Refactor HeatShockTransformation transfer logic and add protocol generator improvements

[Aaron1776]

transformation.py:

• Introduce self.plasmid_name_to_wells (name → [well_obj, ...]) populated at load time, replacing all positional/enumeration-based DNA well lookups
• Introduce self.location_replicates derived from plasmid_locations, unifying 96-well plate and temp module code paths into a single loop structure
• Rewrite _transfer_competent_cells with sequential well_index counter and per-chassis chassis_reaction_count dict for correct tube selection across multiple chassis types
• Rewrite _transfer_DNA as a unified single method with no plate/temp module branching; uses self.transformations as ground truth throughout
• _load_dna_into_dna_plate / _load_dna_into_temp_module now populate plasmid_name_to_wells instead of returning DNA_wells
• _load_reagents_96plate / _load_reagents_temp_module return 2-tuple; _validate_protocol accounts for location_replicates in reaction counts
• Add strain as first entry in dict_of_parts_in_thermocycler per well
• Rewrite Transformation base class docstring to match current parameters; add comprehensive HeatShockTransformation docstring for all subclass params

generate_protocol.py:

• Add generate_param_reference() to append a commented parameter table and full docstrings to every generated protocol for last-minute editing
• Add --plasmid-locations CLI flag and plasmid_locations support throughout
• Raise line-length threshold from 80 to 100 for single-line list formatting
• Improve transformation protocol auto-detection for bare list format

assembly.py:

• Initialize self.product_uri_to_wells = {}