2026-10-26

• The encoders were fixed to leave nothing on the stack when an option is given a value of the wrong type.
• Option passing was improved.
• The performance of most encoders, and of GS1 data processing, input parsing and the Reed-Solomon caches, was improved by sizing their working dictionaries for the entries that they hold.
• The Data Matrix encoder was fixed to make Reader Programming the first codeword, which a latch to C40 could precede.
• The GS1 encoders were fixed to refuse bracketed AI data that ends with an opening parenthesis, which was dropped.
• The Ultracode encoder was fixed to count data positions without ECIs, which let FNC1 in second data position through after a leading ECI and misreported FNC1 in first position after one.
• The DotCode encoder was fixed to keep symbols within its limits of 200 rows and 200 columns, which automatic sizing and fixed rows or columns exceeded, holding a dimension at its limit where the data fits and refusing the data otherwise.
• The DotCode encoder was enhanced with the dotcode.maxrows and dotcode.maxcols global context settings, which change the limits on the rows and columns of a symbol.
• The 2D and stacked encoders were enhanced to report their choices of version, error correction level, mask, columns, rows, format, layers and mode, and the composite encoders their choices of ccversion and cccolumns, in the report dictionary, as the values of the options that select them.
• The Data Matrix encoder was enhanced with the iso144 option, which places the error correction codewords of 144x144 symbols in the order of the ISO/IEC standard.
• The Data Matrix, Code 128, Code 16K, Codablock F, Code One, DotCode, Ultracode and PosiCode encoders were enhanced with the readerinit option, which marks the symbol for programming the reader, and the Data Matrix, Ultracode, PDF417 and MicroPDF417 encoders accept it with raw codewords.