mirror of
https://github.com/curioustorvald/Terrarum.git
synced 2026-06-21 16:04:05 +09:00
bytearray64reader: read length of zero and EOF are properly distinguished (gdx is somewhat pedantic); changed an ascii85 charset; working meta (de)serialisation
This commit is contained in:
minjaesong
@@ -3,7 +3,6 @@ package net.torvald.terrarum.serialise
|
||||
import com.badlogic.gdx.utils.Json
|
||||
import com.badlogic.gdx.utils.JsonValue
|
||||
import com.badlogic.gdx.utils.JsonWriter
|
||||
import net.torvald.terrarum.AppLoader.printdbg
|
||||
import net.torvald.terrarum.console.EchoError
|
||||
import net.torvald.terrarum.gameworld.BlockLayer
|
||||
import net.torvald.terrarum.gameworld.GameWorld
|
||||
@@ -272,215 +271,3 @@ object Common {
|
||||
return unzipdBytes
|
||||
}
|
||||
}
|
||||
|
||||
class ByteArray64Writer(val charset: Charset) : Writer() {
|
||||
|
||||
private val acceptableCharsets = arrayOf(Charsets.UTF_8, Charset.forName("CP437"))
|
||||
|
||||
init {
|
||||
if (!acceptableCharsets.contains(charset))
|
||||
throw UnsupportedCharsetException(charset.name())
|
||||
}
|
||||
|
||||
private val ba64 = ByteArray64()
|
||||
private var closed = false
|
||||
private var surrogateBuf = 0
|
||||
|
||||
init {
|
||||
this.lock = ba64
|
||||
}
|
||||
|
||||
private fun checkOpen() {
|
||||
if (closed) throw ClosedChannelException()
|
||||
}
|
||||
|
||||
private fun Int.isSurroHigh() = this.ushr(10) == 0b110110
|
||||
private fun Int.isSurroLow() = this.ushr(10) == 0b110111
|
||||
private fun Int.toUcode() = 'u' + this.toString(16).uppercase().padStart(4,'0')
|
||||
|
||||
/**
|
||||
* @param c not a freakin' codepoint; just a Java's Char casted into Int
|
||||
*/
|
||||
override fun write(c: Int) {
|
||||
checkOpen()
|
||||
when (charset) {
|
||||
Charsets.UTF_8 -> {
|
||||
if (surrogateBuf == 0 && !c.isSurroHigh() && !c.isSurroLow())
|
||||
writeUtf8Codepoint(c)
|
||||
else if (surrogateBuf == 0 && c.isSurroHigh())
|
||||
surrogateBuf = c
|
||||
else if (surrogateBuf != 0 && c.isSurroLow())
|
||||
writeUtf8Codepoint(65536 + surrogateBuf.and(1023).shl(10) or c.and(1023))
|
||||
// invalid surrogate pair input
|
||||
else
|
||||
throw IllegalStateException("Surrogate high: ${surrogateBuf.toUcode()}, surrogate low: ${c.toUcode()}")
|
||||
}
|
||||
Charset.forName("CP437") -> {
|
||||
ba64.add(c.toByte())
|
||||
}
|
||||
else -> throw UnsupportedCharsetException(charset.name())
|
||||
}
|
||||
}
|
||||
|
||||
fun writeUtf8Codepoint(codepoint: Int) {
|
||||
when (codepoint) {
|
||||
in 0..127 -> ba64.add(codepoint.toByte())
|
||||
in 128..2047 -> {
|
||||
ba64.add((0xC0 or codepoint.ushr(6).and(31)).toByte())
|
||||
ba64.add((0x80 or codepoint.and(63)).toByte())
|
||||
}
|
||||
in 2048..65535 -> {
|
||||
ba64.add((0xE0 or codepoint.ushr(12).and(15)).toByte())
|
||||
ba64.add((0x80 or codepoint.ushr(6).and(63)).toByte())
|
||||
ba64.add((0x80 or codepoint.and(63)).toByte())
|
||||
}
|
||||
in 65536..1114111 -> {
|
||||
ba64.add((0xF0 or codepoint.ushr(18).and(7)).toByte())
|
||||
ba64.add((0x80 or codepoint.ushr(12).and(63)).toByte())
|
||||
ba64.add((0x80 or codepoint.ushr(6).and(63)).toByte())
|
||||
ba64.add((0x80 or codepoint.and(63)).toByte())
|
||||
}
|
||||
else -> throw IllegalArgumentException("Not a unicode code point: U+${codepoint.toString(16).uppercase()}")
|
||||
}
|
||||
}
|
||||
|
||||
override fun write(cbuf: CharArray) {
|
||||
checkOpen()
|
||||
write(String(cbuf))
|
||||
}
|
||||
|
||||
override fun write(str: String) {
|
||||
checkOpen()
|
||||
str.toByteArray(charset).forEach { ba64.add(it) }
|
||||
}
|
||||
|
||||
override fun write(cbuf: CharArray, off: Int, len: Int) {
|
||||
write(cbuf.copyOfRange(off, off + len))
|
||||
}
|
||||
|
||||
override fun write(str: String, off: Int, len: Int) {
|
||||
write(str.substring(off, off + len))
|
||||
}
|
||||
|
||||
override fun close() { closed = true }
|
||||
override fun flush() {}
