Issue 6378 - [emscripten] Make DependentString constexpr

compiled/String.h

 /*
  * This file is part of Adblock Plus <https://adblockplus.org/>,
  * Copyright (C) 2006-present eyeo GmbH
  *
  * Adblock Plus is free software: you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 3 as
  * published by the Free Software Foundation.
  *
  * Adblock Plus is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
@@ skipping 19 matching lines @@
 #include <limits>
 
 #include "base.h"
 #include "debug.h"
 #include "library.h"
 
 ABP_NS_BEGIN
 
 inline void String_assert_writable(bool isWritable);
 
+// hacky because without templates
+#ifdef ABP_UTF8_STRING
+#define ABP_TEXT(val) val
+struct StringTraits
+{
+  typedef char char_type;
+};
+#else
+#define ABP_TEXT(val) u##val
+struct StringTraits
+{
+  typedef char16_t char_type;
+};
+#endif
+
 class String
 {
   friend class DependentString;
   friend class OwnedString;
 
 public:
-  typedef char16_t value_type;
+  typedef StringTraits::char_type value_type;
   typedef size_t size_type;
 
   // Type flags, stored in the top 2 bits of the mLen member
   static constexpr size_type INVALID = 0xC0000000;
   static constexpr size_type DELETED = 0x80000000;
   static constexpr size_type READ_ONLY = 0x40000000;
   static constexpr size_type READ_WRITE = 0x00000000;
 
   static constexpr size_type FLAGS_MASK = 0xC0000000;
   static constexpr size_type LENGTH_MASK = 0x3FFFFFFF;
@@ skipping 32 matching lines @@
   {
     return mBuf;
   }
 
   value_type* data()
   {
     String_assert_writable(is_writable());
     return mBuf;
   }
 
-  const value_type& operator[](size_type pos) const
+  constexpr const value_type& operator[](size_type pos) const
   {
     return mBuf[pos];
   }
 
   value_type& operator[](size_type pos)
   {
     String_assert_writable(is_writable());
     return mBuf[pos];
   }
 
   constexpr bool is_writable() const
   {
     return (mLen & FLAGS_MASK) == READ_WRITE;
   }
 
-  bool equals(const String& other) const
+  constexpr bool equals(const String& other) const
   {
     if (length() != other.length())
       return false;
 
     return std::memcmp(mBuf, other.mBuf, sizeof(value_type) * length()) == 0;
   }
 
-  bool operator==(const String& other) const
+  constexpr bool operator==(const String& other) const
   {
     return equals(other);
   }
 
-  bool operator!=(const String& other) const
+  constexpr bool operator!=(const String& other) const
   {
     return !equals(other);
   }
 
-  size_type find(value_type c, size_type pos = 0) const
+  constexpr size_type find(value_type c, size_type pos = 0) const
   {
     for (size_type i = pos; i < length(); ++i)
       if (mBuf[i] == c)
         return i;
     return npos;
   }
 
-  size_type find(const String& str, size_type pos = 0) const
+  constexpr size_type find(const String& str, size_type pos = 0) const
   {
     return find(str.mBuf, pos, str.length());
   }
 
-  size_type find(const value_type* str, size_type pos, size_type count) const
+  constexpr size_type find(const value_type* str, size_type pos, size_type count) const
   {
     if (pos > LENGTH_MASK || pos + count > length())
       return npos;
 
     if (!count)
       return pos;
 
     for (; pos + count <= length(); ++pos)
     {
       if (mBuf[pos] == str[0] &&
           std::memcmp(mBuf + pos, str, sizeof(value_type) * count) == 0)
       {
         return pos;
       }
     }
 
     return npos;
   }
 
-  size_type rfind(value_type c, size_type pos = npos) const
+  constexpr size_type rfind(value_type c, size_type pos = npos) const
   {
     if (length() == 0)
       return npos;
 
     if (pos >= length())
       pos = length() - 1;
 
     for (int i = pos; i >= 0; --i)
       if (mBuf[i] == c)
         return i;
@@ skipping 12 matching lines @@
 
   void toLower()
   {
     size_type len = length();
     for (size_type i = 0; i < len; ++i)
     {
       value_type currChar = mBuf[i];
 
