Performance, Ordered Collections release

[danlentz]

This branch turns the original interval-tree codebase into a broader ordered-collections library with a more complete API, stronger semantics, significantly expanded documentation, and much deeper testing/benchmarking. It adds several new collection types, rounds out the ordered-set/map interfaces, hardens performance-sensitive internals, and raises the standard of correctness around equality, serialization, EDN, interop, and benchmarks.

• Rebrand/reorganize the library from interval-tree to ordered-collections
• Add a public core API, shared protocols, readers, util/parallel namespaces, and cleaner constructor/re-export structure
• Expand the collection lineup with range-map, segment-tree, priority-queue, ordered-multiset, fuzzy-set, and fuzzy-map
• Strengthen ordered-set / ordered-map with richer ordered-collection operations and better Java/Clojure interface support
• Add specialized constructors and comparator infrastructure for long, double, string, and custom comparator variants
• Add EDN tagged literals / reader support for the collection types
• Add Java serialization support and test coverage across supported collection types
  • Rework parallel set/map execution:
  • explicit-argument hot paths to avoid dynamic binding dependence in forked work
  • lower-overhead RecursiveTask tasking
  • per-operation threshold tuning
  • better benchmark methodology for parallel decisions
• Improve fold support and measurement:
  • node-chunked-fold benchmarking at multiple cardinalities
  • nontrivial fold workload benchmarks
  • comparisons against sorted-set, data.avl, and hash-set
• Greatly expand test coverage:
  • equality / hashing
  • cross-type equivalence / interop
  • coverage tests
  • cookbook and example tests
  • memory tests
  • range-map equivalence tests
  • segment-tree tests
  • priority-queue tests
• Build out benchmark infrastructure:
  • full Criterium runner
  • simple bench runner
  • parallel-threshold bench
  • memory bench
  • range-map bench
• Harden segment-tree:
  • restore generic ordered-key support
  • add broader ordered-collection/API parity
  • add stronger generic-key regression tests
• Harden priority-queue:
  • clarify comparator-relative queue-order semantics
  • strengthen duplicate-priority ordering behavior/tests
  • replace lazy seq wrappers with direct seq adapters
  • re-enable real serialization coverage
• Expand and refresh docs:
  • README overhaul
  • algorithms / performance / benchmarks / competitive analysis docs
  • cookbook and example docs
  • “when to use” and comparison docs
  • research/reference material under doc/papers
• Remove generated doc/api artifacts from the repo and ignore them going forward
• Update changelog and benchmark-sensitive docs to match current behavior