========================================================================= */

Fetch data and put it in the output buffer. Assumes state->x.have is 0. Data is either copied from the input file or decompressed from the input

Use read() to load a buffer -- return -1 on error, otherwise 0. Read from state->fd, and update state->eof, state->err, and state->msg as appropri

Skip len uncompressed bytes of output. Return -1 on error, 0 on success. */

This is the code to hack the first two bytes of the deflate stream (the * deflate header) to correct the windowBits value to match the actual data

Build a single 8-bit table: same as the 16-bit case but much simpler (and * typically much faster). Note that libpng currently does no sBIT processi

The same but for a complete string. */

This is similar to png_text_compress, above, except that it does not require * all of the data at once and, instead of buffering the compressed resul

Remove a channel - this used to be 'png_do_strip_filler' but it used a * somewhat weird combination of flags to determine what to do. All the calls

Clean up (or trim) a linked list of compression buffers. */

Read and check the IDHR chunk */

Read and check the palette */

Note: this does not correctly handle chunks that are > 64K under DOS */

Read the pCAL chunk (described in the PNG Extensions document) */

Read the sCAL chunk */

Note: this does not properly handle chunks that are > 64K under DOS */

Note: this does not correctly handle chunks that are > 64K under DOS */

Do all the *safe* initialization - 'safe' means that png_error won't be * called, so setting up the jmp_buf is not required. This means that anythin

This is used below to find the size of an image to pass to png_deflate_claim, * so it only needs to be accurate if the size is less than 16384 bytes

png_inflate_claim: claim the zstream for some nefarious purpose that involves * decompression. Returns Z_OK on success, else a zlib error code. It

Perform a partial read and decompress, producing 'avail_out' bytes and * reading from the current chunk as required. */

Read the chunk header (length + type name). * Put the type name into png_ptr->chunk_name, and return the length. */

Optional call to update the users info_ptr structure */

Add a filler byte on read, or remove a filler or alpha byte on write. * The filler type has changed in v0.95 to allow future 2-byte fillers * for 48

Compress the data in the compression state input */

Ship the compressed text out via chunk writes */

Write a tEXt chunk */

Write out all the unknown chunks for the current given location */

NON_MATCHING | Score: 10 (lower is better) there's uxtw instead of original sxtw