Fix floating point exception and add --continue-on-error flag

[mosoriob]

Summary

Fixes floating point exception crashes in PotreeConverter and adds a new --continue-on-error flag for robust batch processing.

Problem

PotreeConverter was crashing with Floating point exception (core dumped) when processing certain LAS/LAZ files, particularly when:

• Scale factors in LAS headers are zero (causing division by zero)
• Bounding box dimensions are zero
• Invalid coordinate transformations occur

This made batch processing of large datasets unreliable - a single bad file would crash the entire process.

Solution

🔧 Floating Point Exception Prevention

• Added validation checks for zero scale factors and size values before division operations
• Implemented try-catch blocks around critical floating point operations
• Enhanced error messages with detailed debugging information (coordinates, scales, offsets, file paths)

🚀 Robust Batch Processing

• New --continue-on-error flag allows processing to continue even when individual files fail
• Graceful error handling - failed files are logged and skipped rather than crashing the entire process
• Maintains data integrity - only processes files that can be safely converted

Usage

# Standard behavior (fails on first error)
PotreeConverter input/ -o output/

# New robust mode (continues processing despite errors)  
PotreeConverter input/ -o output/ --continue-on-error

Testing

• Tested with the failing file /data/ibwc06-70cm_2697614c3.laz_converted
• Verified error detection and descriptive messages
• Confirmed graceful skipping with --continue-on-error flag
• Validated normal processing continues for valid files

Files Modified

• Converter/src/main.cpp - Added command line argument and options handling
• Converter/include/converter_utils.h - Extended Options struct
• Converter/include/chunker_countsort_laszip.h - Updated function signature
• Converter/src/chunker_countsort_laszip.cpp - Core validation and error handling logic

Benefits

• 🛡️ Prevents crashes from invalid LAS files
• 📊 Enables large-scale processing of mixed-quality datasets
• 🔍 Better debugging with detailed error information
• ⚡ Maintains performance for valid files
• 🎯 Backward compatible - existing workflows unchanged

Closes #3