mirror of
https://git.code.sf.net/p/libpng/code.git
synced 2025-07-10 18:04:09 +02:00
d52c8eba99
Simplified API: change handling of PNG_IMAGE_FLAG_FAST to use PNG_COMPRESSION_HIGH_SPEED, and PNG_COMPRESSION_HIGH otherwise. Compression: add missing break statements that caused some compression settings to fall through to the 'HIGH' setting. Internal: remove png_struct::flags, it only stored the 'library mismatch' flag and that could never be accessed (because immediately after it was set the png_struct, which was on the stack, was eliminated.) Signed-off-by: John Bowler <jbowler@acm.org>
5049 lines
166 KiB
C
5049 lines
166 KiB
C
/* pngwutil.c - utilities to write a PNG file
|
|
*
|
|
* Last changed in libpng 1.7.0 [(PENDING RELEASE)]
|
|
* Copyright (c) 1998-2002,2004,2006-2016 Glenn Randers-Pehrson
|
|
* (Version 0.96 Copyright (c) 1996, 1997 Andreas Dilger)
|
|
* (Version 0.88 Copyright (c) 1995, 1996 Guy Eric Schalnat, Group 42, Inc.)
|
|
*
|
|
* This code is released under the libpng license.
|
|
* For conditions of distribution and use, see the disclaimer
|
|
* and license in png.h
|
|
*/
|
|
|
|
#include "pngpriv.h"
|
|
#define PNG_SRC_FILE PNG_SRC_FILE_pngwutil
|
|
|
|
#ifdef PNG_WRITE_SUPPORTED
|
|
|
|
#ifdef PNG_WRITE_INT_FUNCTIONS_SUPPORTED
|
|
/* Place a 32-bit number into a buffer in PNG byte order. We work
|
|
* with unsigned numbers for convenience, although one supported
|
|
* ancillary chunk uses signed (two's complement) numbers.
|
|
*/
|
|
void PNGAPI
|
|
png_save_uint_32(png_bytep buf, png_uint_32 i)
|
|
{
|
|
buf[0] = PNG_BYTE(i >> 24);
|
|
buf[1] = PNG_BYTE(i >> 16);
|
|
buf[2] = PNG_BYTE(i >> 8);
|
|
buf[3] = PNG_BYTE(i);
|
|
}
|
|
|
|
/* Place a 16-bit number into a buffer in PNG byte order.
|
|
* The parameter is declared unsigned int, not png_uint_16,
|
|
* just to avoid potential problems on pre-ANSI C compilers.
|
|
*/
|
|
void PNGAPI
|
|
png_save_uint_16(png_bytep buf, unsigned int i)
|
|
{
|
|
buf[0] = PNG_BYTE(i >> 8);
|
|
buf[1] = PNG_BYTE(i);
|
|
}
|
|
#endif /* WRITE_INT_FUNCTIONS */
|
|
|
|
/* Simple function to write the signature. If we have already written
|
|
* the magic bytes of the signature, or more likely, the PNG stream is
|
|
* being embedded into another stream and doesn't need its own signature,
|
|
* we should call png_set_sig_bytes() to tell libpng how many of the
|
|
* bytes have already been written.
|
|
*/
|
|
void PNGAPI
|
|
png_write_sig(png_structrp png_ptr)
|
|
{
|
|
png_byte png_signature[8] = {137, 80, 78, 71, 13, 10, 26, 10};
|
|
|
|
#ifdef PNG_IO_STATE_SUPPORTED
|
|
/* Inform the I/O callback that the signature is being written */
|
|
png_ptr->io_state = PNG_IO_WRITING | PNG_IO_SIGNATURE;
|
|
#endif
|
|
|
|
/* Write the rest of the 8 byte signature */
|
|
png_write_data(png_ptr, &png_signature[png_ptr->sig_bytes],
|
|
(png_size_t)(8 - png_ptr->sig_bytes));
|
|
|
|
if (png_ptr->sig_bytes < 3)
|
|
png_ptr->mode |= PNG_HAVE_PNG_SIGNATURE;
|
|
}
|
|
|
|
/* Write the start of a PNG chunk. The type is the chunk type.
|
|
* The total_length is the sum of the lengths of all the data you will be
|
|
* passing in png_write_chunk_data().
|
|
*/
|
|
static void
|
|
png_write_chunk_header(png_structrp png_ptr, png_uint_32 chunk_name,
|
|
png_uint_32 length)
|
|
{
|
|
png_byte buf[8];
|
|
|
|
#if defined(PNG_DEBUG) && (PNG_DEBUG > 0)
|
|
PNG_CSTRING_FROM_CHUNK(buf, chunk_name);
|
|
png_debug2(0, "Writing %s chunk, length = %lu", buf, (unsigned long)length);
|
|
#endif
|
|
|
|
if (png_ptr == NULL)
|
|
return;
|
|
|
|
#ifdef PNG_IO_STATE_SUPPORTED
|
|
/* Inform the I/O callback that the chunk header is being written.
|
|
* PNG_IO_CHUNK_HDR requires a single I/O call.
|
|
*/
|
|
png_ptr->io_state = PNG_IO_WRITING | PNG_IO_CHUNK_HDR;
|
|
#endif
|
|
|
|
/* Write the length and the chunk name */
|
|
png_save_uint_32(buf, length);
|
|
png_save_uint_32(buf + 4, chunk_name);
|
|
png_write_data(png_ptr, buf, 8);
|
|
|
|
/* Put the chunk name into png_ptr->chunk_name */
|
|
png_ptr->chunk_name = chunk_name;
|
|
|
|
/* Reset the crc and run it over the chunk name */
|
|
png_reset_crc(png_ptr, buf+4);
|
|
|
|
#ifdef PNG_IO_STATE_SUPPORTED
|
|
/* Inform the I/O callback that chunk data will (possibly) be written.
|
|
* PNG_IO_CHUNK_DATA does NOT require a specific number of I/O calls.
|
|
*/
|
|
png_ptr->io_state = PNG_IO_WRITING | PNG_IO_CHUNK_DATA;
|
|
#endif
|
|
}
|
|
|
|
void PNGAPI
|
|
png_write_chunk_start(png_structrp png_ptr, png_const_bytep chunk_string,
|
|
png_uint_32 length)
|
|
{
|
|
png_write_chunk_header(png_ptr, PNG_CHUNK_FROM_STRING(chunk_string), length);
|
|
}
|
|
|
|
/* Write the data of a PNG chunk started with png_write_chunk_header().
|
|
* Note that multiple calls to this function are allowed, and that the
|
|
* sum of the lengths from these calls *must* add up to the total_length
|
|
* given to png_write_chunk_header().
|
|
*/
|
|
void PNGAPI
|
|
png_write_chunk_data(png_structrp png_ptr, png_const_voidp data,
|
|
png_size_t length)
|
|
{
|
|
/* Write the data, and run the CRC over it */
|
|
if (png_ptr == NULL)
|
|
return;
|
|
|
|
if (data != NULL && length > 0)
|
|
{
|
|
png_write_data(png_ptr, data, length);
|
|
|
|
/* Update the CRC after writing the data,
|
|
* in case the user I/O routine alters it.
|
|
*/
|
|
png_calculate_crc(png_ptr, data, length);
|
|
}
|
|
}
|
|
|
|
/* Finish a chunk started with png_write_chunk_header(). */
|
|
void PNGAPI
|
|
png_write_chunk_end(png_structrp png_ptr)
|
|
{
|
|
png_byte buf[4];
|
|
|
|
if (png_ptr == NULL) return;
|
|
|
|
#ifdef PNG_IO_STATE_SUPPORTED
|
|
/* Inform the I/O callback that the chunk CRC is being written.
|
|
* PNG_IO_CHUNK_CRC requires a single I/O function call.
|
|
*/
|
|
png_ptr->io_state = PNG_IO_WRITING | PNG_IO_CHUNK_CRC;
|
|
#endif
|
|
|
|
/* Write the crc in a single operation */
|
|
png_save_uint_32(buf, png_ptr->crc);
|
|
|
|
png_write_data(png_ptr, buf, (png_size_t)4);
|
|
}
|
|
|
|
/* Write a PNG chunk all at once. The type is an array of ASCII characters
|
|
* representing the chunk name. The array must be at least 4 bytes in
|
|
* length, and does not need to be null terminated. To be safe, pass the
|
|
* pre-defined chunk names here, and if you need a new one, define it
|
|
* where the others are defined. The length is the length of the data.
|
|
* All the data must be present. If that is not possible, use the
|
|
* png_write_chunk_start(), png_write_chunk_data(), and png_write_chunk_end()
|
|
* functions instead.
|
|
*/
|
|
static void
|
|
png_write_complete_chunk(png_structrp png_ptr, png_uint_32 chunk_name,
|
|
png_const_voidp data, png_size_t length)
|
|
{
|
|
if (png_ptr == NULL)
|
|
return;
|
|
|
|
/* On 64 bit architectures 'length' may not fit in a png_uint_32. */
|
|
if (length > PNG_UINT_31_MAX)
|
|
png_error(png_ptr, "length exceeds PNG maximum");
|
|
|
|
png_write_chunk_header(png_ptr, chunk_name, (png_uint_32)/*SAFE*/length);
|
|
png_write_chunk_data(png_ptr, data, length);
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
|
|
/* This is the API that calls the internal function above. */
|
|
void PNGAPI
|
|
png_write_chunk(png_structrp png_ptr, png_const_bytep chunk_string,
|
|
png_const_voidp data, png_size_t length)
|
|
{
|
|
png_write_complete_chunk(png_ptr, PNG_CHUNK_FROM_STRING(chunk_string), data,
|
|
length);
|
|
}
|
|
|
|
static png_alloc_size_t
|
|
png_write_row_buffer_size(png_const_structrp png_ptr)
|
|
/* Returns the width of the widest pass in the first row of an interlaced
|
|
* image. Passes in the first row are: 0.5.3.5.1.5.3.5, so the widest row is
|
|
* normally the one from pass 5. The only exception is if the image is only
|
|
* one pixel wide, so:
|
|
*/
|
|
#define PNG_FIRST_ROW_MAX_WIDTH(w) (w > 1U ? PNG_PASS_COLS(w, 5U) : 1U)
|
|
|
|
/* For interlaced images the count of pixels is rounded up to a the number of
|
|
* pixels in the first pass (numbered 0). This ensures that passes before
|
|
* the last can be packed in the buffer without overflow.
|
|
*/
|
|
{
|
|
png_alloc_size_t w;
|
|
|
|
/* If the image is interlaced adjust 'w' for the interlacing: */
|
|
if (png_ptr->interlaced != PNG_INTERLACE_NONE)
|
|
{
|
|
/* Take advantage of the fact that 1-row interlaced PNGs require half the
|
|
* normal row width:
|
|
*/
|
|
if (png_ptr->height == 1U) /* no pass 6 */
|
|
w = PNG_FIRST_ROW_MAX_WIDTH(png_ptr->width);
|
|
|
|
/* Otherwise round up to a multiple of 8. This may waste a few (less
|
|
* than 8) bytes for PNGs with a height less than 57 but this hardly
|
|
* matters.
|
|
*/
|
|
else
|
|
w = (png_ptr->width + 7U) & ~7U;
|
|
}
|
|
|
|
else
|
|
w = png_ptr->width;
|
|
|
|
/* The rounding above may leave 'w' exactly 2^31 */
|
|
debug(w <= 0x80000000U);
|
|
|
|
switch (png_ptr->row_output_pixel_depth)
|
|
{
|
|
/* This would happen if the function is called before png_write_IHDR. */
|
|
default: NOT_REACHED; return 0;
|
|
|
|
case 1: w = (w+7) >> 3; break;
|
|
case 2: w = (w+3) >> 2; break;
|
|
case 4: w = (w+1) >> 1; break;
|
|
case 8: break;
|
|
case 16: w <<= 1; break; /* overflow: w is set to 0, which is OK */
|
|
|
|
/* For the remaining cases the answer is w*bytes; where bytes is 3,4,6
|
|
* or 8. This may overflow 32 bits. There is no way to compute the
|
|
* result on an arbitrary platform, so test the maximum of a (size_t)
|
|
* against w for each possible byte depth:
|
|
*/
|
|
# define CASE(b)\
|
|
case b*8:\
|
|
if (w <= (PNG_SIZE_MAX/b)/*compile-time constant*/)\
|
|
return w * b;\
|
|
return 0;
|
|
|
|
CASE(3)
|
|
CASE(4)
|
|
CASE(6)
|
|
CASE(8)
|
|
|
|
# undef CASE
|
|
}
|
|
|
|
/* This is the low bit depth case. The following can never be false on
|
|
* systems with a 32-bit or greater size_t:
|
|
*/
|
|
if (w <= PNG_SIZE_MAX)
|
|
return w;
|
|
|
|
return 0U;
|
|
}
|
|
|
|
/* Release memory used by the deflate mechanism */
|
|
static void
|
|
png_deflateEnd(png_const_structrp png_ptr, z_stream *zs, int check)
|
|
{
|
|
if (zs->state != NULL)
|
|
{
|
|
int ret = deflateEnd(zs);
|
|
|
|
/* Z_DATA_ERROR means there was pending output. */
|
|
if ((ret != Z_OK && (check || ret != Z_DATA_ERROR)) || zs->state != NULL)
|
|
{
|
|
png_zstream_error(zs, ret);
|
|
|
|
if (check)
|
|
png_error(png_ptr, zs->msg);
|
|
|
|
else
|
|
png_warning(png_ptr, zs->msg);
|
|
|
|
zs->state = NULL;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* compression_buffer (new in 1.6.0) is just a linked list of temporary buffers. * From 1.6.0 it is retained in png_struct so that it will be correctly freed in
|
|
* the event of a write error (previous implementations just leaked memory.)
