Add dense fixture and shove tests

Adds shove_needed.dsn: one signal layer (no vias), where NET_A routes as
a straight wall that boxes NET_C out under greedy ordering. Tests assert
the contrast: without shove NET_C is dropped (result still DRC-clean, just
incomplete); with shove NET_A is moved aside and both nets connect, the
exact has_violation check is clean, endpoints stay on pads, and the result
is deterministic. Regression tests confirm shove is a no-op on the simple/
crossing/ripup/kicad boards (same connectivity, DRC-clean, zero shoves).
Oracle-gated parity is included (skips when the JAR cannot route the
synthetic single-layer board).

tests/dsn/fixtures/shove_needed.dsn

@@ -0,0 +1,64 @@
+(pcb "shove_needed.dsn"
+  (parser
+    (string_quote ")
+    (space_in_quoted_tokens on)
+    (host_cad "freeroute-test")
+    (host_version "1.0")
+  )
+  (resolution um 10)
+  (unit um)
+  (structure
+    (layer F.Cu
+      (type signal)
+      (property
+        (index 0)
+      )
+    )
+    (layer B.Cu
+      (type power)
+      (property
+        (index 1)
+      )
+    )
+    (boundary
+      (path pcb 0  0 0  200000 0  200000 -30000  0 -30000  0 0)
+    )
+    (rule
+      (width 2000)
+      (clearance 2000)
+    )
+  )
+  (library
+    (padstack Rect_Pad
+      (shape (rect F.Cu -1000 -1000 1000 1000))
+      (attach off)
+    )
+    (image PAD
+      (pin Rect_Pad 1 0 0)
+    )
+  )
+  (placement
+    (component PAD
+      (place A1 3000 -15000 front 0)
+      (place A2 197000 -15000 front 0)
+      (place C1 100000 -6000 front 0)
+      (place C2 100000 -24000 front 0)
+    )
+  )
+  (network
+    (net NET_A
+      (pins A1-1 A2-1)
+    )
+    (net NET_C
+      (pins C1-1 C2-1)
+    )
+    (class default
+      (rule
+        (width 2000)
+        (clearance 2000)
+      )
+    )
+  )
+  (wiring
+  )
+)