Skip to main content

ZLibDecompress

Function ZLibDecompress( const Data : TBytes) : TBytes

Example​

procedure ScriptEvent(var Value: variant);
var
  Compressed: TBytes;
begin
  Compressed := ZLibCompressString('Fictional payload');
  Value := Base64EncodeBytes(ZLibDecompress(Compressed));
end;

Usage​

ZLibDecompress expands a zlib-wrapped byte array into its original bytes.

Parameters

NameTypeDescription
DataTBytes, constComplete zlib stream.

Returns

The expanded byte sequence in a newly allocated TBytes value.

Behaviour

A successful result is binary data. If it represents a string, record the encoding separately or use ZLibDecompressString only for the exact UTF-16LE contract of the matching Velox compressor.

Errors

Corrupt headers, checksums, truncated input, extra input, dictionaries and other zlib failures raise EZDecompressionError. Allocation and range errors propagate. The RTL clears the result on its protected exception path.

Additional Technical Info​

ZLibDecompress expands a complete zlib-wrapped deflate stream and returns its original bytes. It does not decode text or read ZIP/gzip containers.

The example represents the expanded UTF-16LE bytes as Base64. It is source-reviewed and is not executed by the documentation workflow.

Implementation

The direct registration calls vxCompression.ZLibDecompress, which selects System.ZLib.ZDecompress(const inBuffer: TBytes; out outBuffer: TBytes; outEstimate) with estimate 0. The installed Studio 37.0 RTL uses the compressed length rounded to 256 bytes as its initial size and growth increment, calls InflateInit, and checks both progress and trailing/unconsumed input before accepting Z_STREAM_END.

Edge cases and quirks

  • Empty input raises EZDecompressionError with a zlib buffer-error message; it does not return empty bytes.
  • A gzip member, raw deflate stream or ZIP archive is a wrapper mismatch and normally raises a decompression error.
  • The RTL rejects a stream that ends while compressed input remains, so concatenated streams or trailing bytes are not silently accepted by this overload.
  • Output size is not capped. A small valid stream can allocate a large result.
  • Buffer sizes and zlib counters are 32-bit; very large results can fail or overflow before completion.

Side effects

No external side effect. The function allocates expanding output memory.

Performance and concurrency

The complete compressed and expanded arrays coexist. Work is synchronous and depends on compressed and expanded sizes. State is local, but untrusted compression ratios can exhaust process resources.

Remarks

Apply an expected maximum expanded size before accepting external compressed data. This API cannot enforce that maximum during inflate, so high-risk inputs should be handled by a bounded integration component rather than directly in a Velox script.

Related entries

External references

Created 2026-07-15