       // This should be more efficient with a lookup table but I couldn't measure
       // any performance difference.
-      if (currChar >= u'A' && currChar <= u'Z')
-        mBuf[i] = currChar + u'a' - u'A';
+      if (currChar >= ABP_TEXT('A') && currChar <= ABP_TEXT('Z'))
+        mBuf[i] = currChar + ABP_TEXT('a') - ABP_TEXT('A');
       else if (currChar >= 128)
       {
         mBuf[i] = CharToLower(currChar);
       }
     }
   }
 };
 
 #ifdef INSIDE_TESTS
 inline std::ostream& operator<<(std::ostream& os, const String& str)
 {
+#ifdef ABP_UTF8_STRING
+  os.write(str.data(), str.length());
+#else
 #if _MSC_VER >= 1900
   std::wstring_convert<std::codecvt_utf8_utf16<int16_t>, int16_t> converter;
   auto p = reinterpret_cast<const int16_t *>(str.data());
   os << converter.to_bytes(p, p + str.length());
 #else
   std::wstring_convert<std::codecvt_utf8<char16_t>, char16_t> converter;
   os << converter.to_bytes(str.data(), str.data() + str.length());
-#endif
+#endif // _MSC_VER >= 1900
+#endif // ABP_UTF8_STRING
   return os;
 }
-#endif
+#endif // INSIDE_TESTS
 
 class DependentString : public String
 {
 public:
-  explicit DependentString()

[sergei, 2018/03/14 16:22:53]

what about this one too?

[René Jeschke, 2018/03/14 16:27:07]

Is there a reason why we would want invalid nullptr constexpr strings? We can't
change them, so they would be absolutely useless.

If there's a use-case for those, I'll make this one constexpr, too.

[sergei, 2018/03/14 17:01:46]

It's difficult to say right now, and it's also up to a compiler as I guess, e.g.
it's used in containers (Map, Set), so maybe it would be safer to have it
constexpr too, so it does not accidentally prevent e.g. some global constant Map
from being constexpr.

+  constexpr explicit DependentString()
       : String(nullptr, 0, INVALID)
   {
   }
 
   template <int N1>
   constexpr explicit DependentString(const value_type (&buf)[N1])
       : String(const_cast<value_type*>(buf), N1 - 1, READ_ONLY)
   {
   }
 
   explicit DependentString(value_type* buf, size_type len)
       : String(buf, len, READ_WRITE)
   {
   }
 
-  explicit DependentString(const value_type* buf, size_type len)
+  constexpr explicit DependentString(const value_type* buf, size_type len)
       : String(const_cast<value_type*>(buf), len, READ_ONLY)
   {
   }
 
   explicit DependentString(String& str, size_type pos = 0, size_type len = npos)
       : String(
           str.mBuf + std::min(pos, str.length()),
           std::min(len, str.length() - std::min(pos, str.length())),
           str.is_writable() ? READ_WRITE: READ_ONLY
         )
@@ skipping 37 matching lines @@
   }
 };
 
 #ifdef INSIDE_TESTS
 inline std::ostream& operator<<(std::ostream& os, const DependentString& str)
 {
   return os << static_cast<const String&>(str);
 }
 #endif
 
-inline DependentString operator "" _str(const String::value_type* str,
+constexpr DependentString operator "" _str(const String::value_type* str,
     String::size_type len)
 {
   return DependentString(str, len);
 }
 
 inline void String_assert_writable(bool isWritable)
 {
-  assert2(isWritable, u"Writing access to a read-only string"_str);
+  assert2(isWritable, ABP_TEXT("Writing access to a read-only string"_str));
 }
 
 class OwnedString : public String
 {
 private:
   void grow(size_type additionalSize)
   {
     OwnedString newValue(length() + additionalSize);
     if (length() > 0)
       std::memcpy(newValue.mBuf, mBuf, sizeof(value_type) * length());
@@ skipping 77 matching lines @@
     std::swap(mBuf, str.mBuf);
     std::swap(mLen, str.mLen);
     return *this;
   }
 
   void append(const value_type* source, size_type sourceLen)
   {
     if (!sourceLen)
       return;
 
-    assert2(source, u"Null buffer passed to OwnedString.append()"_str);
+    assert2(source, ABP_TEXT("Null buffer passed to OwnedString.append()"_str));
     size_t oldLength = length();
     grow(sourceLen);
     std::memcpy(mBuf + oldLength, source, sizeof(value_type) * sourceLen);
   }
 
