refactor: improve de Bruijn graph traversal and longtig generation

- Refactored Node representation using compact bitfields for neighbor counts
  and nucleotides; added count_neighbors helper to compute_degrees()
- Introduced StartIter iterator for unitig/longtigu generation with revised
  traversal semantics (e.g., interior node marking)
- Added nucleotide() accessor to Kmer type for 2-bit extraction at position i
- Renamed unitig.rs → longtigs, updated CLI command and output filenames to reflect "long t ig"
- Extended extract_kmers() in scripts/compare.py with duplication statistics
```

src/obikseq/src/kmer.rs

@@ -234,6 +234,12 @@ impl CanonicalKmer {
         mix64(self.0.0)
     }
 
+    /// Extract nucleotide i (0-based from 5′ end) as a 2-bit value.
+    #[inline]
+    pub fn nucleotide(&self, i: usize) -> u8 {
+        self.0.nucleotide(i)
+    }
+
     /// Return the four left canonical neighbours (each already canonical).
     /// Zero allocation — result lives on the stack.
     pub fn left_canonical_neighbors(&self, k: usize) -> [CanonicalKmer; 4] {