/*
  Copyright 2009 Larry Gritz and the other authors and contributors.
  All Rights Reserved.

  Redistribution and use in source and binary forms, with or without
  modification, are permitted provided that the following conditions are
  met:
  * Redistributions of source code must retain the above copyright
    notice, this list of conditions and the following disclaimer.
  * Redistributions in binary form must reproduce the above copyright
    notice, this list of conditions and the following disclaimer in the
    documentation and/or other materials provided with the distribution.
  * Neither the name of the software's owners nor the names of its
    contributors may be used to endorse or promote products derived from
    this software without specific prior written permission.
  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
  "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
  LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
  A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
  OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
  SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
  LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
  DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
  THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

  (This is the Modified BSD License)
*/

#include <cmath>
#include <cstdio>
#include <cstdlib>
#include <memory>

#include <OpenImageIO/dassert.h>
#include <OpenImageIO/filesystem.h>
#include <OpenImageIO/fmath.h>
#include <OpenImageIO/imageio.h>
#include <OpenImageIO/typedesc.h>

#include "targa_pvt.h"

OIIO_PLUGIN_NAMESPACE_BEGIN

using namespace TGA_pvt;


class TGAInput final : public ImageInput {
public:
    TGAInput() { init(); }
    virtual ~TGAInput() { close(); }
    virtual const char* format_name(void) const override { return "targa"; }
    virtual bool open(const std::string& name, ImageSpec& newspec) override;
    virtual bool open(const std::string& name, ImageSpec& newspec,
                      const ImageSpec& config) override;
    virtual bool close() override;
    virtual bool read_native_scanline(int subimage, int miplevel, int y, int z,
                                      void* data) override;

private:
    std::string m_filename;            ///< Stash the filename
    FILE* m_file;                      ///< Open image handle
    tga_header m_tga;                  ///< Targa header
    tga_footer m_foot;                 ///< Targa 2.0 footer
    unsigned int m_ofs_colcorr_tbl;    ///< Offset to colour correction table
    tga_alpha_type m_alpha;            ///< Alpha type
    bool m_keep_unassociated_alpha;    ///< Do not convert unassociated alpha
    std::vector<unsigned char> m_buf;  ///< Buffer the image pixels

    /// Reset everything to initial state
    ///
    void init()
    {
        m_file = NULL;
        m_buf.clear();
        m_ofs_colcorr_tbl         = 0;
        m_alpha                   = TGA_ALPHA_NONE;
        m_keep_unassociated_alpha = false;
    }

    /// Helper function: read the image.
    ///
    bool readimg();

    /// Helper function: decode a pixel.
    inline void decode_pixel(unsigned char* in, unsigned char* out,
                             unsigned char* palette, int& bytespp,
                             int& palbytespp, int& alphabits);

    /// Helper: read, with error detection
    ///
    bool fread(void* buf, size_t itemsize, size_t nitems)
    {
        size_t n = ::fread(buf, itemsize, nitems, m_file);
        if (n != nitems)
            error("Read error");
        return n == nitems;
    }
};



// Obligatory material to make this a recognizeable imageio plugin:
OIIO_PLUGIN_EXPORTS_BEGIN

OIIO_EXPORT ImageInput*
targa_input_imageio_create()
{
    return new TGAInput;
}

OIIO_EXPORT int targa_imageio_version = OIIO_PLUGIN_VERSION;

OIIO_EXPORT const char* targa_input_extensions[] = { "tga", "tpic", nullptr };

OIIO_EXPORT const char*
targa_imageio_library_version()
{
    return nullptr;
}

OIIO_PLUGIN_EXPORTS_END



bool
TGAInput::open(const std::string& name, ImageSpec& newspec)
{
    m_filename = name;

    m_file = Filesystem::fopen(name, "rb");
    if (!m_file) {
        error("Could not open file \"%s\"", name.c_str());
        return false;
    }

    // due to struct packing, we may get a corrupt header if we just load the
    // struct from file; to adress that, read every member individually
    // save some typing
#define RH(memb)                                                               \
    if (!fread(&m_tga.memb, sizeof(m_tga.memb), 1))                            \
    return false

