src: remove explicit UTF-8 validity check in url

This step was never part of the URL Standard's host parser algorithm,
and is rendered unnecessary after IDNA errors are no longer ignored.

PR-URL: #11859
Refs: c2a302c "src: do not ignore IDNA conversion error"
Refs: https://url.spec.whatwg.org/#concept-host-parser
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Daijiro Wachi <daijiro.wachi@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>

src/node_url.cc

@@ -15,11 +15,6 @@
#include <stdio.h>
#include <cmath>

#if defined(NODE_HAVE_I18N_SUPPORT)
#include <unicode/utf8.h>
#include <unicode/utf.h>
#endif

#define UNICODE_REPLACEMENT_CHARACTER 0xFFFD

namespace node {
@@ -113,21 +108,6 @@ namespace url {
    output->assign(*buf, buf.length());
    return true;
  }

  // Unfortunately there's not really a better way to do this.
  // Iterate through each encoded codepoint and verify that
  // it is a valid unicode codepoint.
  static bool IsValidUTF8(std::string* input) {
    const char* p = input->c_str();
    int32_t len = input->length();
    for (int32_t i = 0; i < len;) {
      UChar32 c;
      U8_NEXT_UNSAFE(p, i, c);
      if (!U_IS_UNICODE_CHAR(c))
        return false;
    }
    return true;
  }
#else
  // Intentional non-ops if ICU is not present.
  static bool ToUnicode(std::string* input, std::string* output) {
@@ -139,10 +119,6 @@ namespace url {
    *output = *input;
    return true;
  }

  static bool IsValidUTF8(std::string* input) {
    return true;
  }
#endif

  // If a UTF-16 character is a low/trailing surrogate.
@@ -395,12 +371,6 @@ namespace url {
    if (PercentDecode(input, length, &decoded) < 0)
      goto end;

    // If there are any invalid UTF8 byte sequences, we have to fail.
    // Unfortunately this means iterating through the string and checking
    // each decoded codepoint.
    if (!IsValidUTF8(&decoded))
      goto end;

    // Then we have to punycode toASCII
    if (!ToASCII(&decoded, &decoded))
      goto end;