Changelog

Version 1.0.0 (February 10, 2023)

Migrated primer3 source to 2.6.1 version
Intergrated all Python C-API related code into thermoanalysis.pyx/pxd
Deprecated camelCase related functions and added snake_case functions / methods in bindings.py and thermoanalysis.pyx
migrated primer3/wrappers.py to tests/wrappers.py to indicate it should not be used.
Expanded test coverage and now using pytest for testing
Added pre-commit hooks
Using primer3 read_boulder_record to setup primer3 data structures instead of reproducing code in primer3-py C or Cython
Intergrated GitHub Actions for CI
Updated documentation
Begin process of making primer3-py code thread-safe
Documentation migrated to MyST Markdown

Added functionality for bindings to expose the ASCII representation of computed dimer and hairpin structures

Changes to upper case sequences in seqtm() in oligotm.c because calls to ThermoAnalysis.calcTm returned 0.0 for strings with lower case letters
disabled existing tests on Windows due to incompatibility with process and memory calls. TODO: create more tests for ThermoAnalysis

Fixes to primerdesign_helpers.c C macro errors.
Added the todict()methods inthermoanalysis.pyx for serialization help for tests there

Migration of Cython extension build process to build_clib to remove Cython as a runtime requirement for setup.py (thanks to @tdsmith)

Fixed improper assignment of "PRIMER_PAIR_MAX_COMPL_END_TH" during design arg parsing

Update thal.c to incorporate traceback fix from Primer3 dev team (thanks Triinu)

Better documentation regarding max. seq. lengths for thermo. calcs
calcHairpin, calcHomodimer, calcHeterodimer, and calcEndStability now raise RuntimeErrors as appropriate

Fixed memory leak in global param struct allocation for design bindings
Integrated C API patch code into libprimer3
Refactored libprimer3 to pure C (khash lib is now used in lieu of hashmap)
Corrected many minor compiler warnings in libprimer3
(POTENTIAL BREAKING CHANGE) Refactored thermodynamic alignment bindings to always return a thermoresult namedtuple (in the previous embodiment, if no structure was found, the return value was None)

Fixed incorrect tm/salt correction method lookups in calcTm function in both bindings and wrappers

Addressed bug in mispriming/mishybridization library parsing in which memory for the global parameters struct was not being properly allocated

Corrected several potential sources of memory leaks in the C API design bindings.

Corrected issue with parsing "SEQUENCE_PRIMER_PAIR_OK_REGION_LIST" as a python sequence object. "SEQUENCE_PRIMER_PAIR_OK_REGION_LIST" should now be provided as a list of lists or tuple of tuples (or a combination of the two). -1 should be used in the case where an index is ommitted. For example:
```
Primer3 boulder IO input:   100,50,300,50 ; 900,60,,
Primer3 python input:       [[100,50,300,50], [900,60,-1,-1]]
```