    RH(idlen);
    RH(cmap_type);
    RH(type);
    RH(cmap_first);
    RH(cmap_length);
    RH(cmap_size);
    RH(x_origin);
    RH(y_origin);
    RH(width);
    RH(height);
    RH(bpp);
    RH(attr);
#undef RH
    if (bigendian()) {
        // TGAs are little-endian
        swap_endian(&m_tga.idlen);
        swap_endian(&m_tga.cmap_type);
        swap_endian(&m_tga.type);
        swap_endian(&m_tga.cmap_first);
        swap_endian(&m_tga.cmap_length);
        swap_endian(&m_tga.cmap_size);
        swap_endian(&m_tga.x_origin);
        swap_endian(&m_tga.y_origin);
        swap_endian(&m_tga.width);
        swap_endian(&m_tga.height);
        swap_endian(&m_tga.bpp);
        swap_endian(&m_tga.attr);
    }

    if (m_tga.bpp != 8 && m_tga.bpp != 15 && m_tga.bpp != 16 && m_tga.bpp != 24
        && m_tga.bpp != 32) {
        error("Illegal pixel size: %d bits per pixel", m_tga.bpp);
        return false;
    }

    if (m_tga.type == TYPE_NODATA) {
        error("Image with no data");
        return false;
    }
    if (m_tga.type != TYPE_PALETTED && m_tga.type != TYPE_RGB
        && m_tga.type != TYPE_GRAY && m_tga.type != TYPE_PALETTED_RLE
        && m_tga.type != TYPE_RGB_RLE && m_tga.type != TYPE_GRAY_RLE) {
        error("Illegal image type: %d", m_tga.type);
        return false;
    }

    if (m_tga.cmap_type
        && (m_tga.type == TYPE_GRAY || m_tga.type == TYPE_GRAY_RLE)) {
        // it should be an error for TYPE_RGB* as well, but apparently some
        // *very* old TGAs can be this way, so we'll hack around it
        error("Palette defined for grayscale image");
        return false;
    }

    if (m_tga.cmap_type
        && (m_tga.cmap_size != 15 && m_tga.cmap_size != 16
            && m_tga.cmap_size != 24 && m_tga.cmap_size != 32)) {
        error("Illegal palette entry size: %d bits", m_tga.cmap_size);
        return false;
    }

    m_alpha = TGA_ALPHA_NONE;
    if (((m_tga.type == TYPE_RGB || m_tga.type == TYPE_RGB_RLE)
         && m_tga.bpp == 32)
        || ((m_tga.type == TYPE_GRAY || m_tga.type == TYPE_GRAY_RLE)
            && m_tga.bpp > 8)) {
        m_alpha = (m_tga.attr & 0x08) > 0 ? TGA_ALPHA_USEFUL : TGA_ALPHA_NONE;
    }


    m_spec = ImageSpec(
        (int)m_tga.width, (int)m_tga.height,
        // colour channels
        ((m_tga.type == TYPE_GRAY || m_tga.type == TYPE_GRAY_RLE) ? 1 : 3)
            // have we got alpha?
            + (m_tga.bpp == 32 || m_alpha >= TGA_ALPHA_UNDEFINED_RETAIN),
        TypeDesc::UINT8);
    m_spec.attribute("oiio:BitsPerSample", m_tga.bpp / m_spec.nchannels);
    m_spec.default_channel_names();
#if 0  // no one seems to adhere to this part of the spec...
    if (m_tga.attr & FLAG_X_FLIP)
        m_spec.x = m_spec.width - m_tga.x_origin - 1;
    else
        m_spec.x = m_tga.x_origin;
    if (m_tga.attr & FLAG_Y_FLIP)
        m_spec.y = m_tga.y_origin;
    else
        m_spec.y = m_spec.width - m_tga.y_origin - 1;
#endif
    if (m_tga.type >= TYPE_PALETTED_RLE)
        m_spec.attribute("compression", "rle");

    /*std::cerr << "[tga] " << m_tga.width << "x" << m_tga.height << "@"
              << (int)m_tga.bpp << " (" << m_spec.nchannels
              << ") type " << (int)m_tga.type << "\n";*/

