/* ScummVM - Graphic Adventure Engine
 *
 * ScummVM is the legal property of its developers, whose names
 * are too numerous to list here. Please refer to the COPYRIGHT
 * file distributed with this source distribution.
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 *
 */

#include "ags/shared/util/compress.h"
#include "common/std/vector.h"
#include "ags/shared/ac/common.h"   // quit, update_polled_stuff
#include "ags/shared/gfx/bitmap.h"
#include "ags/shared/util/file.h"
#include "ags/shared/util/lzw.h"
#include "ags/shared/util/memory_stream.h"
#include "ags/globals.h"
#if AGS_PLATFORM_ENDIAN_BIG
#include "ags/shared/util/bbop.h"
#endif

#include "common/compression/deflate.h"

namespace AGS3 {

using namespace AGS::Shared;

//-----------------------------------------------------------------------------
// RLE
//-----------------------------------------------------------------------------

static void cpackbitl(const uint8_t *line, size_t size, Stream *out) {
	size_t cnt = 0;               // bytes encoded

	while (cnt < size) {
		// IMPORTANT: the algorithm below requires signed operations
		int i = static_cast<int32_t>(cnt);
		int j = i + 1;
		int jmax = i + 126;
		if (static_cast<uint32_t>(jmax) >= size)
			jmax = size - 1;

		if (static_cast<uint32_t>(i) == size - 1) { //......last byte alone
			out->WriteInt8(0);
			out->WriteInt8(line[i]);
			cnt++;

		} else if (line[i] == line[j]) {    //....run
			while ((j < jmax) && (line[j] == line[j + 1]))
				j++;

			out->WriteInt8(i - j);
			out->WriteInt8(line[i]);
			cnt += j - i + 1;

		} else {                    //.............................sequence
			while ((j < jmax) && (line[j] != line[j + 1]))
				j++;

			out->WriteInt8(j - i);
			out->Write(line + i, j - i + 1);
			cnt += j - i + 1;

		}
	} // end while
}

static void cpackbitl16(const uint16_t *line, size_t size, Stream *out) {
	size_t cnt = 0;               // bytes encoded

	while (cnt < size) {
		// IMPORTANT: the algorithm below requires signed operations
		int i = cnt;
		int j = i + 1;
		int jmax = i + 126;
		if (static_cast<uint32_t>(jmax) >= size)
			jmax = size - 1;

		if (static_cast<uint32_t>(i) == size - 1) { //......last byte alone
			out->WriteInt8(0);
			out->WriteInt16(line[i]);
			cnt++;

		} else if (line[i] == line[j]) {    //....run
			while ((j < jmax) && (line[j] == line[j + 1]))
				j++;

			out->WriteInt8(i - j);
			out->WriteInt16(line[i]);
			cnt += j - i + 1;

		} else {                    //.............................sequence
			while ((j < jmax) && (line[j] != line[j + 1]))
				j++;

			out->WriteInt8(j - i);
			out->WriteArray(line + i, j - i + 1, 2);
			cnt += j - i + 1;

		}
	} // end while
}

static void cpackbitl32(const uint32_t *line, size_t size, Stream *out) {
	size_t cnt = 0;               // bytes encoded

	while (cnt < size) {
		// IMPORTANT: the algorithm below requires signed operations
		int i = cnt;
		int j = i + 1;
		int jmax = i + 126;
		if (static_cast<uint32_t>(jmax) >= size)
			jmax = size - 1;

		if (static_cast<uint32_t>(i) == size - 1) { //......last byte alone
			out->WriteInt8(0);
			out->WriteInt32(line[i]);
			cnt++;

		} else if (line[i] == line[j]) {    //....run
			while ((j < jmax) && (line[j] == line[j + 1]))
				j++;

			out->WriteInt8(i - j);
			out->WriteInt32(line[i]);
			cnt += j - i + 1;

		} else {                    //.............................sequence
			while ((j < jmax) && (line[j] != line[j + 1]))
				j++;

			out->WriteInt8(j - i);
			out->WriteArray(line + i, j - i + 1, 4);
			cnt += j - i + 1;

		}
	} // end while
}

static int cunpackbitl(uint8_t *line, size_t size, Stream *in) {
	size_t n = 0;                  // number of bytes decoded

	while (n < size) {
		int ix = in->ReadByte();     // get index byte

		int8 cx = ix;
		if (cx == -128)
			cx = 0;

		if (cx < 0) {                //.............run
			int i = 1 - cx;
			char ch = in->ReadInt8();
			while (i--) {
				// test for buffer overflow
				if (n >= size)
					return -1;

				line[n++] = ch;
			}
		} else {                     //.....................seq
			int i = cx + 1;
			while (i--) {
				// test for buffer overflow
				if (n >= size)
					return -1;