|
|
*
|
|
* From 1.7.0 the size is fixed to the same as the (uncompressed) row buffer
|
|
* size. This avoids allocating a large chunk of memory when compressing small
|
|
* images. This type is also opaque outside this file.
|
|
*/
|
|
typedef struct png_compression_buffer
|
|
{
|
|
struct png_compression_buffer *next;
|
|
png_byte output[PNG_ROW_BUFFER_SIZE];
|
|
} png_compression_buffer, *png_compression_bufferp;
|
|
|
|
/* png_compression_buffer methods */
|
|
/* Deleting a compression buffer deletes the whole list: */
|
|
static void
|
|
png_free_compression_buffer(png_const_structrp png_ptr,
|
|
png_compression_bufferp *listp)
|
|
{
|
|
png_compression_bufferp list = *listp;
|
|
|
|
if (list != NULL)
|
|
{
|
|
*listp = NULL;
|
|
|
|
do
|
|
{
|
|
png_compression_bufferp next = list->next;
|
|
|
|
png_free(png_ptr, list);
|
|
list = next;
|
|
}
|
|
while (list != NULL);
|
|
}
|
|
}
|
|
|
|
/* Return the next compression buffer in the list, allocating it if necessary.
|
|
* The caller must update 'end' if required; this just moves down the list.
|
|
*/
|
|
static png_compression_bufferp
|
|
png_get_compression_buffer(png_const_structrp png_ptr,
|
|
png_compression_bufferp *end)
|
|
{
|
|
png_compression_bufferp next = *end;
|
|
|
|
if (next == NULL)
|
|
{
|
|
next = png_voidcast(png_compression_bufferp, png_malloc_base(png_ptr,
|
|
sizeof *next));
|
|
|
|
/* Check for OOM: this is a recoverable error for non-critical chunks, let
|
|
* the caller decide what to do rather than issuing a png_error here.
|
|
*/
|
|
if (next != NULL)
|
|
{
|
|
next->next = NULL; /* initialize the buffer */
|
|
*end = next;
|
|
}
|
|
}
|
|
|
|
return next; /* may still be NULL on OOM */
|
|
}
|
|
|
|
/* This structure is used to hold all the data for zlib compression of a single
|
|
* stream of data. It may be re-used, it stores the compressed data internally
|
|
* and can handle arbitrary input and output.
|
|
*
|
|
* 'list' is the output data contained in compression buffers, 'end' points to
|
|
* list at the start and is advanced down the compression buffer list (extending
|
|
* it as required) as the data is written. If 'end' points into a compression
|
|
* buffer (does not point to 'list') that is the buffer in use in
|
|
* z_stream::{next,avail}_out.
|
|
*
|
|
* Compression may be performed in multiple steps, '*end' always points to the
|
|
* compression buffer *after* the one that is in use, so 'end' is pointing
|
|
* *into* the one in use.
|
|
*
|
|
* end(on entry) .... end ....... end(on exit)
|
|
* | | |
|
|
* | | |
|
|
* V +----V-----+ +-----V----+ +----------+
|
|
* list ---> | next --+--> | next --+--> | next |
|
|
* | output[] | | output[] | | output[] |
|
|
* +----------+ +----------+ +----------+
|
|
* [in use] [unused]
|
|
*
|
|
* These invariants should always hold:
|
|
*
|
|
* 1) If zs.state is NULL decompression is not in progress, list may be non-NULL
|
|
* but end could be anything;
|
|
*
|
|
* 2) Otherwise if zs.next_out is NULL list will be NULL and end will point at
|
|
* list, len, overflow and start will be 0;
|
|
*
|
|
* 3) Otherwise list is non-NULL and end points at the 'next' element of an
|
|
* in-use compression buffer. zs.next_out points into the 'output' element
|
|
* of the same buffer. {overflow, len} is the amount of compressed data, len
|
|
* being the low 31 bits, overflow being the higher bits. start is used for
|
|
* writing and is the index of the first byte in list->output to write,
|
|
* {overflow, len} does not include start.
|
|
*/
|
|
typedef struct
|
|
{
|
|
z_stream zs; /* zlib compression data */
|
|
png_compression_bufferp list; /* Head of the buffer list */
|
|
png_compression_bufferp *end; /* Pointer to last 'next' pointer */
|
|
png_uint_32 len; /* Bottom 31 bits of data length */
|
|
unsigned int overflow; /* Top bits of data length */
|
|
unsigned int start; /* Start of data in first block */
|
|
} png_zlib_compress, *png_zlib_compressp;
|
|
|
|
/* png_zlib_compress methods */
|
|
/* Initialize the compress structure. The z_stream itself is not initialized,
|
|
* however the the 'user' fields are set, including {next,avail}_{in,out}. The
|
|
* initialization does not change 'list', however it does set 'end' to point to
|
|
* it, effectively truncating the list.
|
|
*/
|
|
static void
|
|
png_zlib_compress_init(png_structrp png_ptr, png_zlib_compressp pz)
|
|
{
|
|
/* png_zlib_compress z_stream: */
|
|
pz->zs.zalloc = png_zalloc;
|
|
pz->zs.zfree = png_zfree;
|
|
/* NOTE: this does not destroy 'restrict' because in all the functions herein
|
|
* *png_ptr is only ever accessed via *either* pz->zs.opaque *or* a passed in
|
|
* png_ptr.
|
|
*/
|
|
pz->zs.opaque = png_ptr;
|
|
|
|
pz->zs.next_in = NULL;
|
|
pz->zs.avail_in = 0U;
|
|
pz->zs.total_in = 0U;
|
|
|
|
pz->zs.next_out = NULL;
|
|
pz->zs.avail_out = 0U;
|
|
pz->zs.total_out = 0U;
|
|
|
|
pz->zs.msg = PNGZ_MSG_CAST("zlib success"); /* safety */
|
|
|
|
/* pz->list preserved */
|
|
pz->end = &pz->list;
|
|
pz->len = 0U;
|
|
pz->overflow = 0U;
|
|
pz->start = 0U;
|
|
}
|
|
|
|
/* Return the png_ptr: this is defined here for all the remaining
|
|
* png_zlib_compress methods because they are only ever called with zs
|
|
* initialized.
|
|
*/
|
|
#define png_ptr png_voidcast(png_const_structrp, pz->zs.opaque)
|
|
|
|
#if PNG_RELEASE_BUILD
|
|
# define png_zlib_compress_validate(pz, in_use) ((void)0)
|
|
#else /* !RELEASE_BUILD */
|
|
static void
|
|
png_zlib_compress_validate(png_zlib_compressp pz, int in_use)
|
|
{
|
|
const uInt o_size = sizeof pz->list->output;
|
|
|
|
affirm(pz->end != NULL && (in_use || (pz->zs.next_in == NULL &&
|
|
pz->zs.avail_in == 0U && *pz->end == NULL)));
|
|
|
|
if (pz->overflow == 0U && pz->len == 0U && pz->start == 0U) /* empty */
|
|
{
|
|
affirm((pz->end == &pz->list && pz->zs.next_out == NULL
|
|
&& pz->zs.avail_out == 0U) ||
|
|
(pz->list != NULL && pz->end == &pz->list->next &&
|
|
pz->zs.next_out == pz->list->output &&
|
|
pz->zs.avail_out == o_size));
|
|
}
|
|
|
|
else /* not empty */
|
|
{
|
|
png_compression_bufferp *ep = &pz->list, list;
|
|
png_uint_32 o, l;
|
|
|
|
affirm(*ep != NULL && pz->zs.next_out != NULL);
|
|
|
|
/* Check the list length: */
|
|
o = pz->overflow;
|
|
l = pz->len;
|
|
affirm((l & 0x80000000U) == 0U && (o & 0x80000000U) == 0U);
|
|
|
|
do
|
|
{
|
|
list = *ep;
|
|
l -= o_size;
|
|
if (l & 0x80000000U) --o, l &= 0x7FFFFFFFU;
|
|
ep = &list->next;
|
|
}
|
|
while (ep != pz->end);
|
|
|
|
l += pz->start;
|
|
l += pz->zs.avail_out;
|
|
if (l & 0x80000000U) ++o, l &= 0x7FFFFFFFU;
|
|
|
|
affirm(o == 0U && l == 0U && pz->zs.next_out >= list->output &&
|
|
pz->zs.next_out + pz->zs.avail_out == list->output + o_size);
|
|
}
|
|
}
|
|
#endif /* !RELEASE_BUILD */
|
|
|
|
/* Destroy one zlib compress structure. */
|
|
static void
|
|
png_zlib_compress_destroy(png_zlib_compressp pz, int check)
|
|
{
|
|
/* If the 'opaque' pointer is NULL this png_zlib_compress was never
|
|
* initialized, so do nothing.
|
|
*/
|
|
if (png_ptr != NULL)
|
|
{
|
|
if (pz->zs.state != NULL)
|
|
{
|
|
if (check)
|
|
png_zlib_compress_validate(pz, 0/*in_use*/);
|
|
|
|
png_deflateEnd(png_ptr, &pz->zs, check);
|
|
}
|
|
|
|
pz->end = &pz->list; /* safety */
|
|
png_free_compression_buffer(png_ptr, &pz->list);
|
|
}
|
|
}
|
|
|
|
/* Ensure that space is available for output, returns the amount of space
|
|
* available, 0 on OOM. This updates pz->zs.avail_out (etc) as required.
|
|
*/
|
|
static uInt
|
|
png_zlib_compress_avail_out(png_zlib_compressp pz)
|
|
{
|
|
uInt avail_out = pz->zs.avail_out;
|
|
|
|
png_zlib_compress_validate(pz, 1/*in_use*/);
|
|
|
|
if (avail_out == 0U)
|
|
{
|
|
png_compression_bufferp next;
|
|
|
|
affirm(pz->end == &pz->list || (pz->end != NULL && pz->list != NULL));
|
|
next = png_get_compression_buffer(png_ptr, pz->end);
|
|
|
|
if (next != NULL)
|
|
{
|
|
pz->zs.next_out = next->output;
|
|
pz->zs.avail_out = avail_out = sizeof next->output;
|
|
pz->end = &next->next;
|
|
}
|
|
|
|
/* else return 0: OOM */
|
|
}
|
|
|
|
else
|
|
affirm(pz->end != NULL && pz->list != NULL);
|
|
|
|
return avail_out;
|
|
}
|
|
|
|
/* Compress the given data given an initialized png_zlib_compress structure.
|
|
* This may be called multiple times, interleaved with writes as required.
|
|
*
|
|
* The input data is passed in in pz->zs.next_in, however the length of the data
|
|
* is in 'input_len' (to avoid the zlib uInt limit) and pz->zs.avail_in is
|
|
* overwritten (and left at 0).
|
|
*
|
|
* The output information is used and the amount of compressed data is added on
|
|
* to pz->{overflow,len}.
|
|
*
|
|
* If 'limit' is a limit on the amount of data to add to the output (not the
|
|
* total amount). The function will retun Z_BUF_ERROR if the limit is reached
|
|
* and the function will never produce more (additional) compressed data than
|
|
* the limit.
|
|
*
|
|
* All of zstream::next_in[input] is consumed if a success code is returned
|
|
* (Z_OK or Z_STREAM_END if flush is Z_FINISH), otherwise next_in may be used to
|
|
* determine how much was compressed.
|
|
*
|
|
* pz->overflow is not checked for overflow, so if 'limit' is not set overflow
|
|
* is possible. The caller must guard against this when supplying a limit of 0.