    // load image ID
    if (m_tga.idlen) {
        // TGA comments can be at most 255 bytes long, but we add 1 extra byte
        // in case the comment lacks null termination
        char id[256];
        memset(id, 0, sizeof(id));
        if (!fread(id, m_tga.idlen, 1))
            return false;
        m_spec.attribute("targa:ImageID", id);
    }

    int ofs = ftell(m_file);
    // now try and see if it's a TGA 2.0 image
    // TGA 2.0 files are identified by a nifty "TRUEVISION-XFILE.\0" signature
    fseek(m_file, -26, SEEK_END);
    if (fread(&m_foot.ofs_ext, sizeof(m_foot.ofs_ext), 1)
        && fread(&m_foot.ofs_dev, sizeof(m_foot.ofs_dev), 1)
        && fread(&m_foot.signature, sizeof(m_foot.signature), 1)
        && !strncmp(m_foot.signature, "TRUEVISION-XFILE.", 17)) {
        //std::cerr << "[tga] this is a TGA 2.0 file\n";
        if (bigendian()) {
            swap_endian(&m_foot.ofs_ext);
            swap_endian(&m_foot.ofs_dev);
        }

        // read the extension area
        fseek(m_file, m_foot.ofs_ext, SEEK_SET);
        // check if this is a TGA 2.0 extension area
        // according to the 2.0 spec, the size for valid 2.0 files is exactly
        // 495 bytes, and the reader should only read as much as it understands
        // for < 495, we ignore this section of the file altogether
        // for > 495, we only read what we know
        uint16_t s;
        if (!fread(&s, 2, 1))
            return false;
        if (bigendian())
            swap_endian(&s);
        //std::cerr << "[tga] extension area size: " << s << "\n";
        if (s >= 495) {
            union {
                unsigned char c[324];  // so as to accomodate the comments
                uint16_t s[6];
                uint32_t l;
            } buf;

            // load image author
            if (!fread(buf.c, 41, 1))
                return false;
            if (buf.c[0])
                m_spec.attribute("Artist", (char*)buf.c);

            // load image comments
            if (!fread(buf.c, 324, 1))
                return false;

            // concatenate the lines into a single string
            std::string tmpstr((const char*)buf.c);
            if (buf.c[81]) {
                tmpstr += "\n";
                tmpstr += (const char*)&buf.c[81];
            }
            if (buf.c[162]) {
                tmpstr += "\n";
                tmpstr += (const char*)&buf.c[162];
            }
            if (buf.c[243]) {
                tmpstr += "\n";
                tmpstr += (const char*)&buf.c[243];
            }
            if (tmpstr.length() > 0)
                m_spec.attribute("ImageDescription", tmpstr);

            // timestamp
            if (!fread(buf.s, 2, 6))
                return false;
            if (buf.s[0] || buf.s[1] || buf.s[2] || buf.s[3] || buf.s[4]
                || buf.s[5]) {
                if (bigendian())
                    swap_endian(&buf.s[0], 6);
                sprintf((char*)&buf.c[12], "%04u:%02u:%02u %02u:%02u:%02u",
                        buf.s[2], buf.s[0], buf.s[1], buf.s[3], buf.s[4],
                        buf.s[5]);
                m_spec.attribute("DateTime", (char*)&buf.c[12]);
            }

            // job name/ID
            if (!fread(buf.c, 41, 1))
                return false;
            if (buf.c[0])
                m_spec.attribute("DocumentName", (char*)buf.c);

            // job time
            if (!fread(buf.s, 2, 3))
                return false;
            if (buf.s[0] || buf.s[1] || buf.s[2]) {
                if (bigendian())
                    swap_endian(&buf.s[0], 3);
                sprintf((char*)&buf.c[6], "%u:%02u:%02u", buf.s[0], buf.s[1],
                        buf.s[2]);
                m_spec.attribute("targa:JobTime", (char*)&buf.c[6]);
            }

            // software
            if (!fread(buf.c, 41, 1))
                return false;
            if (buf.c[0]) {
                // tack on the version number and letter
                uint16_t n;
                char l;
                if (!fread(&n, 2, 1) || !fread(&l, 1, 1))
                    return false;
                sprintf((char*)&buf.c[strlen((char*)buf.c)], " %u.%u%c",
                        n / 100, n % 100, l != ' ' ? l : 0);
                m_spec.attribute("Software", (char*)buf.c);
            }

            // background (key) colour
            if (!fread(buf.c, 4, 1))
                return false;
            // FIXME: what do we do with it?

