API:

import { temporalTimestampNow } from '@protoutil/core';

temporalTimestampNow(); // returns a `Timestamp` object representing the current time with nanosecond resolution

The timestampFromInstant and timestampInstant functions convert Temporal.Instant objects to and from Timestamp protobuf messages:

import { timestampFromInstant, timestampInstant } from '@protoutil/core';

timestampFromInstant(instant); // returns a `Timestamp` object representing the `Temporal.Instant`
timestampInstant(ts); // returns a `Temporal.Instant` object representing the `Timestamp`

UInt32

The assertValidUInt32 and isValidUInt32 functions validate that number values are 32-bit unsigned integers:

import { assertValidUInt32, isValidUInt32 } from '@protoutil/wkt';

assertValidUInt32(num); // throws if `num` is not a 32 bit unsigned integer
isValidUInt32(num); // return true if `num` is a 32-bit unsigned integer or false otherwise

UInt64

The assertValidUInt64 and isValidUInt64 functions validate that BigInt values are 64-bit unsigned integers:

import { assertValidUInt64, isValidUInt64 } from '@protoutil/wkt';

assertValidUInt64(num); // throws if `num` is not a 64 bit unsigned integer
isValidUInt64(num); // return true if `num` is a 64-bit unsinged integer or false otherwise

Contributing

Building

Run nx build core to build the library.

Running unit tests

Run nx test core to execute the unit tests via Vitest.

Changelog

0.2.5 (2025-07-02)

🚀 Features

• cel: initial commit; moving from @bearclaw to ease development (#1)
• cel: add residual ast functionality (5a42b4f)
• cel: add inlining (92b85cd)
• cel: add folding optimizer (f6b0006)
• cel-policy-agent: initial commit (0901db8)
• cel-policy-agent: policy testing (717d34c)
• celql: initial commit (#2)
• celql: support translation from compiled ASTs (2c5e73d)

🩹 Fixes

• cel: fix issue with extended environments; more exports to aid in development (df5f92f)
• cel: expose registry types and descriptors (72068bd)
• cel: fix for unresolved unknown attributes (c30fe02)
• cel: various fixes for issues with inlining (1d92256)

❤️ Thank You

• jafaircl