Levenshtein-CFL intersection using Bar-Hillel construction

breandan · Oct 11, 2023 · 7df6f82 · 7df6f82
1 parent c20e49e
commit 7df6f82
Show file tree

Hide file tree

Showing 3 changed files with 31 additions and 14 deletions.
diff --git a/src/commonMain/kotlin/ai/hypergraph/kaliningraph/parsing/BarHillel.kt b/src/commonMain/kotlin/ai/hypergraph/kaliningraph/parsing/BarHillel.kt
@@ -3,7 +3,7 @@ package ai.hypergraph.kaliningraph.parsing
 import ai.hypergraph.kaliningraph.types.*
 
 infix fun FSA.intersect(cfg: CFG) = cfg.intersect(this)
-// http://www.cs.umd.edu/~gasarch/BLOGPAPERS/cfg.pdf
+// http://www.cs.umd.edu/~gasarch/BLOGPAPERS/cfg.pdf#page=2
 // https://browse.arxiv.org/pdf/2209.06809.pdf#page=5
 
 infix fun CFG.intersect(fsa: FSA): CFG {
@@ -19,7 +19,7 @@ infix fun CFG.intersect(fsa: FSA): CFG {
     nonterminalProductions.map {
       val triples = fsa.states * fsa.states * fsa.states
       val (A, B, C) = it.π1 to it.π2[0] to it.π2[1]
-      triples.map { (p, q, r) -> "[$p,$A,$r] -> [$q,$B,$q] [$q,$C,$r]" }
+      triples.map { (p, q, r) -> "[$p,$A,$r] -> [$p,$B,$q] [$q,$C,$r]" }
     }.flatten()
 
   // For every production A → σ in P, for every (p, σ, q) ∈ Q × Σ × Q
@@ -32,7 +32,8 @@ infix fun CFG.intersect(fsa: FSA): CFG {
 
   return (initFinal + transits + binaryProds + unitProds).joinToString("\n")
     .parseCFG(normalize = false)
-//    .also { println(it.pretty) }
-//    .removeVestigalProductions()
+    .removeVestigalProductions()
+    .normalForm.noNonterminalStubs
+    .also { println(it.pretty) }
 //    .also { println(it.size) }
 }
diff --git a/src/commonMain/kotlin/ai/hypergraph/kaliningraph/parsing/SetValiant.kt b/src/commonMain/kotlin/ai/hypergraph/kaliningraph/parsing/SetValiant.kt
@@ -41,6 +41,9 @@ fun CFG.isValid(str: List<Σᐩ>): 𝔹 =
     //.also { println("Last: ${it.joinToString(",") {if (it) "1" else "0"}}") }
     .let { corner -> corner[bindex[START_SYMBOL]] }
 
+fun CFG.corner(str: Σᐩ) =
+ solveFixedpoint(str.tokenizeByWhitespace())[0].last().map { it.root }.toSet()
+//  START_SYMBOL in solveFixedpoint(str.tokenizeByWhitespace())[0].last().map { it.root }.toSet()
 fun CFG.parseForest(str: Σᐩ): Forest = solveFixedpoint(str.tokenizeByWhitespace())[0].last()
 fun CFG.parseTable(str: Σᐩ): TreeMatrix = solveFixedpoint(str.tokenizeByWhitespace())
 

diff --git a/src/commonTest/kotlin/ai/hypergraph/kaliningraph/parsing/BarHillelTest.kt b/src/commonTest/kotlin/ai/hypergraph/kaliningraph/parsing/BarHillelTest.kt
@@ -1,5 +1,6 @@
 package ai.hypergraph.kaliningraph.parsing
 
+import ai.hypergraph.kaliningraph.levenshtein
 import kotlin.test.*
 import kotlin.time.*
 
@@ -223,21 +224,33 @@ class BarHillelTest {
     val simpleCFG = """
       START -> E
       O -> + | *
-      E -> N O N
+      E -> N O N | E O N
       N -> 1 | 2
     """.parseCFG().noEpsilonOrNonterminalStubs
 
-    val levFSA = makeLevFSA("1 + 1", 1, simpleCFG.terminals)
-//    println(levFSA.graph.toDot())
+    val origStr = "1 + 1"
+    val levFSA = makeLevFSA(origStr, 2, simpleCFG.terminals)
 
-//    println(levFSA.Q.size)
+    val levCFG = levFSA.intersect(simpleCFG)
 
-    val levCFG = (levFSA.intersect(simpleCFG))//.also { print(it.pretty) }
-//    println(levCFG.graph.toDot())
+    fun testLevenshteinAcceptance(s: Σᐩ) {
+      assertTrue(levFSA.recognizes(s))
+      assertTrue(s in simpleCFG.language)
+      assertTrue(s in levCFG.language)
+    }
 
-    val testStr = "1 * 1"
-    assertTrue(levFSA.recognizes(testStr))
-    assertTrue(testStr in simpleCFG.language)
-//    println(levCFG.corner(testStr))
+    val neighbor = "1 * 2"
+    assertEquals(2, levenshtein(origStr, neighbor))
+    testLevenshteinAcceptance(neighbor)
+
+    val foreign = "1 + 1 + 1"
+    testLevenshteinAcceptance(foreign)
+
+    val testFail = "2 * 2"
+    assertFalse(testFail in levCFG.language)
+
+    val template = List(5) { "_" }.joinToString(" ")
+    val solutions = levCFG.solveSeq(template).toList().onEach { println(it) }
+    println("Found ${solutions.size} solutions within Levenshtein distance 2 of \"$origStr\"")
   }
 }