            // aspect ratio
            if (!fread(buf.s, 2, 2))
                return false;
            // if the denominator is zero, it's unused
            if (buf.s[1]) {
                if (bigendian())
                    swap_endian(&buf.s[0], 2);
                m_spec.attribute("PixelAspectRatio",
                                 (float)buf.s[0] / (float)buf.s[1]);
            }

            // gamma
            if (!fread(buf.s, 2, 2))
                return false;
            // if the denominator is zero, it's unused
            if (buf.s[1]) {
                if (bigendian())
                    swap_endian(&buf.s[0], 2);
                float gamma = (float)buf.s[0] / (float)buf.s[1];
                // Round gamma to the nearest hundredth to prevent stupid
                // precision choices and make it easier for apps to make
                // decisions based on known gamma values. For example, you want
                // 2.2, not 2.19998.
                gamma = roundf(100.0 * gamma) / 100.0f;
                if (gamma == 1.f) {
                    m_spec.attribute("oiio:ColorSpace", "linear");
                } else {
                    m_spec.attribute("oiio:ColorSpace",
                                     Strutil::sprintf("GammaCorrected%.2g",
                                                      gamma));
                    m_spec.attribute("oiio:Gamma", gamma);
                }
            }

            // offset to colour correction table
            if (!fread(&buf.l, 4, 1))
                return false;
            if (bigendian())
                swap_endian(&buf.l);
            m_ofs_colcorr_tbl = buf.l;
            /*std::cerr << "[tga] colour correction table offset: "
                      << (int)m_ofs_colcorr_tbl << "\n";*/

            // offset to thumbnail
            if (!fread(&buf.l, 4, 1))
                return false;
            if (bigendian())
                swap_endian(&buf.l);
            unsigned int ofs_thumb = buf.l;

            // offset to scan-line table
            if (!fread(&buf.l, 4, 1))
                return false;
            // TODO: can we find any use for this? we can't advertise random
            // access anyway, because not all RLE-compressed files will have
            // this table

            // alpha type
            if (!fread(buf.c, 1, 1))
                return false;
            m_alpha = (tga_alpha_type)buf.c[0];

            // now load the thumbnail
            if (ofs_thumb) {
                fseek(m_file, ofs_thumb, SEEK_SET);

                // most of this code is a dupe of readimg(); according to the
                // spec, the thumbnail is in the same format as the main image
                // but uncompressed

                // thumbnail dimensions
                if (!fread(&buf.c, 2, 1))
                    return false;
                m_spec.attribute("thumbnail_width", (int)buf.c[0]);
                m_spec.attribute("thumbnail_height", (int)buf.c[1]);
                m_spec.attribute("thumbnail_nchannels", m_spec.nchannels);

                // load image data
                // reuse the image buffer
                m_buf.resize(buf.c[0] * buf.c[1] * m_spec.nchannels);
                int bytespp    = (m_tga.bpp == 15) ? 2 : (m_tga.bpp / 8);
                int palbytespp = (m_tga.cmap_size == 15)
                                     ? 2
                                     : (m_tga.cmap_size / 8);
                int alphabits = m_tga.attr & 0x0F;
                if (alphabits == 0 && m_tga.bpp == 32)
                    alphabits = 8;
                // read palette, if there is any
                std::unique_ptr<unsigned char[]> palette;
                if (m_tga.cmap_type) {
                    fseek(m_file, ofs, SEEK_SET);
                    palette.reset(
                        new unsigned char[palbytespp * m_tga.cmap_length]);
                    if (!fread(palette.get(), palbytespp, m_tga.cmap_length))
                        return false;
                    fseek(m_file, ofs_thumb + 2, SEEK_SET);
                }
                unsigned char pixel[4];
                unsigned char in[4];
                for (int y = buf.c[1] - 1; y >= 0; y--) {
                    for (int x = 0; x < buf.c[0]; x++) {
                        if (!fread(in, bytespp, 1))
                            return false;
                        decode_pixel(in, pixel, palette.get(), bytespp,
                                     palbytespp, alphabits);
                        memcpy(&m_buf[y * buf.c[0] * m_spec.nchannels
                                      + x * m_spec.nchannels],
                               pixel, m_spec.nchannels);
                    }
                }
                //std::cerr << "[tga] buffer size: " << m_buf.size() << "\n";
                // finally, add the thumbnail to attributes
                m_spec.attribute("thumbnail_image",
                                 TypeDesc(TypeDesc::UINT8, m_buf.size()),
                                 &m_buf[0]);
                m_buf.clear();
            }
        }

