---
title: "Encoding and utilities"
description: "estimateSegments, isGsm7, the GSM-7 tables, SEGMENT_LIMITS, isE164, maskPhoneNumber, redactText, and the other root helpers."
---

Synchronous helpers with no network access. The encoding functions live in `@opencoredev/sms-sdk/encoding`. `estimateSegments` and `isGsm7` are also exported from the root.

## `estimateSegments(body)`

Estimates the encoding and segment count for a message body.

```ts
import { estimateSegments } from "@opencoredev/sms-sdk/encoding";

const preview = estimateSegments("Café opens at 9 € 5");
console.log(preview.encoding, preview.units, preview.segments); // "gsm7" 20 1
```

`é` is in the GSM-7 basic table and `€` is in the extension table, so it counts 2 septets.

### Returns: `SegmentPreview`

| Field | Type | Meaning |
| --- | --- | --- |
| `encoding` | `"gsm7" \| "ucs2"` | GSM-7 when every character is in the basic or extension table. |
| `segments` | `number` | Estimated segments. `0` for an empty body. |
| `units` | `number` | Septets (extension characters count 2) or UTF-16 code units. |
| `unitsPerSegment` | `number` | 160 or 153 for GSM-7, 70 or 67 for UCS-2. |
| `remainingInSegment` | `number` | Units left before another segment is needed. |
| `containsUnicode` | `boolean` | At least one character forced UCS-2. |
| `nonGsmCharacters` | `readonly string[]` | Distinct characters that forced UCS-2, in order of first appearance. |

Extension characters and surrogate pairs never split across segments, and nothing is transliterated. The result is an estimate, not a price. See [Segments and billing](/sending/segments-and-billing).

## `isGsm7(text)`

Returns `true` when every character is in the GSM-7 basic or extension table.

## `GSM7_BASIC_CHARACTERS` and `GSM7_EXTENSION_CHARACTERS`

Strings holding the GSM 03.38 basic alphabet, one septet each, and the extension table, two septets each. The extension table is form feed, `^`, `{`, `}`, `\`, `[`, `~`, `]`, `|`, and `€`.

## `SEGMENT_LIMITS`

```ts ignore="value listing"
{ gsm7: { single: 160, concatenated: 153 }, ucs2: { single: 70, concatenated: 67 } }
```

Exported from `/encoding` only.

## `isE164(value)`

```ts
import { isE164 } from "@opencoredev/sms-sdk";

console.log(isE164("+14155550123"), isE164("+1 415 555 0123"), isE164("0044207946000")); // true false false
```

A type guard (`value is E164`) for `^\+[1-9]\d{1,14}$`. It checks format only, not whether the number exists or receives SMS. It accepts `unknown`, so you can pass `process.env` values directly.

## `maskPhoneNumber(value)`

Masks a number for logs, keeping the `+`, the first digit, and the last two digits: `+14155550123` becomes `+1********23`. Short strings are fully masked.

## `redactText(text)`

Makes provider text safe to log. It masks digit runs that look like phone numbers, removes control characters, and truncates to 200 characters. Use it in [custom adapters](/providers/writing-an-adapter) for `details.message`.

```ts
import { maskPhoneNumber, redactText } from "@opencoredev/sms-sdk";

console.log(maskPhoneNumber("+14155550123")); // "+1********23"
console.log(redactText("The 'To' number +14155550123 is not valid")); // "The 'To' number +1********23 is not valid"
```

## `resolveSender(from)` and `supportsSender(capabilities, sender)`

`resolveSender` validates an `SmsFrom` and returns `{ kind: "ok", sender: SmsSender }`, `{ kind: "missing" }`, or `{ kind: "invalid", message }`. `supportsSender` returns whether an adapter's `capabilities.senderTypes` allow that sender. The client uses both. They are exported for adapter authors.

## `isSmsError(value)` and `isFallbackEligible(category)`

`isSmsError` is a type guard for any SDK error. `isFallbackEligible` returns `true` for `auth`, `rate_limited`, `sender`, and `account`. See [Errors](/reference/errors).

## Types

`E164`, `SegmentPreview`, `SmsEncoding`, and `SenderResolution` are exported as types from `@opencoredev/sms-sdk`.