|
|
*/
|
|
static int
|
|
png_compress(
|
|
png_zlib_compressp pz,
|
|
png_alloc_size_t input_len, /* Length of data to be compressed */
|
|
png_uint_32 limit, /* Limit on amount of compressed data made */
|
|
int flush) /* Flush parameter at end of input */
|
|
{
|
|
const int unlimited = (limit == 0U);
|
|
|
|
/* Sanity checking: */
|
|
affirm(pz->zs.state != NULL &&
|
|
(pz->zs.next_out == NULL
|
|
? pz->end == &pz->list && pz->len == 0U && pz->overflow == 0U
|
|
: pz->list != NULL && pz->end != NULL));
|
|
implies(pz->zs.next_out == NULL, pz->zs.avail_out == 0);
|
|
|
|
for (;;)
|
|
{
|
|
uInt extra;
|
|
|
|
/* OUTPUT: make sure some space is available: */
|
|
if (png_zlib_compress_avail_out(pz) == 0U)
|
|
return Z_MEM_ERROR;
|
|
|
|
/* INPUT: limit the deflate call input to ZLIB_IO_MAX: */
|
|
/* Adjust the input counters: */
|
|
{
|
|
uInt avail_in = ZLIB_IO_MAX;
|
|
|
|
if (avail_in > input_len)
|
|
avail_in = (uInt)/*SAFE*/input_len;
|
|
|
|
input_len -= avail_in;
|
|
pz->zs.avail_in = avail_in;
|
|
}
|
|
|
|
if (!unlimited && pz->zs.avail_out > limit)
|
|
{
|
|
extra = (uInt)/*SAFE*/(pz->zs.avail_out - limit); /* unused bytes */
|
|
pz->zs.avail_out = (uInt)/*SAFE*/limit;
|
|
limit = 0U;
|
|
}
|
|
|
|
else
|
|
{
|
|
extra = 0U;
|
|
limit -= pz->zs.avail_out; /* limit >= 0U */
|
|
}
|
|
|
|
pz->len += pz->zs.avail_out; /* maximum that can be produced */
|
|
|
|
/* Compress the data */
|
|
{
|
|
int ret = deflate(&pz->zs, input_len > 0U ? Z_NO_FLUSH : flush);
|
|
|
|
/* Claw back input data that was not consumed (because avail_in is
|
|
* reset above every time round the loop) and correct the output
|
|
* length.
|
|
*/
|
|
input_len += pz->zs.avail_in;
|
|
pz->zs.avail_in = 0; /* safety */
|
|
pz->len -= pz->zs.avail_out;
|
|
|
|
if (pz->len & 0x80000000U)
|
|
++pz->overflow, pz->len &= 0x7FFFFFFFU;
|
|
|
|
limit += pz->zs.avail_out;
|
|
pz->zs.avail_out += extra;
|
|
|
|
/* Check the error code: */
|
|
switch (ret)
|
|
{
|
|
case Z_OK:
|
|
if (pz->zs.avail_out > extra)
|
|
{
|
|
/* zlib had output space, so all the input should have been
|
|
* consumed:
|
|
*/
|
|
affirm(input_len == 0U /* else unexpected stop */ &&
|
|
flush != Z_FINISH/* ret != Z_STREAM_END */);
|
|
return Z_OK;
|
|
}
|
|
|
|
else
|
|
{
|
|
/* zlib ran out of output space, produce some more. If the
|
|
* limit is 0 at this point, however, no more space is
|
|
* available.
|
|
*/
|
|
if (unlimited || limit > 0U)
|
|
break; /* Allocate more output */
|
|
|
|
/* No more output space available, but the input may have all
|
|
* been consumed.
|
|
*/
|
|
if (input_len == 0U && flush != Z_FINISH)
|
|
return Z_OK;
|
|
|
|
/* Input all consumed, but insufficient space to flush the
|
|
* output; this is the Z_BUF_ERROR case.
|
|
*/
|
|
return Z_BUF_ERROR;
|
|
}
|
|
|
|
case Z_STREAM_END:
|
|
affirm(input_len == 0U && flush == Z_FINISH);
|
|
return Z_STREAM_END;
|
|
|
|
case Z_BUF_ERROR:
|
|
/* This means that we are flushing all the output; expect
|
|
* avail_out and input_len to be 0.
|
|
*
|
|
* NOTE: if png_compress is called with input_len 0 and flush set
|
|
* to Z_NO_FLUSH this affirm will fire because zlib will have no
|
|
* work to do.
|
|
*/
|
|
affirm(input_len == 0U && pz->zs.avail_out == extra);
|
|
/* Allocate another buffer */
|
|
break;
|
|
|
|
default:
|
|
/* An error */
|
|
return ret;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
#undef png_ptr /* remove definition using a png_zlib_compressp */
|
|
|
|
/* All the compression state is held here, it is allocated when required. This
|
|
* ensures that the read code doesn't carry the overhead of the much less
|
|
* frequently used write stuff.
|
|
*
|
|
* TODO: make png_create_write_struct allocate this stuff after the main
|
|
* png_struct.
|
|
*/
|
|
struct filter_selector; /* Used only for filter selection */
|
|
|
|
typedef struct png_zlib_state
|
|
{
|
|
png_zlib_compress s; /* Primary compression state */
|
|
png_compression_bufferp stash; /* Unused compression buffers */
|
|
|
|
# define ps_png_ptr(ps) png_upcast(png_const_structrp, (ps)->s.zs.opaque)
|
|
/* A png_ptr, used below in functions that only have a png_zlib_state.
|
|
* NOTE: the png_zlib_compress must have been initialized!
|
|
*/
|
|
|
|
png_uint_32 zlib_max_pixels;
|
|
/* Maximum number of pixels that zlib can handle at once; the lesser of
|
|
* the PNG maximum and the maximum that will fit in (uInt)-1 bytes. This
|
|
* number of pixels may not be byte aligned.
|
|
*/
|
|
png_uint_32 zlib_max_aligned_pixels;
|
|
/* The maximum number of pixels that zlib can handle while maintaining a
|
|
* buffer byte alignment of PNG_ROW_BUFFER_BYTE_ALIGN; <= the previous
|
|
* value.
|
|
*/
|
|
|
|
png_alloc_size_t write_row_size;
|
|
/* Size of the PNG row (without the filter byte) in bytes or 0 if it is
|
|
* too large to be cached.
|
|
*/
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* During write libpng needs the previous row when writing a new row with
|
|
* up, avg or paeth and one or more image rows when performing filter
|
|
* selection. So if performing filter selection typically two or more
|
|
* rows are required while if no filter selection is to be done only the
|
|
* previous row pointer is required.
|
|
*/
|
|
png_bytep previous_write_row; /* Last row written, if any */
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
png_bytep current_write_row; /* Row being written */
|
|
struct filter_selector *selector; /* Data for filter selection */
|
|
png_uint_32 filter_select_window;
|
|
/* The number of bytes of uncompressed PNG data which are assumed to
|
|
* be relevant when doing filter selection. Limited to 8453377
|
|
* (about 2^23); the maximum number of bytes that can be encoded in
|
|
* the largest deflate window.
|
|
*/
|
|
# define PNG_FILTER_SELECT_WINDOW_MAX 8453377U
|
|
png_byte filter_select_threshold;
|
|
/* If the number of distinct codes seen in the PNG data are below
|
|
* this threshold the PNG data will not be filtered (if the 'none'
|
|
* filter is allowed). If this is still true and a particular
|
|
* filter does not add new codes that filter will be used.
|
|
*/
|
|
png_byte filter_select_threshold2;
|
|
/* If the number of distinct codes that result by using a particular
|
|
* filter is below this second threshold that filter will be used.
|
|
* (When multiple filters pass this criterion the lowest numbered
|
|
* one producing the lowest number of new codes will be
|
|
* chosen.)
|
|
*/
|
|
# endif /* SELECT_FILTER */
|
|
|
|
unsigned int row_buffer_max_pixels;
|
|
/* The maximum number of pixels that can fit in PNG_ROW_BUFFER_SIZE
|
|
* bytes; not necessary a whole number of bytes.
|
|
*/
|
|
unsigned int row_buffer_max_aligned_pixels;
|
|
/* The maximum number of pixels that can fit in PNG_ROW_BUFFER_SIZE
|
|
* bytes while maintaining PNG_ROW_BUFFER_BYTE_ALIGN alignment.
|
|
*/
|
|
|
|
unsigned int filter_mask :8; /* mask of filters to consider on NEXT row */
|
|
# define PREVIOUS_ROW_FILTERS\
|
|
(PNG_FILTER_UP|PNG_FILTER_AVG|PNG_FILTER_PAETH)
|
|
unsigned int filters :8; /* Filters for current row */
|
|
unsigned int save_row :2; /* As below: */
|
|
# define SAVE_ROW_UNSET 0U
|
|
# define SAVE_ROW_OFF 1U /* Previous-row filters will be ignored */
|
|
# define SAVE_ROW_DEFAULT 2U /* Default to save rows set by libpng */
|
|
# define SAVE_ROW_ON 3U /* Force rows to be saved */
|
|
# define SAVE_ROW(ps) ((ps)->save_row >= SAVE_ROW_DEFAULT)
|
|
# endif /* WRITE_FILTER */
|
|
|
|
/* Compression settings: see below for how these are encoded. */
|
|
png_uint_32 pz_IDAT; /* Settings for the image */
|
|
png_uint_32 pz_iCCP; /* Settings for iCCP chunks */
|
|
png_uint_32 pz_text; /* Settings for text chunks */
|
|
png_uint_32 pz_current; /* Last set settings */
|
|
|
|
# ifdef PNG_WRITE_FLUSH_SUPPORTED
|
|
png_uint_32 flush_dist; /* how many rows apart to flush, 0 - no flush */
|
|
png_uint_32 flush_rows; /* number of rows written since last flush */
|
|
# endif /* WRITE_FLUSH */
|
|
} png_zlib_state;
|
|
|
|
/* Create the zlib state: */
|
|
static void
|
|
png_create_zlib_state(png_structrp png_ptr)
|
|
{
|
|
png_zlib_statep ps = png_voidcast(png_zlib_state*,
|
|
png_malloc(png_ptr, sizeof *ps));
|
|
|
|
/* Clear to NULL/0: */
|
|
memset(ps, 0, sizeof *ps);
|
|
|
|
debug(png_ptr->zlib_state == NULL);
|
|
png_ptr->zlib_state = ps;
|
|
png_zlib_compress_init(png_ptr, &ps->s);
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
ps->previous_write_row = NULL;
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
ps->current_write_row = NULL;
|
|
ps->selector = NULL;
|
|
# endif /* SELECT_FILTER */
|
|
# endif /* WRITE_FILTER */
|
|
# ifdef PNG_WRITE_FLUSH_SUPPORTED
|
|
/* Set this to prevent flushing by making it larger than the number
|
|
* of rows in the largest interlaced PNG; PNG_UINT_31_MAX times
|
|
* (1/8+1/8+1/8+1/4+1/4+1/2+1/2); 1.875, or 15/8
|
|
*/
|
|
ps->flush_dist = 0xEFFFFFFFU;
|
|
# endif /* WRITE_FLUSH */
|
|
}
|
|
|
|
static void
|
|
png_zlib_state_set_buffer_limits(png_const_structrp png_ptr, png_zlib_statep ps)
|
|
/* Delayed initialization of the zlib state maxima; this is not done above in
|
|
* case the zlib_state is created before the IHDR has been written, which
|
|
* would lead to the various png_struct fields used below being
|
|
* uninitialized.
|
|
*/
|
|
{
|
|
/* Initialization of the buffer size constants. */
|
|
const unsigned int bpp = PNG_PIXEL_DEPTH(*png_ptr);
|
|
const unsigned int byte_pp = bpp >> 3; /* May be 0 */
|
|
const unsigned int pixel_block =
|
|
/* Number of pixels required to maintain PNG_ROW_BUFFER_BYTE_ALIGN
|
|
* alignment. For multi-byte pixels use the first set bit to determine
|
|
* if the pixels have a greater alignment already.
|
|
*/
|
|
bpp < 8U ?
|
|
PNG_ROW_BUFFER_BYTE_ALIGN * (8U/bpp) :
|
|
PNG_ROW_BUFFER_BYTE_ALIGN <= (byte_pp & -byte_pp) ?
|
|
1U :
|
|
PNG_ROW_BUFFER_BYTE_ALIGN / (byte_pp & -byte_pp);
|
|
|
|
/* pixel_block must always be a power of two: */
|
|
debug(bpp > 0 && pixel_block > 0 &&
|
|
(pixel_block & -pixel_block) == pixel_block &&
|
|
((8U*PNG_ROW_BUFFER_BYTE_ALIGN-1U) & (pixel_block*bpp)) == 0U);
|
|
|
|
/* Zlib maxima */
|
|
{
|
|
png_uint_32 max = (uInt)-1; /* max bytes */
|
|
|
|
if (bpp <= 8U)
|
|
{
|
|
/* Maximum number of bytes PNG can generate in the lower bit depth
|
|
* cases:
|
|
*/
|
|
png_uint_32 png_max =
|
|
(0x7FFFFFFF + PNG_ADDOF(bpp)) >> PNG_SHIFTOF(bpp);
|
|
|
|
if (png_max < max)
|
|
max = 0x7FFFFFFF;
|
|
}
|
|
|
|
else /* bpp > 8U */
|
|
{
|
|
max /= byte_pp;
|
|
if (max > 0x7FFFFFFF)
|
|
max = 0x7FFFFFFF;
|
|
}
|
|
|
|
/* So this is the maximum number of pixels regardless of alignment: */
|
|
ps->zlib_max_pixels = max;
|
|
|
|
/* For byte alignment the value has to be a multiple of pixel_block and
|
|
* that is a power of 2, so:
|
|
*/
|
|
ps->zlib_max_aligned_pixels = max & ~(pixel_block-1U);
|
|
}
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* PNG_ROW_BUFFER maxima; this is easier because PNG_ROW_BUFFER_SIZE is
|
|
* limited so that the number of bits fits in any ANSI-C (unsigned int).