        // FIXME: provide access to the developer area; according to Larry,
        // it's probably safe to ignore it altogether until someone complains
        // that it's missing :)
    }

    if (m_spec.alpha_channel != -1 && m_alpha != TGA_ALPHA_PREMULTIPLIED)
        if (m_keep_unassociated_alpha)
            m_spec.attribute("oiio:UnassociatedAlpha", 1);

    fseek(m_file, ofs, SEEK_SET);

    newspec = spec();
    return true;
}



bool
TGAInput::open(const std::string& name, ImageSpec& newspec,
               const ImageSpec& config)
{
    // Check 'config' for any special requests
    if (config.get_int_attribute("oiio:UnassociatedAlpha", 0) == 1)
        m_keep_unassociated_alpha = true;
    return open(name, newspec);
}



inline void
TGAInput::decode_pixel(unsigned char* in, unsigned char* out,
                       unsigned char* palette, int& bytespp, int& palbytespp,
                       int& alphabits)
{
    unsigned int k = 0;
    // I hate nested switches...
    switch (m_tga.type) {
    case TYPE_PALETTED:
    case TYPE_PALETTED_RLE:
        for (int i = 0; i < bytespp; ++i)
            k |= in[i] << (8 * i);  // Assemble it in little endian order
        k = (m_tga.cmap_first + k) * palbytespp;
        switch (palbytespp) {
        case 2:
            // see the comment for 16bpp RGB below for an explanation of this
            out[0] = bit_range_convert<5, 8>((palette[k + 1] & 0x7C) >> 2);
            out[1] = bit_range_convert<5, 8>(((palette[k + 0] & 0xE0) >> 5)
                                             | ((palette[k + 1] & 0x03) << 3));
            out[2] = bit_range_convert<5, 8>(palette[k + 0] & 0x1F);
            break;
        case 3:
            out[0] = palette[k + 2];
            out[1] = palette[k + 1];
            out[2] = palette[k + 0];
            break;
        case 4:
            out[0] = palette[k + 2];
            out[1] = palette[k + 1];
            out[2] = palette[k + 0];
            out[3] = palette[k + 3];
            break;
        }
        break;
    case TYPE_RGB:
    case TYPE_RGB_RLE:
        switch (bytespp) {
        case 2:
            // This format is pretty funky. It's a 1A-5R-5G-5B layout,
            // with the first bit alpha (or unused if only 3 channels),
            // but thanks to the little-endianness, the order is pretty
            // bizarre. The bits are non-contiguous, so we have to
            // extract the relevant ones and synthesize the colour
            // values from the two bytes.
            // NOTE: This way of handling the pixel as two independent bytes
            // (as opposed to a single short int) makes it independent from
            // endianness.
            // Here's what the layout looks like:
            // MSb       unused   LSb
            //  v           v      v
            //  GGGBBBBB     RRRRRGG
            // [||||||||]  [||||||||]
            // While red and blue channels are quite self-explanatory, the
            // green one needs a few words. The 5 bits are composed of the
            // 2 from the second byte as the more significant and the 3 from
            // the first one as the less significant ones.

