Fix: Support for chromosomes >1GB in size

[ad3002]

Problem

TRF crashed when processing T2T (telomere-to-telomere) chromosome assemblies longer than 1-2 GB. The errors varied depending on chromosome size, initially appearing to be a genome size issue but actually stemming from individual chromosome length limitations. Example Lissotriton helveticus genome assembly, chromosome: 4_1

The root cause was the use of 32-bit integer types (int, unsigned int) for sequence indices and lengths, which limited processing to ~2GB (2^31 bytes for signed int).

Solution

Replaced 32-bit integer types with 64-bit size_t types throughout the codebase for all sequence-related indices, positions, and lengths. This enables TRF to process sequences up to system memory limits (theoretically 2^64 bytes).

Changes by file

tr30dat.h - Core data type replacements:

• int Length → size_t Length (global sequence length variable)
• unsigned int maxwraplength → size_t maxwraplength
• Updated structures: FASTASEQUENCE, pairalign, TRFPARAMSET, bestperiodlistelement, distanceentry, distancelist, distanceseenarrayelement, distancelistelement

trfrun.h - Index and position variables:

• Updated index_list structure (repeat positions)
• Local variables: charcount, flankstart, flankend
• Functions: LoadSequenceFromFileBenson, LoadSequenceFromFileEugene

trf.c - Command-line parsing:

• Added ParseSize() function using strtoull() for 64-bit values
• Updated -l parameter parsing for maxwraplength

tr30dat.c - Memory allocation:

• Updated new1Darrayfunc macro for correct memory allocation
• Updated format specifiers: %d → %zu for all output

Testing

Successfully processes amphibian chromosomes and T2T assemblies exceeding 1-2 GB without crashes.