|
|
*/
|
|
{
|
|
const unsigned int max = (8U * PNG_ROW_BUFFER_SIZE) / bpp;
|
|
|
|
ps->row_buffer_max_pixels = max;
|
|
ps->row_buffer_max_aligned_pixels = max & ~(pixel_block-1U);
|
|
}
|
|
# endif /* WRITE_FILTER */
|
|
|
|
/* NOTE: this will be 0 for very long rows on 32-bit or less systems */
|
|
ps->write_row_size = png_write_row_buffer_size(png_ptr);
|
|
}
|
|
|
|
static png_zlib_statep
|
|
get_zlib_state(png_structrp png_ptr)
|
|
{
|
|
if (png_ptr->zlib_state == NULL)
|
|
png_create_zlib_state(png_ptr);
|
|
|
|
return png_ptr->zlib_state;
|
|
}
|
|
|
|
/* Internal API to clean up all the deflate related stuff, including the buffer
|
|
* lists.
|
|
*/
|
|
static void /* PRIVATE */
|
|
png_deflate_release(png_structrp png_ptr, png_zlib_statep ps, int check)
|
|
{
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* Free any mode-specific data that is owned here: */
|
|
if (ps->previous_write_row != NULL)
|
|
{
|
|
png_bytep p = ps->previous_write_row;
|
|
ps->previous_write_row = NULL;
|
|
png_free(png_ptr, p);
|
|
}
|
|
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
if (ps->current_write_row != NULL)
|
|
{
|
|
png_bytep p = ps->current_write_row;
|
|
ps->current_write_row = NULL;
|
|
png_free(png_ptr, p);
|
|
}
|
|
|
|
if (ps->selector != NULL)
|
|
{
|
|
struct filter_selector *s = ps->selector;
|
|
ps->selector = NULL;
|
|
png_free(png_ptr, s);
|
|
}
|
|
# endif /* SELECT_FILTER */
|
|
# endif /* WRITE_FILTER */
|
|
|
|
/* The main z_stream opaque pointer needs to remain set to png_ptr; it is
|
|
* only set once.
|
|
*/
|
|
png_zlib_compress_destroy(&ps->s, check);
|
|
png_free_compression_buffer(png_ptr, &ps->stash);
|
|
}
|
|
|
|
void /* PRIVATE */
|
|
png_deflate_destroy(png_structrp png_ptr)
|
|
{
|
|
png_zlib_statep ps = png_ptr->zlib_state;
|
|
|
|
if (ps != NULL)
|
|
{
|
|
png_deflate_release(png_ptr, ps, 0/*check*/);
|
|
png_ptr->zlib_state = NULL;
|
|
png_free(png_ptr, ps);
|
|
}
|
|
}
|
|
|
|
/* Compression settings.
|
|
*
|
|
* These are stored packed into a png_uint_32 to make comparison with the
|
|
* current setting quick. The packing method uses four bits for each setting
|
|
* and reserves '0' for unset.
|
|
*
|
|
* ps_<setting>_base: The lowest valid value (encoded as 1).
|
|
* ps_<setting>_max: The highest valid value.
|
|
* ps_<setting>_pos: The position in the range 0..3 (shift of 0..12).
|
|
*
|
|
* The low 16 bits are the zlib compression parameters:
|
|
*/
|
|
#define pz_level_base (-1)
|
|
#define pz_level_max 9
|
|
#define pz_level_pos 0
|
|
#define pz_windowBits_base 8
|
|
#define pz_windowBits_max 15
|
|
#define pz_windowBits_pos 1
|
|
#define pz_memLevel_base 1
|
|
#define pz_memLevel_max 9
|
|
#define pz_memLevel_pos 2
|
|
#define pz_strategy_base 0
|
|
#define pz_strategy_max 4
|
|
#define pz_strategy_pos 3
|
|
#define pz_zlib_bits 0xFFFFU
|
|
/* Anything below this is not used directly by zlib: */
|
|
#define pz_png_level_base 0
|
|
#define pz_png_level_max 6
|
|
#define pz_png_level_pos 4
|
|
|
|
#define pz_offset(name) (pz_ ## name ## _base - 1)
|
|
/* setting_value == pz_offset(setting)+encoded_value */
|
|
#define pz_min(name) pz_ ## name ## _base
|
|
#define pz_max(name) pz_ ## name ## _max
|
|
#define pz_shift(name) (4 * pz_ ## name ## _pos)
|
|
|
|
#define pz_bits(name,x) ((int)(((x)>>pz_shift(name))&0xF))
|
|
/* the encoded value, or 0 if unset */
|
|
|
|
/* Enquiries: */
|
|
#define pz_isset(name,x) (pz_bits(name,x) != 0)
|
|
#define pz_value(name,x) (pz_bits(name,x)+pz_offset(name))
|
|
|
|
/* Assignments: */
|
|
#define pz_clear(name,x) ((x)&~((png_uint_32)0xFU<<pz_shift(name)))
|
|
#define pz_encode(name,v) ((png_uint_32)((v)-pz_offset(name))<<pz_shift(name))
|
|
#define pz_change(name,x,v) (pz_clear(name,x) | pz_encode(name, v))
|
|
|
|
/* Direct use/modification: */
|
|
#define pz_var(ps, type) ((ps)->pz_ ## type)
|
|
#define pz_get(ps, type, name, def)\
|
|
(pz_isset(name, pz_var(ps, type)) ? pz_value(name, pz_var(ps, type)) : (def))
|
|
/* pz_assign checks for out-of-range values and clears the setting if these are
|
|
* given. No warning or error is generated.
|
|
*/
|
|
#define pz_assign(ps, type, name, value)\
|
|
(pz_var(ps, type) = pz_clear(name, pz_var(ps, type)) |\
|
|
((value) >= pz_min(name) && (value) <= pz_max(name) ?\
|
|
pz_encode(name, value) : 0))
|
|
|
|
static png_int_32
|
|
pz_compression_setting(png_structrp png_ptr, png_uint_32 owner,
|
|
int min, int max, int shift, png_int_32 value, int only_get, int unset)
|
|
/* This is a support function for png_write_setting below. */
|
|
{
|
|
png_zlib_statep ps;
|
|
png_uint_32p psettings;
|
|
|
|
/* The value is only required for a 'set', eliminate out-of-range values
|
|
* first:
|
|
*/
|
|
if (!only_get && (value < min || value > max))
|
|
return PNG_EDOM;
|
|
|
|
/* If setting a value make sure the state exists: */
|
|
if (!only_get)
|
|
ps = get_zlib_state(png_ptr);
|
|
|
|
else if (owner != 0U) /* ps may be NULL */
|
|
ps = png_ptr->zlib_state;
|
|
|
|
else /* get and owner is 0U */
|
|
return 0; /* supported */
|
|
|
|
psettings = NULL;
|
|
switch (owner)
|
|
{
|
|
png_int_32 res;
|
|
|
|
case png_IDAT:
|
|
if (ps != NULL) psettings = &ps->pz_IDAT;
|
|
break;
|
|
|
|
case png_iCCP:
|
|
if (ps != NULL) psettings = &ps->pz_iCCP;
|
|
break;
|
|
|
|
case 0U:
|
|
/* All the settings. At this point the 'get' case has returned 0
|
|
* above, the value has been checked and the paramter is 0, therefore
|
|
* valid. Each of the following calls should succeed and it would be
|
|
* reasonable to eliminate the PNG_FAILED tests in a world where
|
|
* software engineers never made mistakes.
|
|
*/
|
|
res = pz_compression_setting(png_ptr, png_IDAT, min, max, shift,
|
|
value, 0/*set*/, 1/*iff unset*/);
|
|
|
|
if (PNG_FAILED(res))
|
|
return res;
|
|
|
|
res = pz_compression_setting(png_ptr, png_iCCP, min, max, shift,
|
|
value, 0/*set*/, 1/*iff unset*/);
|
|
|
|
if (PNG_FAILED(res))
|
|
return res;
|
|
|
|
/* The text settings are changed regardless of the customize support
|
|
* because if WRITE_CUSTOMIZE_ZTXT_COMPRESSION is not supported the old
|
|
* behavior was to use the WRITE_CUSTOMIZE_COMPRESSION setting.
|
|
*
|
|
* However, when we get png_zTXt directly (from png_write_setting) and
|
|
* the support is not compiled in return PNG_ENOSYS.
|
|
*/
|
|
unset = 1; /* i.e. only if not already set */
|
|
|
|
# ifdef PNG_WRITE_CUSTOMIZE_ZTXT_COMPRESSION_SUPPORTED
|
|
case png_zTXt:
|
|
case png_iTXt:
|
|
# endif /* WRITE_CUSTOMIZE_ZTXT_COMPRESSION */
|
|
if (ps != NULL) psettings = &ps->pz_text;
|
|
break;
|
|
|
|
default:
|
|
/* Return PNG_ENOSYS, not PNG_EINVAL, to support future addition of new
|
|
* compressed chunks and the fact that zTXt and iTXt customization can
|
|
* be disabled.
|
|
*/
|
|
return PNG_ENOSYS;
|
|
}
|
|
|
|
if (psettings == NULL)
|
|
return PNG_UNSET; /* valid setting that is not set */
|
|
|
|
{
|
|
png_uint_32 settings = *psettings;
|
|
png_uint_32 mask = 0xFU << shift;
|
|
|
|
/* Do not set it if 'only_get' was passed in or if 'unset' is true and the
|
|
* setting is not currently set:
|
|
*/
|
|
if (!only_get && ((settings & mask) == 0U || !unset))
|
|
*psettings = (settings & ~mask) +
|
|
((png_uint_32)/*SAFE*/(value-min+1) << shift);
|
|
|
|
settings &= mask;
|
|
|
|
if (settings == 0U)
|
|
return PNG_UNSET;
|
|
|
|
else
|
|
return (int)/*SAFE*/((settings >> shift)-1U) + min;
|
|
}
|
|
}
|
|
|
|
#define compression_setting(pp, owner, setting, value, get)\
|
|
pz_compression_setting(pp, owner, pz_min(setting), pz_max(setting),\
|
|
pz_shift(setting), value, get, 0/*always*/)
|
|
|
|
/* There is (as of zlib 1.2.8) a bug in the implementation of compression with a
|
|
* window size of 256 which zlib works round by resetting windowBits from 8 to 9
|
|
* whenever deflateInit2 is called with that value. Fix this up here.
|
|
*/
|
|
static void
|
|
fix_cinfo(png_zlib_statep ps, png_bytep data, png_alloc_size_t data_size)
|
|
{
|
|
/* Do this if the CINFO field is '1', meaning windowBits of 9. The first
|
|
* byte of the stream is the CMF value, CINFO is in the upper four bits.
|
|
*
|
|
* If zlib didn't futz with the value then it should match the value in
|
|
* pz_current; check this is debug. (See below for why this works in the
|
|
* pz_default_settings call.)
|
|
*/
|
|
# define png_ptr png_voidcast(png_const_structrp, ps->s.zs.opaque)
|
|
if (data[0] == 0x18U &&
|
|
pz_get(ps, current, windowBits, 0) == 8 /* i.e. it was requested */)
|
|
{
|
|
/* Double check this here; the fixup only works if the data was 256 bytes
|
|
* or shorter *or* the window is never used. For safety repeat the checks
|
|
* done in pz_default_settings; technically we should be able to just skip
|
|
* this test.
|
|
*
|
|
* TODO: set a 'fixup' flag in zlib_state to make this quicker?