            // extract the bits to valid 5-bit integers and expand to full range
            out[0] = bit_range_convert<5, 8>((in[1] & 0x7C) >> 2);
            out[1] = bit_range_convert<5, 8>(((in[0] & 0xE0) >> 5)
                                             | ((in[1] & 0x03) << 3));
            out[2] = bit_range_convert<5, 8>(in[0] & 0x1F);
            if (m_spec.nchannels > 3)
                out[3] = (in[0] & 0x80) ? 255 : 0;
            break;
        case 3:
            out[0] = in[2];
            out[1] = in[1];
            out[2] = in[0];
            break;
        case 4:
            out[0] = in[2];
            out[1] = in[1];
            out[2] = in[0];
            out[3] = in[3];
            break;
        }
        break;
    case TYPE_GRAY:
    case TYPE_GRAY_RLE:
        if (bigendian()) {
            for (int i = bytespp - 1; i >= 0; i--)
                out[i] = in[bytespp - i - 1];
        } else
            memcpy(out, in, bytespp);
        break;
    }
}



template<class T>
static void
associateAlpha(T* data, int size, int channels, int alpha_channel, float gamma)
{
    T max = std::numeric_limits<T>::max();
    if (gamma == 1) {
        for (int x = 0; x < size; ++x, data += channels)
            for (int c = 0; c < channels; c++)
                if (c != alpha_channel) {
                    unsigned int f = data[c];
                    data[c]        = (f * data[alpha_channel]) / max;
                }
    } else {  //With gamma correction
        float inv_max = 1.0 / max;
        for (int x = 0; x < size; ++x, data += channels) {
            float alpha_associate = pow(data[alpha_channel] * inv_max, gamma);
            // We need to transform to linear space, associate the alpha, and
            // then transform back.  That is, if D = data[c], we want
            //
            // D' = max * ( (D/max)^(1/gamma) * (alpha/max) ) ^ gamma
            //
            // This happens to simplify to something which looks like
            // multiplying by a nonlinear alpha:
            //
            // D' = D * (alpha/max)^gamma
            for (int c = 0; c < channels; c++)
                if (c != alpha_channel)
                    data[c] = static_cast<T>(data[c] * alpha_associate);
        }
    }
}



bool
TGAInput::readimg()
{
    // how many bytes we actually read
    // for 15-bit read 2 bytes and ignore the 16th bit
    int bytespp    = (m_tga.bpp == 15) ? 2 : (m_tga.bpp / 8);
    int palbytespp = (m_tga.cmap_size == 15) ? 2 : (m_tga.cmap_size / 8);
    int alphabits  = m_tga.attr & 0x0F;
    if (alphabits == 0 && m_tga.bpp == 32)
        alphabits = 8;

    /*std::cerr << "[tga] bytespp = " << bytespp
              << " palbytespp = " << palbytespp
              << " alphabits = " << alphabits
              << "\n";*/

    m_buf.resize(m_spec.image_bytes());

    // read palette, if there is any
    unsigned char* palette = NULL;
    if (m_tga.cmap_type) {
        palette = new unsigned char[palbytespp * m_tga.cmap_length];
        if (!fread(palette, palbytespp, m_tga.cmap_length))
            return false;
    }

