From https://www.khronos.org/opengl/wiki/Load_OpenGL_Functions

argument c is void* so that this dtor can be given as function pointer to the vector resize function*/

read a PNG, the result will be in the same color type as the PNG (hence "generic")*/

Deflate for a block of type "dynamic", that is, with freely, optimally, created huffman trees*/

Change ',' to '.' everywhere in buffer. * * We had a sophisticated way, but it did not work in WinCE. * @see https://github.com/open-source-parsers

log2 approximation. A slight bit faster than std::log. */

palette must have 4 * palettesize bytes allocated, and given in format RGBARGBARGBARGBA... returns 0 if the palette is opaque, returns 1 if the palett

get the tree of a deflated block with dynamic tree, the tree itself is also Huffman compressed with a known tree*/

get the tree of a deflated block with fixed tree, as specified in the deflate specification*/

Returns how many bits needed to represent given value (max 8 bit)*/

profile must already have been inited with mode. It's ok to set some parameters of profile to done already.*/

inflate a block with dynamic of fixed Huffman tree*/

Automatically chooses color type that gives smallest amount of bits in the output image, e.g. grey if there are only greyscale pixels, palette if ther

in an idat chunk, each scanline is a multiple of 8 bits, unlike the lodepng output buffer*/

read the information from the header and store it in the LodePNGInfo. return value is error*/

write given buffer to the file, overwriting the file, it doesn't append to it.*/

out must be buffer big enough to contain full image, and in must contain the full decompressed data from the IDAT chunks (with filter index bytes and

background color chunk (bKGD)*/

international text chunk (iTXt)*/

text chunk (tEXt)*/

compressed text chunk (zTXt)*/

put a pixel, given its RGBA16 color, into image of any color 16-bitdepth type*/

put a pixel, given its RGBA color, into image of any color type*/

write given buffer to the file, overwriting the file, it doesn't append to it.*/

build a table that decodes both magnitude and value of small ACs in one go.

decode image to YCbCr format