|
|
*/
|
|
if (data_size <= 256U ||
|
|
pz_get(ps, current, strategy, Z_RLE) == Z_HUFFMAN_ONLY ||
|
|
pz_get(ps, current, level, 1) == Z_NO_COMPRESSION)
|
|
{
|
|
unsigned int d1;
|
|
|
|
data[0] = 0x08U;
|
|
/* The header checksum must be fixed too. The FCHECK (low 5 bits) make
|
|
* CMF.FLG a multiple of 31:
|
|
*/
|
|
d1 = data[1] & 0xE0U; /* top three bits */
|
|
d1 += 31U - (0x0800U + d1) % 31U;
|
|
data[1] = PNG_BYTE(d1);
|
|
}
|
|
|
|
else /* pz_default_settings is expected to guarantee the above */
|
|
NOT_REACHED;
|
|
}
|
|
|
|
else if (data_size > 0U)
|
|
{
|
|
/* Prior to 1.7.0 libpng would shrink the windowBits even if the
|
|
* application requested a particular value, so:
|
|
*/
|
|
unsigned int z_cinfo = data[0] >> 4;
|
|
unsigned int half_z_window_size = 1U << (z_cinfo + 7);
|
|
|
|
if (data_size <= half_z_window_size && z_cinfo > 0)
|
|
{
|
|
unsigned int tmp;
|
|
|
|
do
|
|
{
|
|
half_z_window_size >>= 1;
|
|
--z_cinfo;
|
|
}
|
|
while (z_cinfo > 0 && data_size <= half_z_window_size);
|
|
|
|
data[0] = PNG_BYTE((z_cinfo << 4) + 0x8U);
|
|
tmp = data[1] & 0xE0U; /* top three bits */
|
|
tmp += 31U - ((data[0] << 8) + tmp) % 31U;
|
|
data[1] = PNG_BYTE(tmp);
|
|
}
|
|
}
|
|
|
|
else
|
|
NOT_REACHED; /* invalid data size (0) */
|
|
# undef png_ptr
|
|
}
|
|
|
|
static png_uint_32
|
|
pz_default_settings(png_uint_32 settings, const png_uint_32 owner,
|
|
const png_alloc_size_t data_size, const unsigned int filters/*for IDAT*/)
|
|
{
|
|
int png_level, strategy, zlib_level, windowBits;
|
|
|
|
/* The png 'level' parameter controls the defaults below. It uses the same
|
|
* numbering scheme as the Zlib compression level except that -1 invokes the
|
|
* set of options and, in some cases, libpng behavior of libpng 1.6 and
|
|
* earlier.
|
|
*
|
|
* In the comments below reference is made to the differences beteen the
|
|
* legacy compression sizes from libpng 1.6 and earlier and the result of
|
|
* using the various options. These are quoted as an overall size change in
|
|
* the compression of 147323 PNG test files. The set of test files is
|
|
* slightly restricted because pre-1.7 versions of png_read_png leave random
|
|
* bits into the final byte of a row which ends with a partial byte. This
|
|
* affects the compression unpredictably so such files were omitted from the
|
|
* measurements.
|
|
*/
|
|
if (!pz_isset(png_level, settings))
|
|
{
|
|
png_level = PNG_DEFAULT_COMPRESSION_LEVEL;
|
|
settings |= pz_encode(png_level, png_level);
|
|
}
|
|
|
|
else
|
|
png_level = pz_value(png_level, settings);
|
|
|
|
/* First default the strategy. At lower data sizes other strategies do as
|
|
* well as the zlib default compression strategy but they never seem to
|
|
* improve on it with the 1.7 filtering.
|
|
*/
|
|
if (!pz_isset(strategy, settings))
|
|
{
|
|
switch (png_level)
|
|
{
|
|
case PNG_COMPRESSION_COMPAT: /* Legacy setting */
|
|
/* The pre-1.7 code used Z_FILTERED normally but uses
|
|
* Z_DEFAULT_STRATEGY for palette or low-bit-depth images.
|
|
*
|
|
* In fact Z_DEFAULT_STRATEGY works best for filtered images as
|
|
* well, however the change in results is small:
|
|
*
|
|
* Z_DEFAULT_STRATEGY: -0.1%
|
|
* Z_FILTERED: +0.1%
|
|
*
|
|
* NOTE: this happened even if WRITE_FILTER was *not* supported.
|
|
*/
|
|
if (owner != png_IDAT || filters == PNG_FILTER_NONE)
|
|
strategy = Z_DEFAULT_STRATEGY;
|
|
|
|
else
|
|
strategy = Z_FILTERED;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_HIGH_SPEED:
|
|
/* RLE is as fast as HUFFMAN_ONLY and can reduce size a lot in a few
|
|
* cases.
|
|
*/
|
|
strategy = Z_RLE;
|
|
break;
|
|
|
|
default: /* For GCC */
|
|
case PNG_COMPRESSION_LOW:
|
|
case PNG_COMPRESSION_MEDIUM:
|
|
/* Z_FILTERED is almost as good as the default and can be
|
|
* significantly faster. It biases the algorithm towards smaller
|
|
* byte values.
|
|
*
|
|
* Using Z_DEFAULT_STRATEGY here, rather than Z_FILTERED, benefits
|
|
* smaller 8 and 16-bit gray and larger 8 and 16-bit RGB images,
|
|
* however the overall gain is only 0.1% because it is offset by
|
|
* losses in larger 8-bit gray and alpha images. It is extremely
|
|
* difficult to deduce a pattern other than biases in the test set
|
|
* of images.
|
|
*
|
|
* Looking at the pattern of behavior with the 1.6 filter selection
|
|
* algorithm (none of palette or low-bit-depth, else all) produces
|
|
* results as follows:
|
|
*/
|
|
if (owner == png_IDAT)
|
|
{
|
|
if (filters == PNG_FILTER_NONE)
|
|
strategy = Z_DEFAULT_STRATEGY;
|
|
|
|
else
|
|
strategy = Z_FILTERED;
|
|
}
|
|
|
|
else if (owner == png_iCCP)
|
|
strategy = Z_DEFAULT_STRATEGY;
|
|
|
|
/* TODO: investigate this, the observed behavior is suspicious: */
|
|
else /* text chunk */
|
|
strategy = Z_FILTERED; /* Always better for some reason */
|
|
break;
|
|
|
|
case PNG_COMPRESSION_LOW_MEMORY:
|
|
/* Reduce memory at all costs, speed doesn't matter. */
|
|
case PNG_COMPRESSION_HIGH_READ_SPEED:
|
|
case PNG_COMPRESSION_HIGH:
|
|
if (owner == png_IDAT || owner == png_iCCP)
|
|
strategy = Z_DEFAULT_STRATEGY;
|
|
|
|
else
|
|
strategy = Z_FILTERED;
|
|
break;
|
|
}
|
|
|
|
settings |= pz_encode(strategy, strategy);
|
|
}
|
|
|
|
else
|
|
strategy = pz_value(strategy, settings);
|
|
|
|
/* Next the zlib level; this just defaults to the png level, except that for
|
|
* Huffman or RLE encoding the level setting for Zlib doesn't matter.
|
|
*/
|
|
if (!pz_isset(level, settings))
|
|
{
|
|
switch (strategy)
|
|
{
|
|
case Z_HUFFMAN_ONLY:
|
|
case Z_RLE:
|
|
/* The 'level' doesn't make any significant difference to the
|
|
* compression with these strategies; in a test set of about 3GByte
|
|
* of PNG files the total compressed size changed under 20 bytes
|
|
* with libpng 1.6!
|
|
*/
|
|
zlib_level = 1;
|
|
break;
|
|
|
|
default: /* Z_FIXED, Z_FILTERED, Z_DEFAULT_STRATEGY */
|
|
/* Everything that uses the window seems to show rapidly diminishing
|
|
* returns above level 6 (at least with libpng 1.6).
|
|
* Z_DEFAULT_COMPRESSION is, in fact, level 6 so Mark seems to
|
|
* concur. With libpng 1.6 the following results were obtained
|
|
* using the full test set of files (including those with a partial
|
|
* byte at the end of the row) and just varying the zlib level:
|
|
*
|
|
* LEVEL SIZE(bytes) CHANGE TIME(s) CHANGE METRIC
|
|
* 9 2550246600 -1.19% 1972 +227% -77%
|
|
* 8 2556675866 -0.94% 1215 +101% -59%
|
|
* 7 2572685552 -0.32% 679 +12% -15%
|
|
* 6 2581196708 0% 604 0% 0%
|
|
* 5 2602831249 +0.84% 414 -30% +87%
|
|
* 4 2625206800 +1.71% 358 -40% +153%
|
|
* 3 2674752349 +3.62% 298 -50% +303%
|
|
* 2 2716261483 +5.23% 262 -56% +537%
|
|
* 1 2749875805 +6.53% 251 -57% +662%
|
|
* 0 7174488347 202 -66%
|
|
*
|
|
* The CHANGE columns express the change in compressed size
|
|
* (positive is an increase; a decrease in compression) and time
|
|
* (positive is an increase; an increase in time) relative to level
|
|
* 6. The METRIC column is a measure of the compression-per-second
|
|
* relative to level 6; positive is an increase in
|
|
* compression-per-second.
|
|
*
|
|
* The metric is derived by assuming the difference in time between
|
|
* level 0 (which does no compression) and the level being
|
|
* considered is spent doing the compression. (Reasonable, since
|
|
* only the level changed). Just the inverse of the product of the
|
|
* size and the time difference is a measure of compression per
|
|
* second. It can be seen that time dominates the metric;
|
|
* compression only varies slightly (under 8%) across the level
|
|
* range.
|
|
*/
|
|
switch (png_level)
|
|
{
|
|
case PNG_COMPRESSION_COMPAT:
|
|
zlib_level = Z_DEFAULT_COMPRESSION; /* NOTE: -1 */
|
|
break;
|
|
|
|
case PNG_COMPRESSION_HIGH_SPEED:
|
|
zlib_level = 1;
|
|
break;
|
|
|
|
default: /* For GCC */
|
|
case PNG_COMPRESSION_LOW:
|
|
zlib_level = 3;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_MEDIUM:
|
|
zlib_level = 6; /* Old default! */
|
|
break;
|
|
|
|
case PNG_COMPRESSION_LOW_MEMORY:
|
|
case PNG_COMPRESSION_HIGH_READ_SPEED:
|
|
case PNG_COMPRESSION_HIGH:
|
|
zlib_level = 9;
|
|
break;
|
|
}
|
|
break;
|
|
}
|
|
|
|
settings |= pz_encode(level, zlib_level);
|
|
}
|
|
|
|
else
|
|
zlib_level = pz_value(level, settings);
|
|
|
|
/* Now default windowBits. This is probably the most important of the
|
|
* settings because it is pretty much the only one that affects decode
|
|
* performance. The smaller the better:
|
|
*/
|
|
if (!pz_isset(windowBits, settings))
|
|
{
|
|
if (png_level == PNG_COMPRESSION_COMPAT/* Legacy */)
|
|
{
|
|
/* This is the libpng16 calculation (it is wrong; a misunderstanding of
|
|
* what zlib actually requires!)
|
|
*
|
|
* Using the code below with the legacy choice of Z_FILTERED or
|
|
* Z_DEFAULT_STRATEGY increases the size of the test files by only
|
|
* 0.04%, however the settings below considerably reduce the windowBits
|
|
* used potentially benefitting read code a lot.
|
|
*
|
|
* NOTE: the algorithm below was determined by experiment and
|
|
* observation with the same set of test files; there is some
|
|
* considerable possibility that a different set might show different
|
|
* results. Obtaining large, representative, test sets is both a
|
|
* considerable amount of work and very error prone. [JB 20160518]
|
|
*/
|
|
windowBits = 15;
|
|
|
|
if (data_size <= 16384U)
|
|
{
|
|
unsigned int half_window_size = 1U << (windowBits-1);
|
|
|
|
while (data_size + 262U <= half_window_size)
|
|
{
|
|
half_window_size >>= 1;
|
|
--windowBits;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* The window size affects the memory used on both read and write but also
|
|
* the time on write (but not normally read). Handle the low memory
|
|
* requirement first:
|
|
*/
|
|
else if (zlib_level == Z_NO_COMPRESSION ||
|
|
png_level == PNG_COMPRESSION_LOW_MEMORY)
|
|
windowBits = 8;
|
|
|
|
/* If the strategy has been set to something that doesn't benefit from
|
|
* higher windowBits values take advantage of this. Note that pz_value
|
|
* returns an invalid value if pz_isset is false.
|
|
*
|
|
* The only png_level that affects this decision is HIGH_SPEED, because
|
|
* a smaller windowBits should speed up the search, however the code above
|
|
* chose zlib_level based on this so ignore that consideration and just
|
|
* use zlib_level below.
|
|
*/
|
|
else switch (strategy)
|
|
{
|
|
png_alloc_size_t test_size;
|
|
|
|
case Z_HUFFMAN_ONLY:
|
|
/* Use the minimum; the window doesn't get used */
|
|
windowBits = 8;
|
|
break;
|
|
|
|
case Z_RLE:
|
|
/* The longest length code is 258 bytes, the shortest string that
|
|
* can achieve this is 259 bytes long; 259 copies of the same byte
|
|
* which can be encoded as a code for the byte value then a string
|
|
* of length 258 starting at the first byte. So if the data is
|
|
* longer than 256 bytes use '9' for the windowBits, otherwise use
|
|
* 8:
|
|
*/
|
|
if (data_size <= 256U)
|
|
windowBits = 8;
|
|
|
|
else
|
|
windowBits = 9;
|
|
break;
|
|
|
|
/* By experiment using about 150,000 files the optimal windowBits
|
|
* value across a range of files is somewhat less than implied by
|
|
* the data size and depends on the zlib level and the strategy
|
|
* used, the following values were determined by experiment using
|
|
* those files:
|
|
*/
|
|
case Z_FILTERED:
|
|
/* The Z_FILTERED case changes suddenly at (zlib) level 4 to
|
|
* benefit from looking at all the data:
|
|
*/
|
|
if (zlib_level < 4 && zlib_level != Z_DEFAULT_COMPRESSION/*-1: 6*/)
|
|
test_size = data_size / 8U;
|
|
|
|
else
|
|
test_size = data_size;
|
|
|
|
goto check_test_size;
|
|
|
|
case Z_FIXED:
|
|
/* With the fixed Huffman tables better compression only ever comes
|
|
* from looking for matches, so, logically:
|
|
*/
|
|
test_size = data_size;
|
|
goto check_test_size;
|
|
|
|
default:
|
|
/* The default algorithm always does better with a window smaller
|
|
* than all the data and shows jumps at level 4 and level 8. The
|
|
* net effect with the test set of images is a very minor overall
|
|
* improvement compared to the pre-1.7 calculation (data size +
|
|
* 262). The benefit is less than 0.01%, however smaller window
|
|
* sizes reduce the memory zlib has to allocate in the decoder.