    unsigned char pixel[4];
    if (m_tga.type < TYPE_PALETTED_RLE) {
        // uncompressed image data
        unsigned char in[4];
        for (int y = m_spec.height - 1; y >= 0; y--) {
            for (int x = 0; x < m_spec.width; x++) {
                if (!fread(in, bytespp, 1))
                    return false;
                decode_pixel(in, pixel, palette, bytespp, palbytespp,
                             alphabits);
                memcpy(&m_buf[y * m_spec.width * m_spec.nchannels
                              + x * m_spec.nchannels],
                       pixel, m_spec.nchannels);
            }
        }
    } else {
        // Run Length Encoded image
        unsigned char in[5];
        int packet_size;
        for (int y = m_spec.height - 1; y >= 0; y--) {
            for (int x = 0; x < m_spec.width; x++) {
                if (!fread(in, 1 + bytespp, 1))
                    return false;
                packet_size = 1 + (in[0] & 0x7f);
                decode_pixel(&in[1], pixel, palette, bytespp, palbytespp,
                             alphabits);
                if (in[0] & 0x80) {  // run length packet
                    /*std::cerr << "[tga] run length packet "
                              << packet_size << "\n";*/
                    for (int i = 0; i < packet_size; i++) {
                        memcpy(&m_buf[y * m_spec.width * m_spec.nchannels
                                      + x * m_spec.nchannels],
                               pixel, m_spec.nchannels);
                        if (i < packet_size - 1) {
                            x++;
                            if (x >= m_spec.width) {
                                // run spans across multiple scanlines
                                x = 0;
                                if (y > 0)
                                    y--;
                                else
                                    goto loop_break;
                            }
                        }
                    }
                } else {  // non-rle packet
                    /*std::cerr << "[tga] non-run length packet "
                              << packet_size << "\n";*/
                    for (int i = 0; i < packet_size; i++) {
                        memcpy(&m_buf[y * m_spec.width * m_spec.nchannels
                                      + x * m_spec.nchannels],
                               pixel, m_spec.nchannels);
                        if (i < packet_size - 1) {
                            x++;
                            if (x >= m_spec.width) {
                                // run spans across multiple scanlines
                                x = 0;
                                if (y > 0)
                                    y--;
                                else
                                    goto loop_break;
                            }
                            // skip the packet header byte
                            if (!fread(&in[1], bytespp, 1))
                                return false;
                            decode_pixel(&in[1], pixel, palette, bytespp,
                                         palbytespp, alphabits);
                        }
                    }
                }
            }
        loop_break:;
        }
    }

    delete[] palette;

    // flip the image, if necessary
    if (m_tga.cmap_type)
        bytespp = palbytespp;
    // Y-flipping is now done in read_native_scanline instead
    /*if (m_tga.attr & FLAG_Y_FLIP) {
        //std::cerr << "[tga] y flipping\n";

        std::vector<unsigned char> flip (m_spec.width * bytespp);
        unsigned char *src, *dst, *tmp = &flip[0];
        for (int y = 0; y < m_spec.height / 2; y++) {
            src = &m_buf[(m_spec.height - y - 1) * m_spec.width * bytespp];
            dst = &m_buf[y * m_spec.width * bytespp];

            memcpy(tmp, src, m_spec.width * bytespp);
            memcpy(src, dst, m_spec.width * bytespp);
            memcpy(dst, tmp, m_spec.width * bytespp);
        }
    }*/
    if (m_tga.attr & FLAG_X_FLIP) {
        //std::cerr << "[tga] x flipping\n";

        std::vector<unsigned char> flip(bytespp * m_spec.width / 2);
        unsigned char *src, *dst, *tmp = &flip[0];
        for (int y = 0; y < m_spec.height; y++) {
            src = &m_buf[y * m_spec.width * bytespp];
            dst = &m_buf[(y * m_spec.width + m_spec.width / 2) * bytespp];

            memcpy(tmp, src, bytespp * m_spec.width / 2);
            memcpy(src, dst, bytespp * m_spec.width / 2);
            memcpy(dst, tmp, bytespp * m_spec.width / 2);
        }
    }

    if (m_alpha != TGA_ALPHA_PREMULTIPLIED) {
        // Convert to associated unless we were requested not to do so
        if (m_spec.alpha_channel != -1 && !m_keep_unassociated_alpha) {
            int size    = m_spec.width * m_spec.height;
            float gamma = m_spec.get_float_attribute("oiio:Gamma", 1.0f);

            associateAlpha((unsigned char*)&m_buf[0], size, m_spec.nchannels,
                           m_spec.alpha_channel, gamma);
        }
    }

    return true;
}



bool
TGAInput::close()
{
    if (m_file) {
        fclose(m_file);
        m_file = NULL;
    }

    init();  // Reset to initial state
    return true;
}



bool
TGAInput::read_native_scanline(int subimage, int miplevel, int y, int z,
                               void* data)
{
    lock_guard lock(m_mutex);
    if (!seek_subimage(subimage, miplevel))
        return false;

    if (m_buf.empty())
        readimg();

    if (m_tga.attr & FLAG_Y_FLIP)
        y = m_spec.height - y - 1;
    size_t size = spec().scanline_bytes();
    memcpy(data, &m_buf[0] + y * size, size);
    return true;
}

OIIO_PLUGIN_NAMESPACE_END