				line[n++] = in->ReadByte();
			}
		}
	}

	return 0;
}

static int cunpackbitl16(uint16_t *line, size_t size, Stream *in) {
	size_t n = 0;                  // number of bytes decoded

	while (n < size) {
		int ix = in->ReadByte();     // get index byte

		int8 cx = ix;
		if (cx == -128)
			cx = 0;

		if (cx < 0) {                //.............run
			int i = 1 - cx;
			unsigned short ch = in->ReadInt16();
			while (i--) {
				// test for buffer overflow
				if (n >= size)
					return -1;

				line[n++] = ch;
			}
		} else {                     //.....................seq
			int i = cx + 1;
			while (i--) {
				// test for buffer overflow
				if (n >= size)
					return -1;

				line[n++] = in->ReadInt16();
			}
		}
	}

	return 0;
}

static int cunpackbitl32(uint32_t *line, size_t size, Stream *in) {
	size_t n = 0;                  // number of bytes decoded

	while (n < size) {
		int ix = in->ReadByte();     // get index byte

		int8 cx = ix;
		if (cx == -128)
			cx = 0;

		if (cx < 0) {                //.............run
			int i = 1 - cx;
			unsigned int ch = in->ReadInt32();
			while (i--) {
				// test for buffer overflow
				if (n >= size)
					return -1;

				line[n++] = ch;
			}
		} else {                     //.....................seq
			int i = cx + 1;
			while (i--) {
				// test for buffer overflow
				if (n >= size)
					return -1;

				line[n++] = (unsigned int)in->ReadInt32();
			}
		}
	}

	return 0;
}

bool rle_compress(const uint8_t *data, size_t data_sz, int image_bpp, Stream *out) {
	switch (image_bpp) {
	case 1: cpackbitl(data, data_sz, out); break;
	case 2: cpackbitl16(reinterpret_cast<const uint16_t *>(data), data_sz / sizeof(uint16_t), out); break;
	case 4: cpackbitl32(reinterpret_cast<const uint32_t *>(data), data_sz / sizeof(uint32_t), out); break;
	default: assert(0); break;
	}
	return true;
}

bool rle_decompress(uint8_t *data, size_t data_sz, int image_bpp, Stream *in) {
	switch (image_bpp) {
	case 1: cunpackbitl(data, data_sz, in); break;
	case 2: cunpackbitl16(reinterpret_cast<uint16_t *>(data), data_sz / sizeof(uint16_t), in); break;
	case 4: cunpackbitl32(reinterpret_cast<uint32_t *>(data), data_sz / sizeof(uint32_t), in); break;
	default: assert(0); break;
	}
	return true;
}

void save_rle_bitmap8(Stream *out, const Bitmap *bmp, const RGB(*pal)[256]) {
	assert(bmp->GetBPP() == 1);
	out->WriteInt16(static_cast<uint16_t>(bmp->GetWidth()));
	out->WriteInt16(static_cast<uint16_t>(bmp->GetHeight()));
	// Pack the pixels
	cpackbitl(bmp->GetData(), bmp->GetWidth() * bmp->GetHeight(), out);
	// Save palette
	if (!pal) { // if no pal, write dummy palette, because we have to
		out->WriteByteCount(0, 256 * 3);
		return;
	}
	const RGB *ppal = *pal;
	for (int i = 0; i < 256; ++i) {
		out->WriteInt8(ppal[i].r);
		out->WriteInt8(ppal[i].g);
		out->WriteInt8(ppal[i].b);
	}
}

Shared::Bitmap *load_rle_bitmap8(Stream *in, RGB(*pal)[256]) {
	int w = in->ReadInt16();
	int h = in->ReadInt16();
	Bitmap *bmp = BitmapHelper::CreateBitmap(w, h, 8);
	if (!bmp) return nullptr;
	// Unpack the pixels
	cunpackbitl(bmp->GetDataForWriting(), w * h, in);
	// Load or skip the palette
	if (!pal) {
		in->Seek(3 * 256);
		return bmp;
	}
	RGB *ppal = *pal;
	for (int i = 0; i < 256; ++i) {
		ppal[i].r = in->ReadInt8();
		ppal[i].g = in->ReadInt8();
		ppal[i].b = in->ReadInt8();
	}
	return bmp;
}

void skip_rle_bitmap8(Stream *in) {
	int w = in->ReadInt16();
	int h = in->ReadInt16();
	// Unpack the pixels into temp buf
	std::vector<uint8_t> buf;
	buf.resize(w * h);
	cunpackbitl(&buf[0], w * h, in);
	// Skip RGB palette
	in->Seek(3 * 256);
}