|
|
*/
|
|
switch (zlib_level)
|
|
{
|
|
case 1: case 2: case 3:
|
|
test_size = data_size / 8U;
|
|
break;
|
|
|
|
default: /* -1(Z_DEFAULT_COMPRESSION) == 6, 4..7 */
|
|
/* This includes, implicitly, ZLIB_NO_COMPRESSION, but that
|
|
* was eliminated in the 'if' above.
|
|
*/
|
|
test_size = data_size / 4U;
|
|
break;
|
|
|
|
case 8: case 9:
|
|
test_size = data_size / 3U;
|
|
break;
|
|
}
|
|
|
|
goto check_test_size;
|
|
|
|
check_test_size:
|
|
/* Find the smallest window that covers 'test_size' bytes, subject
|
|
* to the constraint that if the actual data size is more than 256
|
|
* bytes the minimum windowBits that can be supported is 9:
|
|
*/
|
|
if (data_size <= 256U)
|
|
windowBits = 8;
|
|
|
|
else
|
|
windowBits = 9;
|
|
|
|
while (windowBits < 15 && (1U << windowBits) < test_size)
|
|
++windowBits;
|
|
|
|
break;
|
|
}
|
|
|
|
settings |= pz_encode(windowBits, windowBits);
|
|
}
|
|
|
|
else
|
|
windowBits = pz_value(windowBits, settings);
|
|
|
|
/* zlib has a problem with 256 byte windows; 512 is used instead.
|
|
* We can't work round this if the data size is more than 256 bytes and
|
|
* the strategy actually uses the window (everything except huffman-only)
|
|
* so fix the problem here.
|
|
*/
|
|
if (windowBits == 8 && data_size > 256U && strategy != Z_HUFFMAN_ONLY &&
|
|
zlib_level != Z_NO_COMPRESSION)
|
|
settings = pz_change(windowBits, settings, 9);
|
|
|
|
/* For memLevel this just increases the memory used but can help with the
|
|
* Huffman code generation even to level 9 (the maximum), so just set the
|
|
* max. This affects memory used, not (apparently) compression speed so
|
|
* the only relevant png_level is LOW_MEMORY.
|
|
*
|
|
* The legacy setting is '8'; this is the level that Zlib defaults to because
|
|
* 16-bit iAPX86 systems could not handle '9'. Because MAX_MEM_LEVEL is used
|
|
* below this does not matter; zconf.h selects 8 or 9 as appropriate.
|
|
*
|
|
* In fact using '9' with the legacy settings increases the size of the test
|
|
* set minutely; +0.007%. This is hardly significant; 0.007% of the test
|
|
* images equals 10 images. (Nevertheless it is interesting, just as the
|
|
* observation that decreasing windowBits can result in smaller compressed
|
|
* sizes is interesting.)
|
|
*/
|
|
if (!pz_isset(memLevel, settings))
|
|
{
|
|
int memLevel;
|
|
|
|
switch (png_level)
|
|
{
|
|
case PNG_COMPRESSION_COMPAT:
|
|
memLevel = 8;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_LOW_MEMORY:
|
|
memLevel = 1;
|
|
break;
|
|
|
|
default:
|
|
memLevel = MAX_MEM_LEVEL/*from zconf.h*/;
|
|
break;
|
|
}
|
|
|
|
settings |= pz_encode(memLevel, memLevel);
|
|
}
|
|
|
|
return settings;
|
|
}
|
|
|
|
/* This is used below to find the size of an image to pass to png_deflate_claim.
|
|
* It returns 0 for images whose size would overflow a 32-bit integer or have
|
|
* rows which cannot be allocated.
|
|
*/
|
|
static png_alloc_size_t
|
|
png_image_size(png_const_structrp png_ptr)
|
|
{
|
|
/* The size returned here is limited to PNG_SIZE_MAX, if the size would
|
|
* exceed that (or is close to exceeding that) 0 is returned. See below for
|
|
* a variant that limits the size of 0xFFFFFFFFU.
|
|
*/
|
|
const png_alloc_size_t rowbytes = png_ptr->zlib_state->write_row_size;
|
|
|
|
/* NON-INTERLACED: (1+rowbytes) * h
|
|
* INTERLACED: Each pixel is transmitted exactly once, so the size is
|
|
* (rowbytes * h) + the count of filter bytes. Each complete
|
|
* block of 8 image rows generates at most 15 output rows
|
|
* (less for narrow images), so the filter byte count is
|
|
* at most (15*h/8)+14. Because the original rows are split
|
|
* extra byte passing may be introduced. Account for this by
|
|
* allowing an extra 1 byte per output row; that's two bytes
|
|
* including the filer byte.
|
|
*
|
|
* So:
|
|
* NON-INTERLACED: (rowbytes * h) + h
|
|
* INTERLACED: < (rowbytes * h) + 2*(15 * h/8) + 2*15
|
|
*
|
|
* Hence:
|
|
*/
|
|
if (rowbytes != 0)
|
|
{
|
|
const png_uint_32 h = png_ptr->height;
|
|
|
|
if (png_ptr->interlaced == PNG_INTERLACE_NONE)
|
|
{
|
|
const png_alloc_size_t limit = PNG_SIZE_MAX / h;
|
|
|
|
/* On 16-bit systems the above might be 0, so: */
|
|
if (rowbytes </*allow 1 for filter byte*/ limit)
|
|
return (rowbytes+1U) * h;
|
|
}
|
|
|
|
else /* INTERLACED */
|
|
{
|
|
const png_uint_32 w = png_ptr->width;
|
|
|
|
/* Interlacing makes the image larger because of the replication of
|
|
* both the filter byte and the padding to a byte boundary.
|
|
*/
|
|
png_alloc_size_t cb_base;
|
|
int pass;
|
|
|
|
for (cb_base=0, pass=0; pass<PNG_INTERLACE_ADAM7_PASSES; ++pass)
|
|
{
|
|
const png_uint_32 pass_w = PNG_PASS_COLS(w, pass);
|
|
|
|
if (pass_w > 0)
|
|
{
|
|
const png_uint_32 pass_h = PNG_PASS_ROWS(h, pass);
|
|
|
|
if (pass_h > 0)
|
|
{
|
|
/* This is the number of bytes available for each row of this
|
|
* pass:
|
|
*/
|
|
const png_alloc_size_t limit = (PNG_SIZE_MAX - cb_base)/pass_h;
|
|
/* This cannot overflow because if it did rowbytes would
|
|
* have been 0 above.
|
|
*/
|
|
const png_alloc_size_t pass_bytes =
|
|
PNG_ROWBYTES(png_ptr->row_output_pixel_depth, pass_w);
|
|
|
|
if (pass_bytes </*allow 1 for filter byte*/ limit)
|
|
cb_base += (pass_bytes+1U) * pass_h;
|
|
|
|
else
|
|
return 0U; /* insufficient address space left */
|
|
}
|
|
}
|
|
}
|
|
|
|
return cb_base;
|
|
}
|
|
}
|
|
|
|
/* Failure case: */
|
|
return 0U;
|
|
}
|
|
|
|
/* Initialize the compressor for the appropriate type of compression. */
|
|
static png_zlib_statep
|
|
png_deflate_claim(png_structrp png_ptr, png_uint_32 owner,
|
|
png_alloc_size_t data_size)
|
|
{
|
|
png_zlib_statep ps = get_zlib_state(png_ptr);
|
|
|
|
affirm(png_ptr->zowner == 0);
|
|
|
|
{
|
|
int ret; /* zlib return code */
|
|
unsigned int filters = 0U;
|
|
png_uint_32 settings;
|
|
|
|
switch (owner)
|
|
{
|
|
case png_IDAT:
|
|
debug(data_size == 0U);
|
|
data_size = png_image_size(png_ptr);
|
|
|
|
if (data_size == 0U)
|
|
data_size = PNG_SIZE_MAX;
|
|
|
|
settings = ps->pz_IDAT;
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
filters = ps->filter_mask;
|
|
debug(filters != 0U);
|
|
# else /* !WRITE_FILTER */
|
|
filters = PNG_FILTER_NONE;
|
|
# endif /* !WRITE_FILTER */
|
|
break;
|
|
|
|
case png_iCCP:
|
|
settings = ps->pz_iCCP;
|
|
break;
|
|
|
|
default: /* text chunk */
|
|
settings = ps->pz_text;
|
|
break;
|
|
}
|
|
|
|
settings = pz_default_settings(settings, owner, data_size, filters);
|
|
|
|
/* Check against the previous initialized values, if any. The relevant
|
|
* settings are in the low 16 bits.
|
|
*/
|
|
if (ps->s.zs.state != NULL &&
|
|
((settings ^ ps->pz_current) & pz_zlib_bits) != 0U)
|
|
png_deflateEnd(png_ptr, &ps->s.zs, 0/*check*/);
|
|
|
|
/* For safety clear out the input and output pointers (currently zlib
|
|
* doesn't use them on Init, but it might in the future).
|
|
*/
|
|
ps->s.zs.next_in = NULL;
|
|
ps->s.zs.avail_in = 0;
|
|
ps->s.zs.next_out = NULL;
|
|
ps->s.zs.avail_out = 0;
|
|
|
|
/* The length fields must be cleared too and the lists reset: */
|
|
ps->s.overflow = ps->s.len = ps->s.start = 0U;
|
|
|
|
if (ps->s.list != NULL) /* error in prior chunk writing */
|
|
{
|
|
debug(ps->stash == NULL);
|
|
ps->stash = ps->s.list;
|
|
ps->s.list = NULL;
|
|
}
|
|
|
|
ps->s.end = &ps->s.list;
|
|
|
|
/* Now initialize if required, setting the new parameters, otherwise just
|
|
* do a simple reset to the previous parameters.
|
|
*/
|
|
if (ps->s.zs.state != NULL)
|
|
ret = deflateReset(&ps->s.zs);
|
|
|
|
else
|
|
ret = deflateInit2(&ps->s.zs, pz_value(level, settings), Z_DEFLATED,
|
|
pz_value(windowBits, settings), pz_value(memLevel, settings),
|
|
pz_value(strategy, settings));
|
|
|
|
ps->pz_current = settings;
|
|
|
|
/* The return code is from either deflateReset or deflateInit2; they have
|
|
* pretty much the same set of error codes.
|
|
*/
|
|
if (ret == Z_OK && ps->s.zs.state != NULL)
|
|
png_ptr->zowner = owner;
|
|
|
|
else
|
|
{
|
|
png_zstream_error(&ps->s.zs, ret);
|
|
png_error(png_ptr, ps->s.zs.msg);
|
|
}
|
|
}
|
|
|
|
return ps;
|
|
}
|
|
|
|
#ifdef PNG_WRITE_COMPRESSED_TEXT_SUPPORTED /* includes iCCP */
|
|
/* Compress the block of data at the end of a chunk. This claims and releases
|
|
* png_struct::z_stream. It returns the amount of data in the chunk list or
|
|
* zero on error (a zlib stream always contains some bytes!)
|
|
*
|
|
* prefix_len is the amount of (uncompressed) data before the start of the
|
|
* compressed data. The routine will return 0 if the total of the compressed
|
|
* data and the prefix exceeds PNG_UINT_MAX_31.
|
|
*
|
|
* NOTE: this function may not return; it only returns 0 if
|
|
* png_chunk_report(PNG_CHUNK_WRITE_ERROR) returns (not the default).
|
|
*/
|
|
static int /* success */
|
|
png_compress_chunk_data(png_structrp png_ptr, png_uint_32 chunk_name,
|
|
png_uint_32 prefix_len, png_const_voidp input, png_alloc_size_t input_len)
|
|
{
|
|
/* To find the length of the output it is necessary to first compress the
|
|
* input. The result is buffered rather than using the two-pass algorithm
|
|
* that is used on the inflate side; deflate is assumed to be slower and a
|
|
* PNG writer is assumed to have more memory available than a PNG reader.
|
|
*
|
|
* IMPLEMENTATION NOTE: the zlib API deflateBound() can be used to find an
|
|
* upper limit on the output size, but it is always bigger than the input
|
|
* size so it is likely to be more efficient to use this linked-list
|
|
* approach.