+#ifndef ABP_UTF8_STRING
   void append(const char* source, size_type sourceLen)
   {
     if (!sourceLen)
       return;
 
-    assert2(source, u"Null buffer passed to OwnedString.append()"_str);
+    assert2(source, ABP_TEXT("Null buffer passed to OwnedString.append()"_str));
     size_t oldLength = length();
     grow(sourceLen);
     for (size_t i = 0; i < sourceLen; i++)
       mBuf[oldLength + i] = source[i];
   }
+#endif // !ABP_UTF8_STRING
 
   void append(const String& str)
   {
     append(str.mBuf, str.length());
   }
 
   void append(value_type c)
   {
     append(&c, 1);
   }
@@ skipping 11 matching lines @@
 
     size_type size = 0;
     for (T i = num; i; i /= 10)
       size++;
     size = (size ? size : 1);
 
     size_type pos = length();
     grow((negative ? 1 : 0) + size);
 
     if (negative)
-      mBuf[pos++] = '-';
+      mBuf[pos++] = ABP_TEXT('-');
 
     for (int i = size - 1; i >= 0; i--)
     {
-      mBuf[pos + i] = '0' + (num % 10);
+      mBuf[pos + i] = ABP_TEXT('0') + (num % 10);
       num /= 10;
     }
   }
 };
 
 #ifdef INSIDE_TESTS
 inline std::ostream& operator<<(std::ostream& os, const OwnedString& str)
 {
   return os << static_cast<const String&>(str);
 }
 #endif
 
 template<typename T>
 struct LexicalCastImpl;
 
 /// Performs common conversions of a text represented value.
 template<typename T>
 inline T lexical_cast(const String& value)
 {
   return LexicalCastImpl<T>::Convert(value);
 }
 
 template<>
 struct LexicalCastImpl<bool>
 {
   static bool Convert(const String& value)
   {
-    return value == u"true"_str;
+    return value == ABP_TEXT("true"_str);
   }
 };
 
 template<typename T>
 struct LexicalCastImpl
 {
   static_assert(std::is_integral<T>::value, "T should be a number");
   static T Convert(const String& value)
   {
     String::size_type len = value.length();
     if (len == 0)
       return 0;
     String::size_type pos = 0;
-    bool negative = std::numeric_limits<T>::is_signed && value[0] == u'-';
+    bool negative = std::numeric_limits<T>::is_signed && value[0] == ABP_TEXT('-');
     if (negative)
     {
       ++pos;
     }
     T result = 0;
     for (; pos < len; ++pos)
     {
       auto c = value[pos];
-      if (c < u'0' || c > u'9')
+      if (c < ABP_TEXT('0') || c > ABP_TEXT('9'))
         return 0;
       // isDangerous is the optimization because there is no need for some checks
       // when the values are far from edge cases.
       // It targets the normal values, when a value is prefixed with several
       // zeros additional checks start to work earlier than the actual value of
       // result reaches an edge case, but it does not affect the result.
       bool isDangerous = pos >= std::numeric_limits<T>::digits10;
       // It also invalidates the parsing of too big numbers in comparison with
       // stopping when it encounters a non numerical character.
       // cast<uint8_t>(u"1230"_str) -> 0
       // cast<uint8_t>(u"123E"_str) -> 123
       if (isDangerous && std::numeric_limits<T>::max() / 10 < result)
       {
         return 0;
       }
       result *= 10;
-      uint8_t digit = c - u'0';
+      uint8_t digit = c - ABP_TEXT('0');
       if (isDangerous && (std::numeric_limits<T>::max() - digit < result - (negative ? 1 : 0)))
       {
         return 0;
       }
       result += digit;
     }
     return negative ? -result : result;
   }
 };
 
 template<>
 inline OwnedString lexical_cast<OwnedString>(const String& value)
 {
   return OwnedString{value};
 }
 
 DependentString TrimSpaces(const String& value);
 
 // Splits the `value` string into two `DependentString`s excluding the character staying at `separatorPos`.
 // Useful for parsing.
 std::pair<DependentString, DependentString> SplitString(const String& value, String::size_type separatorPos);
 
 ABP_NS_END