Skip to main content

SqidDecodeLower

Function SqidDecodeLower( const aToken : string) : Int64

Example​

procedure ScriptEvent(var Value: Variant);
var
  Token: String;
begin
  Token := SqidEncodeLower(1001, 8);
  Value := SqidDecodeLower(Token); // 1001
end;

Usage​

SqidDecodeLower decodes one lowercase-and-digit Velox Sqid token to its Int64 value.

Parameters

NameTypeDescription
aTokenstring, constCase-sensitive token to decode with the exact alphabet abcdefghijklmnopqrstuvwxyz0123456789.

Returns

The single decoded value as Int64. A token created by SqidEncodeLower from a supported value round-trips to that value.

Behaviour

  • Only the lowercase-and-digit alphabet, in this exact order, is used.
  • Tokens are case-sensitive; uppercasing a token changes it or makes it invalid.
  • Padding introduced by a minimum length does not change the decoded number.
  • Decoding is deterministic for the Velox version and alphabet.

Errors

ESqidsException, range-check, overflow, allocation and script conversion failures propagate. Invalid input is not converted to zero or -1.

Usage notes

Use the lowercase encoder and decoder as a pair. Validate the decoded entity, tenant and caller permissions separately; a valid token is only a reversible representation.

Additional Technical Info​

SqidDecodeLower reverses one token created with Velox's lowercase-and-digit Sqids alphabet and returns its non-negative numeric value as Int64. It is useful where a transport requires lowercase identifiers. The token remains reversible and should not contain sensitive information.

The example is source-reviewed and was not executed by the documentation workflow.

Implementation

Velox creates the shipped TSqids class with the ordered 36-character alphabet abcdefghijklmnopqrstuvwxyz0123456789, minimum length zero and the shipped default blocklist. It calls DecodeSingle, which requires the decoder to produce exactly one UInt64, converts the result to Int64, and frees the instance in a finally block.

Edge cases and quirks

  • Empty text, an out-of-alphabet character, or a token yielding zero or multiple numbers causes an exception rather than a sentinel result.
  • A numeric result does not prove that a token is the one canonical value emitted by the encoder. Decode/re-encode/compare exact text when canonical form matters.
  • The blocklist is not a decode-time rejection list. Some non-canonical or blocklisted alphabet-valid strings can still decode.
  • Long or adversarial input can overflow the port's UInt64 calculation; a value above MaxInt64 can fail the wrapper conversion.
  • The output of the paired encoder is random-looking but exposes the number to any party using the same algorithm and alphabet.

Side effects

None beyond a per-call Sqids object and temporary arrays.

Performance and concurrency

Work scales mainly with token length, plus per-call alphabet/blocklist setup. Bound untrusted token lengths. Separate calls own separate instances and share no mutable instance state.

Related entries

External references

  • Sqids FAQ - official guidance on case-sensitive alphabets, padding, reversibility and validation.
Created 2026-07-15