|
|
*/
|
|
png_zlib_statep ps = png_deflate_claim(png_ptr, chunk_name, input_len);
|
|
|
|
affirm(ps != NULL);
|
|
|
|
/* The data compression function always returns so that we can clean up. */
|
|
ps->s.zs.next_in = PNGZ_INPUT_CAST(png_voidcast(const Bytef*, input));
|
|
|
|
/* Use the stash, if available: */
|
|
debug(ps->s.list == NULL);
|
|
ps->s.list = ps->stash;
|
|
ps->stash = NULL;
|
|
|
|
{
|
|
int ret = png_compress(&ps->s, input_len, PNG_UINT_31_MAX-prefix_len,
|
|
Z_FINISH);
|
|
|
|
ps->s.zs.next_out = NULL; /* safety */
|
|
ps->s.zs.avail_out = 0;
|
|
ps->s.zs.next_in = NULL;
|
|
ps->s.zs.avail_in = 0;
|
|
png_ptr->zowner = 0; /* release png_ptr::zstream */
|
|
|
|
/* Since Z_FINISH was passed as the flush parameter any result other than
|
|
* Z_STREAM_END is an error. In any case in the event of an error free
|
|
* the whole compression state; the only expected error is Z_MEM_ERROR.
|
|
*/
|
|
if (ret != Z_STREAM_END)
|
|
{
|
|
png_zlib_compress_destroy(&ps->s, 0/*check*/);
|
|
|
|
/* This is not very likely given the PNG_UINT_31_MAX limit above, but
|
|
* if code is added to limit the size of the chunks produced it can
|
|
* start to happen.
|
|
*/
|
|
if (ret == Z_BUF_ERROR)
|
|
ps->s.zs.msg = PNGZ_MSG_CAST("compressed chunk too long");
|
|
|
|
else
|
|
png_zstream_error(&ps->s.zs, ret);
|
|
|
|
png_chunk_report(png_ptr, ps->s.zs.msg, PNG_CHUNK_WRITE_ERROR);
|
|
return 0;
|
|
}
|
|
}
|
|
|
|
/* png_compress is meant to guarantee this on a successful return: */
|
|
affirm(ps->s.overflow == 0U && ps->s.len <= PNG_UINT_31_MAX - prefix_len);
|
|
|
|
/* Correct the zlib CINFO field: */
|
|
if (ps->s.len >= 2U)
|
|
fix_cinfo(ps, ps->s.list->output, input_len);
|
|
|
|
return 1;
|
|
}
|
|
|
|
/* Return the length of the compressed data; this is effectively a debug
|
|
* function to catch inconsistencies caused by internal errors. It will
|
|
* disappear in a release build.
|
|
*/
|
|
#if PNG_RELEASE_BUILD
|
|
# define png_length_compressed_chunk_data(pp, p) ((pp)->zlib_state->s.len)
|
|
#else /* !RELEASE_BUILD */
|
|
static png_uint_32
|
|
png_length_compressed_chunk_data(png_structrp png_ptr, png_uint_32 p)
|
|
{
|
|
png_zlib_statep ps = png_ptr->zlib_state;
|
|
|
|
debug(ps != NULL && ps->s.overflow == 0U && ps->s.len <= PNG_UINT_31_MAX-p);
|
|
return ps->s.len;
|
|
}
|
|
#endif /* !RELEASE_BUILD */
|
|
|
|
/* Write all the data produced by the above function; the caller must write the
|
|
* prefix and chunk header.
|
|
*/
|
|
static void
|
|
png_write_compressed_chunk_data(png_structrp png_ptr)
|
|
{
|
|
png_zlib_statep ps = png_ptr->zlib_state;
|
|
png_compression_bufferp next;
|
|
png_uint_32 output_len;
|
|
|
|
affirm(ps != NULL && ps->s.overflow == 0U);
|
|
next = ps->s.list;
|
|
|
|
for (output_len = ps->s.len; output_len > 0U; next = next->next)
|
|
{
|
|
png_uint_32 size = PNG_ROW_BUFFER_SIZE;
|
|
|
|
/* If this affirm fails there is a bug in the calculation of
|
|
* output_length above, or in the buffer_limit code in png_compress.
|
|
*/
|
|
affirm(next != NULL && output_len > 0U);
|
|
|
|
if (size > output_len)
|
|
size = output_len;
|
|
|
|
png_write_chunk_data(png_ptr, next->output, size);
|
|
|
|
output_len -= size;
|
|
}
|
|
|
|
/* Release the list back to the stash. */
|
|
debug(ps->stash == NULL);
|
|
ps->stash = ps->s.list;
|
|
ps->s.list = NULL;
|
|
ps->s.end = &ps->s.list;
|
|
}
|
|
#endif /* WRITE_COMPRESSED_TEXT */
|
|
|
|
#if defined(PNG_WRITE_TEXT_SUPPORTED) || defined(PNG_WRITE_pCAL_SUPPORTED) || \
|
|
defined(PNG_WRITE_iCCP_SUPPORTED) || defined(PNG_WRITE_sPLT_SUPPORTED)
|
|
/* Check that the tEXt or zTXt keyword is valid per PNG 1.0 specification,
|
|
* and if invalid, correct the keyword rather than discarding the entire
|
|
* chunk. The PNG 1.0 specification requires keywords 1-79 characters in
|
|
* length, forbids leading or trailing whitespace, multiple internal spaces,
|
|
* and the non-break space (0x80) from ISO 8859-1. Returns keyword length.
|
|
*
|
|
* The 'new_key' buffer must be at least 80 characters in size (for the keyword
|
|
* plus a trailing '\0'). If this routine returns 0 then there was no keyword,
|
|
* or a valid one could not be generated, and the caller must CHUNK_WRITE_ERROR.
|
|
*/
|
|
static unsigned int
|
|
png_check_keyword(png_structrp png_ptr, png_const_charp key, png_bytep new_key)
|
|
{
|
|
png_const_charp orig_key = key;
|
|
unsigned int key_len = 0;
|
|
int bad_character = 0;
|
|
int space = 1;
|
|
|
|
png_debug(1, "in png_check_keyword");
|
|
|
|
if (key == NULL)
|
|
{
|
|
*new_key = 0;
|
|
return 0;
|
|
}
|
|
|
|
while (*key && key_len < 79)
|
|
{
|
|
png_byte ch = (png_byte)(0xff & *key++);
|
|
|
|
if ((ch > 32 && ch <= 126) || (ch >= 161 /*&& ch <= 255*/))
|
|
*new_key++ = ch, ++key_len, space = 0;
|
|
|
|
else if (space == 0)
|
|
{
|
|
/* A space or an invalid character when one wasn't seen immediately
|
|
* before; output just a space.
|
|
*/
|
|
*new_key++ = 32, ++key_len, space = 1;
|
|
|
|
/* If the character was not a space then it is invalid. */
|
|
if (ch != 32)
|
|
bad_character = ch;
|
|
}
|
|
|
|
else if (bad_character == 0)
|
|
bad_character = ch; /* just skip it, record the first error */
|
|
}
|
|
|
|
if (key_len > 0 && space != 0) /* trailing space */
|
|
{
|
|
--key_len, --new_key;
|
|
if (bad_character == 0)
|
|
bad_character = 32;
|
|
}
|
|
|
|
/* Terminate the keyword */
|
|
*new_key = 0;
|
|
|
|
if (key_len == 0)
|
|
return 0;
|
|
|
|
#ifdef PNG_WARNINGS_SUPPORTED
|
|
/* Try to only output one warning per keyword: */
|
|
if (*key != 0) /* keyword too long */
|
|
png_app_warning(png_ptr, "keyword truncated");
|
|
|
|
else if (bad_character != 0)
|
|
{
|
|
PNG_WARNING_PARAMETERS(p)
|
|
|
|
png_warning_parameter(p, 1, orig_key);
|
|
png_warning_parameter_signed(p, 2, PNG_NUMBER_FORMAT_02x, bad_character);
|
|
|
|
png_formatted_warning(png_ptr, p, "keyword \"@1\": bad character '0x@2'");
|
|
}
|
|
#endif /* WARNINGS */
|
|
|
|
return key_len;
|
|
}
|
|
#endif /* WRITE_TEXT || WRITE_pCAL || WRITE_iCCP || WRITE_sPLT */
|
|
|
|
/* Write the IHDR chunk, and update the png_struct with the necessary
|
|
* information. Note that the rest of this code depends upon this
|
|
* information being correct.
|
|
*/
|
|
void /* PRIVATE */
|
|
png_write_IHDR(png_structrp png_ptr, png_uint_32 width, png_uint_32 height,
|
|
int bit_depth, int color_type, int compression_type, int filter_method,
|
|
int interlace_type)
|
|
{
|
|
png_byte buf[13]; /* Buffer to store the IHDR info */
|
|
|
|
png_debug(1, "in png_write_IHDR");
|
|
|
|
/* Check that we have valid input data from the application info */
|
|
switch (color_type)
|
|
{
|
|
case PNG_COLOR_TYPE_GRAY:
|
|
switch (bit_depth)
|
|
{
|
|
case 1:
|
|
case 2:
|
|
case 4:
|
|
case 8:
|
|
#ifdef PNG_WRITE_16BIT_SUPPORTED
|
|
case 16:
|
|
#endif
|
|
break;
|
|
|
|
default:
|
|
png_error(png_ptr, "Invalid bit depth for grayscale image");
|
|
}
|
|
break;
|
|
|
|
case PNG_COLOR_TYPE_RGB:
|
|
#ifdef PNG_WRITE_16BIT_SUPPORTED
|
|
if (bit_depth != 8 && bit_depth != 16)
|
|
#else
|
|
if (bit_depth != 8)
|
|
#endif
|
|
png_error(png_ptr, "Invalid bit depth for RGB image");
|
|
|
|
break;
|
|
|
|
case PNG_COLOR_TYPE_PALETTE:
|
|
switch (bit_depth)
|
|
{
|
|
case 1:
|
|
case 2:
|
|
case 4:
|
|
case 8:
|
|
break;
|
|
|
|
default:
|
|
png_error(png_ptr, "Invalid bit depth for paletted image");
|
|
}
|
|
break;
|
|
|
|
case PNG_COLOR_TYPE_GRAY_ALPHA:
|
|
if (bit_depth != 8 && bit_depth != 16)
|
|
png_error(png_ptr, "Invalid bit depth for grayscale+alpha image");
|
|
|
|
break;
|
|
|
|
case PNG_COLOR_TYPE_RGB_ALPHA:
|
|
#ifdef PNG_WRITE_16BIT_SUPPORTED
|
|
if (bit_depth != 8 && bit_depth != 16)
|
|
#else
|
|
if (bit_depth != 8)
|
|
#endif
|
|
png_error(png_ptr, "Invalid bit depth for RGBA image");
|
|
|
|
break;
|
|
|
|
default:
|
|
png_error(png_ptr, "Invalid image color type specified");
|
|
}
|
|
|
|
if (compression_type != PNG_COMPRESSION_TYPE_BASE)
|
|
{
|
|
png_app_error(png_ptr, "Invalid compression type specified");
|
|
compression_type = PNG_COMPRESSION_TYPE_BASE;
|
|
}
|
|
|
|
/* Write filter_method 64 (intrapixel differencing) only if
|
|
* 1. Libpng was compiled with PNG_MNG_FEATURES_SUPPORTED and
|
|
* 2. Libpng did not write a PNG signature (this filter_method is only
|
|
* used in PNG datastreams that are embedded in MNG datastreams) and
|
|
* 3. The application called png_permit_mng_features with a mask that
|
|
* included PNG_FLAG_MNG_FILTER_64 and
|
|
* 4. The filter_method is 64 and
|
|
* 5. The color_type is RGB or RGBA
|
|
*/
|
|
if (
|
|
# ifdef PNG_MNG_FEATURES_SUPPORTED
|
|
!((png_ptr->mng_features_permitted & PNG_FLAG_MNG_FILTER_64) != 0 &&
|
|
((png_ptr->mode & PNG_HAVE_PNG_SIGNATURE) == 0) &&
|
|
(color_type == PNG_COLOR_TYPE_RGB ||
|
|
color_type == PNG_COLOR_TYPE_RGB_ALPHA) &&
|
|
(filter_method == PNG_INTRAPIXEL_DIFFERENCING)) &&
|
|
# endif /* MNG_FEATURES */
|
|
filter_method != PNG_FILTER_TYPE_BASE)
|
|
{
|
|
png_app_error(png_ptr, "Invalid filter type specified");
|
|
filter_method = PNG_FILTER_TYPE_BASE;
|
|
}
|
|
|
|
if (interlace_type != PNG_INTERLACE_NONE &&
|
|
interlace_type != PNG_INTERLACE_ADAM7)
|
|
{
|
|
png_app_error(png_ptr, "Invalid interlace type specified");
|
|
interlace_type = PNG_INTERLACE_ADAM7;
|
|
}
|
|
|
|
/* Save the relevant information */
|
|
png_ptr->bit_depth = png_check_byte(png_ptr, bit_depth);
|
|
png_ptr->color_type = png_check_byte(png_ptr, color_type);
|
|
png_ptr->interlaced = png_check_byte(png_ptr, interlace_type);
|
|
png_ptr->filter_method = png_check_byte(png_ptr, filter_method);
|
|
png_ptr->compression_type = png_check_byte(png_ptr, compression_type);
|
|
png_ptr->width = width;
|
|
png_ptr->height = height;
|
|
|
|
/* Pack the header information into the buffer */
|
|
png_save_uint_32(buf, width);
|
|
png_save_uint_32(buf + 4, height);
|
|
buf[8] = png_check_byte(png_ptr, bit_depth);
|
|
buf[9] = png_check_byte(png_ptr, color_type);
|
|
buf[10] = png_check_byte(png_ptr, compression_type);
|
|
buf[11] = png_check_byte(png_ptr, filter_method);
|
|
buf[12] = png_check_byte(png_ptr, interlace_type);
|
|
|
|
/* Write the chunk */
|
|
png_write_complete_chunk(png_ptr, png_IHDR, buf, (png_size_t)13);
|
|
png_ptr->mode |= PNG_HAVE_IHDR;
|
|
}
|
|
|
|
/* Write the palette. We are careful not to trust png_color to be in the
|
|
* correct order for PNG, so people can redefine it to any convenient
|
|
* structure.
