Add Decoder class to decode code points

This commit is contained in:
Josh Holtrop 2021-07-05 18:47:10 -04:00
parent 651461c570
commit 2f1cb47bea

View File

@ -4,6 +4,71 @@ module <%= @grammar.modulename %>;
<% end %>
class <%= classname %>
{
class Decoder
{
enum
{
CODE_POINT_INVALID = 0xFFFFFFFE,
CODE_POINT_EOF = 0xFFFFFFFF,
}
static uint decode_code_point(const(ubyte) * input, size_t input_length, size_t * code_point_length)
{
if (input_length == 0u)
{
return CODE_POINT_EOF;
}
ubyte c = *input;
uint result;
if ((c & 0x80u) == 0u)
{
result = c;
*code_point_length = 1u;
}
else
{
ubyte following_bytes;
if ((c & 0xE0u) == 0xC0u)
{
result = c & 0x1Fu;
following_bytes = 1u;
}
else if ((c & 0xF0u) == 0xE0u)
{
result = c & 0x0Fu;
following_bytes = 2u;
}
else if ((c & 0xF8u) == 0xF0u)
{
result = c & 0x07u;
following_bytes = 3u;
}
else if ((c & 0xFCu) == 0xF8u)
{
result = c & 0x03u;
following_bytes = 4u;
}
else if ((c & 0xFEu) == 0xFCu)
{
result = c & 0x01u;
following_bytes = 5u;
}
if (input_length <= following_bytes)
{
return CODE_POINT_INVALID;
}
*code_point_length = following_bytes + 1u;
while (following_bytes-- > 0u)
{
input++;
result <<= 6u;
result |= *input & 0x3Fu;
}
}
return result;
}
}
enum
{
<% @grammar.tokens.each_with_index do |token, index| %>
@ -53,7 +118,7 @@ class <%= classname %>
static string[] lex(const(ubyte)[] input)
{
string[] tokens;
string token = lex_token(&input);
string token = lex_token(input.ptr, input.length);
if (token !is null)
{
tokens ~= token;
@ -61,15 +126,10 @@ class <%= classname %>
return tokens;
}
private static string lex_token(const(ubyte)[] * input)
private static string lex_token(const(ubyte) * input, size_t input_length)
{
uint code_point_length;
uint code_point = decode_code_point(input, &code_point_length);
size_t code_point_length;
uint code_point = Decoder.decode_code_point(input, input_length, &code_point_length);
return null;
}
private static uint decode_code_point(const(ubyte)[] * input, uint * code_point_length)
{
return 0u;
}
}