|
||||
|
||||
fun toByteArray64() = if (closed) ba64 else throw IllegalAccessException("Writer not closed")
|
||||
}
|
||||
|
||||
class ByteArray64Reader(val ba: ByteArray64, val charset: Charset) : Reader() {
|
||||
|
||||
private val acceptableCharsets = arrayOf(Charsets.UTF_8, Charset.forName("CP437"))
|
||||
|
||||
init {
|
||||
if (!acceptableCharsets.contains(charset))
|
||||
throw UnsupportedCharsetException(charset.name())
|
||||
}
|
||||
|
||||
private var readCursor = 0L
|
||||
private val remaining
|
||||
get() = ba.size - readCursor
|
||||
|
||||
/**
|
||||
* U+0000 .. U+007F 0xxxxxxx
|
||||
* U+0080 .. U+07FF 110xxxxx 10xxxxxx
|
||||
* U+0800 .. U+FFFF 1110xxxx 10xxxxxx 10xxxxxx
|
||||
* U+10000 .. U+10FFFF 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
|
||||
*/
|
||||
private fun utf8GetCharLen(head: Byte) = when (head.toInt() and 255) {
|
||||
in 0b11110_000..0b11110_111 -> 4
|
||||
in 0b1110_0000..0b1110_1111 -> 3
|
||||
in 0b110_00000..0b110_11111 -> 2
|
||||
in 0b0_0000000..0b0_1111111 -> 1
|
||||
else -> throw IllegalArgumentException("Invalid UTF-8 Character head byte: ${head.toInt() and 255}")
|
||||
}
|
||||
|
||||
/**
|
||||
* @param list of bytes that encodes one unicode character. Get required byte length using [utf8GetCharLen].
|
||||
* @return A codepoint of the character.
|
||||
*/
|
||||
private fun utf8decode(bytes0: List<Byte>): Int {
|
||||
val bytes = bytes0.map { it.toInt() and 255 }
|
||||
var ret = when (bytes.size) {
|
||||
4 -> (bytes[0] and 7) shl 18
|
||||
3 -> (bytes[0] and 15) shl 12
|
||||
2 -> (bytes[0] and 31) shl 6
|
||||
1 -> (bytes[0] and 127)
|
||||
else -> throw IllegalArgumentException("Expected bytes size: 1..4, got ${bytes.size}")
|
||||
}
|
||||
bytes.tail().reversed().forEachIndexed { index, byte ->
|
||||
ret = ret or (byte and 63).shl(6 * index)
|
||||
}
|
||||
return ret
|
||||
}
|
||||
|
||||
private var surrogateLeftover = ' '
|
||||
|
||||
override fun read(cbuf: CharArray, off: Int, len: Int): Int {
|
||||
var readCount = 0
|
||||
|
||||
when (charset) {
|
||||
Charsets.UTF_8 -> {
|
||||
while (readCount < len && remaining > 0) {
|
||||
if (surrogateLeftover != ' ') {
|
||||
cbuf[off + readCount] = surrogateLeftover
|
||||
|
||||
readCount += 1
|
||||
surrogateLeftover = ' '
|
||||
}
|
||||
else {
|
||||
val bbuf = (0 until minOf(4L, remaining)).map { ba[readCursor + it] }
|
||||
val charLen = utf8GetCharLen(bbuf[0])
|
||||
val codePoint = utf8decode(bbuf.subList(0, charLen))
|
||||
|
||||
if (codePoint < 65536) {
|
||||
cbuf[off + readCount] = codePoint.toChar()
|
||||
|
||||
readCount += 1
|
||||
readCursor += charLen
|
||||
}
|
||||
else {
|
||||
/*
|
||||
* U' = yyyyyyyyyyxxxxxxxxxx // U - 0x10000
|
||||
* W1 = 110110yyyyyyyyyy // 0xD800 + yyyyyyyyyy
|
||||
* W2 = 110111xxxxxxxxxx // 0xDC00 + xxxxxxxxxx
|
||||
*/
|
||||
val codPoin = codePoint - 65536
|
||||
val surroLead = (0xD800 or codPoin.ushr(10)).toChar()
|
||||
val surroTrail = (0xDC00 or codPoin.and(1023)).toChar()
|
||||
|
||||
cbuf[off + readCount] = surroLead
|
||||
|
||||
if (off + readCount + 1 < cbuf.size) {
|
||||
cbuf[off + readCount + 1] = surroTrail
|
||||
|
||||
readCount += 2
|
||||
readCursor += 4
|
||||
}
|
||||
else {
|
||||
readCount += 1
|
||||
readCursor += 4
|
||||
surrogateLeftover = surroTrail
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
Charset.forName("CP437") -> {
|
||||
for (i in 0 until minOf(len.toLong(), remaining)) {
|
||||
cbuf[(off + i).toInt()] = ba[readCursor].toChar()
|
||||
readCursor += 1
|
||||
readCount += 1
|
||||
}
|
||||
}
|
||||
else -> throw UnsupportedCharsetException(charset.name())
|
||||
}
|
||||
|
||||
return if (readCount == 0) -1 else readCount
|
||||
}
|
||||
|
||||
override fun close() { readCursor = 0L }
|
||||
override fun reset() { readCursor = 0L }
|
||||
override fun markSupported() = false
|
||||
|
||||
}
|
||||
Reference in New Issue
Block a user