int red = color.r;
int green = color.g;
int blue = color.b;
+#ifdef USE_CORRECT_RGB_CONVERSION
+ int const scale = 255;
+#else
// the color values are given in the range of 0-255, so to get
// an output of "0.5" for the value 127 we need to do the following
+ // FIXME: This is wrong, since it creates a nonlinear mapping:
+ // There is a gap between 0/256 and 2/256!
+ // 0.5 cannot be represented in 8bit hex RGB, it would be 127.5.
if (red != 0)
++red;
if (green != 0)
++green;
if (blue != 0)
++blue;
+ int const scale = 256;
+#endif
string output;
- output = convert<string>(float(red) / 256) + ", "
- + convert<string>(float(green) / 256) + ", "
- + convert<string>(float(blue) / 256);
+ output = convert<string>(float(red) / scale) + ", "
+ + convert<string>(float(green) / scale) + ", "
+ + convert<string>(float(blue) / scale);
return output;
}
+RGBColor const RGBColorFromLaTeX(string const & color)
+{
+ vector<string> rgb = getVectorFromString(color);
+ while (rgb.size() < 3)
+ rgb.push_back("0");
+ RGBColor c;
+ for (int i = 0; i < 3; ++i) {
+ rgb[i] = trim(rgb[i]);
+ if (!isStrDbl(rgb[i]))
+ return c;
+ }
+#ifdef USE_CORRECT_RGB_CONVERSION
+ int const scale = 255;
+#else
+ // FIXME: This is wrong, since it creates a nonlinear mapping:
+ // Both 0/256 and 1/256 are mapped to 0!
+ // The wrong code exists only to match outputLaTeXColor().
+ int const scale = 256;
+#endif
+ c.r = static_cast<unsigned int>(scale * convert<double>(rgb[0]) + 0.5);
+ c.g = static_cast<unsigned int>(scale * convert<double>(rgb[1]) + 0.5);
+ c.b = static_cast<unsigned int>(scale * convert<double>(rgb[2]) + 0.5);
+#ifndef USE_CORRECT_RGB_CONVERSION
+ if (c.r != 0)
+ c.r--;
+ if (c.g != 0)
+ c.g--;
+ if (c.b != 0)
+ c.b--;
+#endif
+ return c;
+}
+
+
Color::Color(ColorCode base_color) : baseColor(base_color),
mergeColor(Color_ignore)
{}