Scanner now uses @nogc UTF decoding.
This commit is contained in:
parent
e1209711af
commit
e6fdade4a6
|
@ -20,7 +20,6 @@ import std.exception;
|
||||||
import std.string;
|
import std.string;
|
||||||
import std.typecons;
|
import std.typecons;
|
||||||
import std.traits : Unqual;
|
import std.traits : Unqual;
|
||||||
import std.utf;
|
|
||||||
|
|
||||||
import dyaml.fastcharsearch;
|
import dyaml.fastcharsearch;
|
||||||
import dyaml.escapes;
|
import dyaml.escapes;
|
||||||
|
@ -1942,11 +1941,24 @@ final class Scanner
|
||||||
//
|
//
|
||||||
// This is probably slow, but simple and URI escapes are extremely uncommon
|
// This is probably slow, but simple and URI escapes are extremely uncommon
|
||||||
// in YAML.
|
// in YAML.
|
||||||
|
//
|
||||||
|
// Returns the number of bytes used by the dchar in bytes on success,
|
||||||
|
// size_t.max on failure.
|
||||||
static size_t getDchar(char[] bytes, Reader reader_)
|
static size_t getDchar(char[] bytes, Reader reader_)
|
||||||
{
|
{
|
||||||
import std.utf;
|
|
||||||
size_t nextChar;
|
size_t nextChar;
|
||||||
const c = std.utf.decode(bytes[], nextChar);
|
dchar c;
|
||||||
|
if(bytes[0] < 0x80)
|
||||||
|
{
|
||||||
|
c = bytes[0];
|
||||||
|
++nextChar;
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
const decoded = decodeUTF8NoGC!(No.validated)(bytes[], nextChar);
|
||||||
|
if(decoded.errorMessage !is null) { return size_t.max; }
|
||||||
|
c = decoded.decoded;
|
||||||
|
}
|
||||||
reader_.sliceBuilder.write(c);
|
reader_.sliceBuilder.write(c);
|
||||||
if(bytes.length - nextChar > 0)
|
if(bytes.length - nextChar > 0)
|
||||||
{
|
{
|
||||||
|
@ -1957,14 +1969,19 @@ final class Scanner
|
||||||
}
|
}
|
||||||
|
|
||||||
enum contextMsg = "While scanning a " ~ name;
|
enum contextMsg = "While scanning a " ~ name;
|
||||||
try
|
|
||||||
{
|
|
||||||
while(reader_.peek() == '%')
|
while(reader_.peek() == '%')
|
||||||
{
|
{
|
||||||
reader_.forward();
|
reader_.forward();
|
||||||
if(bytesUsed == bytes.length)
|
if(bytesUsed == bytes.length)
|
||||||
{
|
{
|
||||||
bytesUsed = getDchar(bytes[], reader_);
|
bytesUsed = getDchar(bytes[], reader_);
|
||||||
|
if(bytesUsed == size_t.max)
|
||||||
|
{
|
||||||
|
error(contextMsg, startMark,
|
||||||
|
"Invalid UTF-8 data encoded in URI escape sequence",
|
||||||
|
reader_.mark);
|
||||||
|
return;
|
||||||
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
char b = 0;
|
char b = 0;
|
||||||
|
@ -1996,16 +2013,6 @@ final class Scanner
|
||||||
|
|
||||||
bytesUsed = getDchar(bytes[0 .. bytesUsed], reader_);
|
bytesUsed = getDchar(bytes[0 .. bytesUsed], reader_);
|
||||||
}
|
}
|
||||||
catch(UTFException e)
|
|
||||||
{
|
|
||||||
error(contextMsg, startMark, e.msg, mark);
|
|
||||||
return;
|
|
||||||
}
|
|
||||||
catch(Exception e)
|
|
||||||
{
|
|
||||||
assert(false, "Unexpected exception in scanURIEscapesToSlice");
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
|
|
||||||
/// Scan a line break, if any.
|
/// Scan a line break, if any.
|
||||||
|
|
Loading…
Reference in a new issue