Ok.

src/arm64/dom_parser_implementation.cpp

@@ -105,7 +105,7 @@ WARN_UNUSED error_code dom_parser_implementation::stage1(const uint8_t *_buf, si
   this->len = _len;
   return arm64::stage1::json_structural_indexer::index<64>(buf, len, *this, streaming);
 }
-
+#include "generic/stage1/utf8_validator.h"
 } // namespace arm64
 } // namespace simdjson
 

src/generic/stage1/buf_block_reader.h

@@ -6,6 +6,9 @@ public:
   really_inline size_t block_index();
   really_inline bool has_full_block() const;
   really_inline const uint8_t *full_block() const;
+  really_inline bool has_remainder() const {
+    return idx < len;
+  }
   /**
    * Get the last block, padded with spaces.
    *

src/generic/stage1/utf8_validator.h

@@ -0,0 +1,29 @@
+namespace stage1 {
+/**
+ * Validates that the string is actual UTF-8.
+ */
+template<class checker>
+bool utf8_validate(const uint8_t * input, size_t length) {
+    checker c{};
+    buf_block_reader<64> reader(input, length);
+    while (reader.has_full_block()) {
+      simd::simd8x64<uint8_t> in(reader.full_block());
+      c.check_next_input(in);
+      reader.advance();
+    }
+
+    if (likely(reader.has_remainder())) {
+      uint8_t block[64]{};
+      reader.get_remainder(block);
+      simd::simd8x64<uint8_t> in(block);
+      c.check_next_input(in);
+      reader.advance();
+    }
+    return c.errors() == error_code::SUCCESS;
+}
+
+bool utf8_validate(const char * input, size_t length) {
+    return utf8_validate<utf8_checker>((const uint8_t *)input,length);
+}
+
+} // namespace stage1

src/haswell/dom_parser_implementation.cpp

@@ -93,7 +93,7 @@ WARN_UNUSED error_code dom_parser_implementation::stage1(const uint8_t *_buf, si
   this->len = _len;
   return haswell::stage1::json_structural_indexer::index<128>(_buf, _len, *this, streaming);
 }
-
+#include "generic/stage1/utf8_validator.h"
 } // namespace haswell
 } // namespace simdjson
 UNTARGET_REGION

src/westmere/dom_parser_implementation.cpp

@@ -94,7 +94,7 @@ WARN_UNUSED error_code dom_parser_implementation::stage1(const uint8_t *_buf, si
   this->len = _len;
   return westmere::stage1::json_structural_indexer::index<64>(_buf, _len, *this, streaming);
 }
-
+#include "generic/stage1/utf8_validator.h"
 } // namespace westmere
 } // namespace simdjson
 UNTARGET_REGION