[mlir:PDLL] Fix error handling of eof within a string literal

[rriddle]

Summary

We currently aren't handling this properly, and in the case
of a string block just crash. This commit adds proper error handling
and detection for eof.

Depends on D124584

mlir/lib/Tools/PDLL/Parser/Lexer.cpp

if (curPtr == codeCompletionLocation) {
       return formToken(Token::code_complete_string,
                        tokStart + (isStringBlock ? 2 : 1));
     }
 
     switch (*curPtr++) {
       case '"':
         // If this is a string block, we only end the string when we encounter a
         // `}]`.
-        if (!isStringBlock) return formToken(Token::string, tokStart);
+        if (!isStringBlock)
+          return formToken(Token::string, tokStart);
         continue;
       case '}':
         // If this is a string block, we only end the string when we encounter a
         // `}]`.
-        if (!isStringBlock || *curPtr != ']') continue;
+        if (!isStringBlock || *curPtr != ']')
+          continue;
         ++curPtr;
         return formToken(Token::string_block, tokStart);
-      case 0:
+      case 0: {
         // If this is a random nul character in the middle of a string, just
-        // include it.  If it is the end of file, then it is an error.
-        if (curPtr - 1 != curBuffer.end()) continue;
-        LLVM_FALLTHROUGH;
+        // include it. If it is the end of file, then it is an error.
+        if (curPtr - 1 != curBuffer.end())
+          continue;
+        --curPtr;
+
+        StringRef expectedEndStr = isStringBlock ? "}]" : "\"";
+        return emitError(curPtr - 1,
+                         "expected '" + expectedEndStr + "' in string literal");
+      }
+
       case '\n':
       case '\v':
       case '\f':
         // String blocks allow multiple lines.
         if (!isStringBlock)
           return emitError(curPtr - 1, "expected '\"' in string literal");
         continue;
 

mlir/lib/Tools/PDLL/Parser/Parser.cpp

FailureOr<T *> Parser::parseUserNativeConstraintOrRewriteDecl(
   Optional<StringRef> optCodeStr;
   if (curToken.isString()) {
     codeStrStorage = curToken.getStringValue();
     optCodeStr = codeStrStorage;
     consumeToken();
   } else if (isInline) {
     return emitError(name.getLoc(),
                      "external declarations must be declared in global scope");
+  } else if (curToken.is(Token::error)) {
+    return failure();
   }
   if (failed(parseToken(Token::semicolon,
                         "expected `;` after native declaration")))
     return failure();
   // TODO: PDL should be able to support constraint results in certain
   // situations, we should revise this.
   if (std::is_same<ast::UserConstraintDecl, T>::value && !results.empty()) {
     return emitError(

mlir/test/mlir-pdll/Parser/string-eof.pdll

+// RUN: not mlir-pdll %s -I %S -split-input-file 2>&1 | FileCheck %s
+
+// CHECK: expected '}]' in string literal
+Constraint Cst() [{
+
+// -----
+
+// CHECK: expected '"' in string literal
+Constraint Cst() "
 No newline at end of file