//-----------------------------------------------------------------------------
// LZW
//-----------------------------------------------------------------------------

bool lzw_compress(const uint8_t *data, size_t data_sz, int /*image_bpp*/, Shared::Stream *out) {
	// LZW algorithm that we use fails on sequence less than 16 bytes.
	if (data_sz < 16) {
		out->Write(data, data_sz);
		return true;
	}
	MemoryStream mem_in(data, data_sz);
	return lzwcompress(&mem_in, out);
}

bool lzw_decompress(uint8_t *data, size_t data_sz, int /*image_bpp*/, Shared::Stream *in, size_t in_sz) {
	// LZW algorithm that we use fails on sequence less than 16 bytes.
	if (data_sz < 16) {
		in->Read(data, data_sz);
		return true;
	}
	std::vector<uint8_t> in_buf(in_sz);
	in->Read(in_buf.data(), in_sz);
	return lzwexpand(in_buf.data(), in_sz, data, data_sz);
}

void save_lzw(Stream *out, const Bitmap *bmpp, const RGB(*pal)[256]) {
	// First write original bitmap's info and data into the memory buffer
	// NOTE: we must do this purely for backward compatibility with old room formats:
	// because they also included bmp width and height into compressed data!
	std::vector<uint8_t> membuf;
	{
		VectorStream memws(membuf, kStream_Write);
		int w = bmpp->GetWidth(), h = bmpp->GetHeight(), bpp = bmpp->GetBPP();
		memws.WriteInt32(w * bpp); // stride
		memws.WriteInt32(h);
		switch (bpp) {
		case 1: memws.Write(bmpp->GetData(), w * h * bpp); break;
		case 2: memws.WriteArrayOfInt16(reinterpret_cast<const int16_t *>(bmpp->GetData()), w *h); break;
		case 4: memws.WriteArrayOfInt32(reinterpret_cast<const int32_t *>(bmpp->GetData()), w *h); break;
		default: assert(0); break;
		}
	}

	// Open same buffer for reading, and begin writing compressed data into the output
	VectorStream mem_in(membuf);
	// NOTE: old format saves full RGB struct here (4 bytes, including the filler)
	if (pal)
		out->WriteArray(*pal, sizeof(RGB), 256);
	else
		out->WriteByteCount(0, sizeof(RGB) * 256);
	out->WriteInt32((uint32_t)mem_in.GetLength());

	// reserve space for compressed size
	soff_t cmpsz_at = out->GetPosition();
	out->WriteInt32(0);
	lzwcompress(&mem_in, out);
	soff_t toret = out->GetPosition();
	out->Seek(cmpsz_at, kSeekBegin);
	soff_t compressed_sz = (toret - cmpsz_at) - sizeof(uint32_t);
	out->WriteInt32(compressed_sz); // write compressed size
	// seek back to the end of the output stream
	out->Seek(toret, kSeekBegin);
}

Bitmap *load_lzw(Stream *in, int dst_bpp, RGB(*pal)[256]) {
	// NOTE: old format saves full RGB struct here (4 bytes, including the filler)
	if (pal)
		in->Read(*pal, sizeof(RGB) * 256);
	else
		in->Seek(sizeof(RGB) * 256);
	const size_t uncomp_sz = in->ReadInt32();
	const size_t comp_sz = in->ReadInt32();
	const soff_t end_pos = in->GetPosition() + comp_sz;

	// First decompress data into the memory buffer
	std::vector<uint8_t> inbuf(comp_sz);
	std::vector<uint8_t> membuf(uncomp_sz);
	in->Read(inbuf.data(), comp_sz);
	lzwexpand(inbuf.data(), comp_sz, membuf.data(), uncomp_sz);

	// Open same buffer for reading and get params and pixels
	VectorStream mem_in(membuf);
	int stride = mem_in.ReadInt32(); // width * bpp
	int height = mem_in.ReadInt32();
	Bitmap *bmm = BitmapHelper::CreateBitmap((stride / dst_bpp), height, dst_bpp * 8);
	if (!bmm) return nullptr; // out of mem?

	size_t num_pixels = stride * height / dst_bpp;
	uint8_t *bmp_data = bmm->GetDataForWriting();
	switch (dst_bpp) {
	case 1: mem_in.Read(bmp_data, num_pixels); break;
	case 2: mem_in.ReadArrayOfInt16(reinterpret_cast<int16_t *>(bmp_data), num_pixels); break;
	case 4: mem_in.ReadArrayOfInt32(reinterpret_cast<int32_t *>(bmp_data), num_pixels); break;
	default: assert(0); break;
	}

	if (in->GetPosition() != end_pos)
		in->Seek(end_pos, kSeekBegin);

	return bmm;
}

bool deflate_compress(const uint8_t *data, size_t data_sz, int /*image_bpp*/, Stream *out) {
	// TODO
	return false;
}

bool inflate_decompress(uint8_t *data, size_t data_sz, int /*image_bpp*/, Stream *in, size_t in_sz) {
	std::vector<uint8_t> in_buf(in_sz);
	in->Read(in_buf.data(), in_sz);
	return Common::inflateZlib(data, (unsigned long)data_sz, in_buf.data(), in_sz);
}

} // namespace AGS3