|
|
*/
|
|
void /* PRIVATE */
|
|
png_write_PLTE(png_structrp png_ptr, png_const_colorp palette,
|
|
unsigned int num_pal)
|
|
{
|
|
png_uint_32 max_palette_length, i;
|
|
png_const_colorp pal_ptr;
|
|
png_byte buf[3];
|
|
|
|
png_debug(1, "in png_write_PLTE");
|
|
|
|
max_palette_length = (png_ptr->color_type == PNG_COLOR_TYPE_PALETTE) ?
|
|
(1 << png_ptr->bit_depth) : PNG_MAX_PALETTE_LENGTH;
|
|
|
|
if ((
|
|
# ifdef PNG_MNG_FEATURES_SUPPORTED
|
|
(png_ptr->mng_features_permitted & PNG_FLAG_MNG_EMPTY_PLTE) == 0 &&
|
|
# endif /* MNG_FEATURES */
|
|
num_pal == 0) || num_pal > max_palette_length)
|
|
{
|
|
if (png_ptr->color_type == PNG_COLOR_TYPE_PALETTE)
|
|
{
|
|
png_error(png_ptr, "Invalid number of colors in palette");
|
|
}
|
|
|
|
else
|
|
{
|
|
png_warning(png_ptr, "Invalid number of colors in palette");
|
|
return;
|
|
}
|
|
}
|
|
|
|
if ((png_ptr->color_type & PNG_COLOR_MASK_COLOR) == 0)
|
|
{
|
|
png_warning(png_ptr,
|
|
"Ignoring request to write a PLTE chunk in grayscale PNG");
|
|
|
|
return;
|
|
}
|
|
|
|
png_ptr->num_palette = png_check_bits(png_ptr, num_pal, 9);
|
|
png_debug1(3, "num_palette = %d", png_ptr->num_palette);
|
|
|
|
png_write_chunk_header(png_ptr, png_PLTE, num_pal * 3U);
|
|
|
|
for (i = 0, pal_ptr = palette; i < num_pal; i++, pal_ptr++)
|
|
{
|
|
buf[0] = pal_ptr->red;
|
|
buf[1] = pal_ptr->green;
|
|
buf[2] = pal_ptr->blue;
|
|
png_write_chunk_data(png_ptr, buf, 3U);
|
|
}
|
|
|
|
png_write_chunk_end(png_ptr);
|
|
png_ptr->mode |= PNG_HAVE_PLTE;
|
|
}
|
|
|
|
/* Write an IEND chunk */
|
|
void /* PRIVATE */
|
|
png_write_IEND(png_structrp png_ptr)
|
|
{
|
|
png_debug(1, "in png_write_IEND");
|
|
|
|
png_write_complete_chunk(png_ptr, png_IEND, NULL, (png_size_t)0);
|
|
png_ptr->mode |= PNG_HAVE_IEND;
|
|
}
|
|
|
|
#if defined(PNG_WRITE_gAMA_SUPPORTED) || defined(PNG_WRITE_cHRM_SUPPORTED)
|
|
static int
|
|
png_save_int_31(png_structrp png_ptr, png_bytep buf, png_int_32 i)
|
|
/* Save a signed value as a PNG unsigned value; the argument is required to
|
|
* be in the range 0..0x7FFFFFFFU. If not a *warning* is produced and false
|
|
* is returned. Because this is only called from png_write_cHRM_fixed and
|
|
* png_write_gAMA_fixed below this is safe (we don't need either chunk,
|
|
* particularly if the value is bogus.)
|
|
*
|
|
* The warning is png_app_error; it may return if the app tells it to but the
|
|
* app can have it error out. JB 20150821: I believe the checking in png.c
|
|
* actually makes this error impossible, but this is safe.
|
|
*/
|
|
{
|
|
#ifndef __COVERITY__
|
|
if (i >= 0 && i <= 0x7FFFFFFF)
|
|
#else
|
|
/* Supress bogus Coverity complaint */
|
|
if (i >= 0)
|
|
#endif
|
|
{
|
|
png_save_uint_32(buf, (png_uint_32)/*SAFE*/i);
|
|
return 1;
|
|
}
|
|
|
|
else
|
|
{
|
|
png_chunk_report(png_ptr, "negative value in cHRM or gAMA",
|
|
PNG_CHUNK_WRITE_ERROR);
|
|
return 0;
|
|
}
|
|
}
|
|
#endif /* WRITE_gAMA || WRITE_cHRM */
|
|
|
|
#ifdef PNG_WRITE_gAMA_SUPPORTED
|
|
/* Write a gAMA chunk */
|
|
void /* PRIVATE */
|
|
png_write_gAMA_fixed(png_structrp png_ptr, png_fixed_point file_gamma)
|
|
{
|
|
png_byte buf[4];
|
|
|
|
png_debug(1, "in png_write_gAMA");
|
|
|
|
/* file_gamma is saved in 1/100,000ths */
|
|
if (png_save_int_31(png_ptr, buf, file_gamma))
|
|
png_write_complete_chunk(png_ptr, png_gAMA, buf, (png_size_t)4);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_sRGB_SUPPORTED
|
|
/* Write a sRGB chunk */
|
|
void /* PRIVATE */
|
|
png_write_sRGB(png_structrp png_ptr, int srgb_intent)
|
|
{
|
|
png_byte buf[1];
|
|
|
|
png_debug(1, "in png_write_sRGB");
|
|
|
|
if (srgb_intent >= PNG_sRGB_INTENT_LAST)
|
|
png_chunk_report(png_ptr, "Invalid sRGB rendering intent specified",
|
|
PNG_CHUNK_WRITE_ERROR);
|
|
|
|
buf[0] = png_check_byte(png_ptr, srgb_intent);
|
|
png_write_complete_chunk(png_ptr, png_sRGB, buf, (png_size_t)1);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_iCCP_SUPPORTED
|
|
/* Write an iCCP chunk */
|
|
void /* PRIVATE */
|
|
png_write_iCCP(png_structrp png_ptr, png_const_charp name,
|
|
png_const_voidp profile)
|
|
{
|
|
png_uint_32 name_len;
|
|
png_uint_32 profile_len;
|
|
png_byte new_name[81]; /* 1 byte for the compression byte */
|
|
|
|
png_debug(1, "in png_write_iCCP");
|
|
|
|
affirm(profile != NULL);
|
|
|
|
profile_len = png_get_uint_32(profile);
|
|
name_len = png_check_keyword(png_ptr, name, new_name);
|
|
|
|
if (name_len == 0)
|
|
{
|
|
png_chunk_report(png_ptr, "iCCP: invalid keyword", PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
++name_len; /* trailing '\0' */
|
|
new_name[name_len++] = PNG_COMPRESSION_TYPE_BASE;
|
|
|
|
if (png_compress_chunk_data(png_ptr, png_iCCP, name_len, profile,
|
|
profile_len))
|
|
{
|
|
png_write_chunk_header(png_ptr, png_iCCP,
|
|
name_len+png_length_compressed_chunk_data(png_ptr, name_len));
|
|
png_write_chunk_data(png_ptr, new_name, name_len);
|
|
png_write_compressed_chunk_data(png_ptr);
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_sPLT_SUPPORTED
|
|
/* Write a sPLT chunk */
|
|
void /* PRIVATE */
|
|
png_write_sPLT(png_structrp png_ptr, png_const_sPLT_tp spalette)
|
|
{
|
|
png_uint_32 name_len;
|
|
png_byte new_name[80];
|
|
png_byte entrybuf[10];
|
|
png_size_t entry_size = (spalette->depth == 8 ? 6 : 10);
|
|
png_size_t palette_size = entry_size * spalette->nentries;
|
|
png_sPLT_entryp ep;
|
|
|
|
png_debug(1, "in png_write_sPLT");
|
|
|
|
name_len = png_check_keyword(png_ptr, spalette->name, new_name);
|
|
|
|
if (name_len == 0)
|
|
png_error(png_ptr, "sPLT: invalid keyword");
|
|
|
|
/* Make sure we include the NULL after the name */
|
|
png_write_chunk_header(png_ptr, png_sPLT,
|
|
(png_uint_32)(name_len + 2 + palette_size));
|
|
|
|
png_write_chunk_data(png_ptr, new_name, name_len + 1);
|
|
|
|
png_write_chunk_data(png_ptr, &spalette->depth, 1);
|
|
|
|
/* Loop through each palette entry, writing appropriately */
|
|
for (ep = spalette->entries; ep<spalette->entries + spalette->nentries; ep++)
|
|
{
|
|
if (spalette->depth == 8)
|
|
{
|
|
entrybuf[0] = png_check_byte(png_ptr, ep->red);
|
|
entrybuf[1] = png_check_byte(png_ptr, ep->green);
|
|
entrybuf[2] = png_check_byte(png_ptr, ep->blue);
|
|
entrybuf[3] = png_check_byte(png_ptr, ep->alpha);
|
|
png_save_uint_16(entrybuf + 4, ep->frequency);
|
|
}
|
|
|
|
else
|
|
{
|
|
png_save_uint_16(entrybuf + 0, ep->red);
|
|
png_save_uint_16(entrybuf + 2, ep->green);
|
|
png_save_uint_16(entrybuf + 4, ep->blue);
|
|
png_save_uint_16(entrybuf + 6, ep->alpha);
|
|
png_save_uint_16(entrybuf + 8, ep->frequency);
|
|
}
|
|
|
|
png_write_chunk_data(png_ptr, entrybuf, entry_size);
|
|
}
|
|
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_sBIT_SUPPORTED
|
|
/* Write the sBIT chunk */
|
|
void /* PRIVATE */
|
|
png_write_sBIT(png_structrp png_ptr, png_const_color_8p sbit, int color_type)
|
|
{
|
|
png_byte buf[4];
|
|
png_size_t size;
|
|
|
|
png_debug(1, "in png_write_sBIT");
|
|
|
|
/* Make sure we don't depend upon the order of PNG_COLOR_8 */
|
|
if ((color_type & PNG_COLOR_MASK_COLOR) != 0)
|
|
{
|
|
unsigned int maxbits;
|
|
|
|
maxbits = color_type==PNG_COLOR_TYPE_PALETTE ? 8 : png_ptr->bit_depth;
|
|
|
|
if (sbit->red == 0 || sbit->red > maxbits ||
|
|
sbit->green == 0 || sbit->green > maxbits ||
|
|
sbit->blue == 0 || sbit->blue > maxbits)
|
|
{
|
|
png_app_error(png_ptr, "Invalid sBIT depth specified");
|
|
return;
|
|
}
|
|
|
|
buf[0] = sbit->red;
|
|
buf[1] = sbit->green;
|
|
buf[2] = sbit->blue;
|
|
size = 3;
|
|
}
|
|
|
|
else
|
|
{
|
|
if (sbit->gray == 0 || sbit->gray > png_ptr->bit_depth)
|
|
{
|
|
png_app_error(png_ptr, "Invalid sBIT depth specified");
|
|
return;
|
|
}
|
|
|
|
buf[0] = sbit->gray;
|
|
size = 1;
|
|
}
|
|
|
|
if ((color_type & PNG_COLOR_MASK_ALPHA) != 0)
|
|
{
|
|
if (sbit->alpha == 0 || sbit->alpha > png_ptr->bit_depth)
|
|
{
|
|
png_app_error(png_ptr, "Invalid sBIT depth specified");
|
|
return;
|
|
}
|
|
|
|
buf[size++] = sbit->alpha;
|
|
}
|
|
|
|
png_write_complete_chunk(png_ptr, png_sBIT, buf, size);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_cHRM_SUPPORTED
|
|
/* Write the cHRM chunk */
|
|
void /* PRIVATE */
|
|
png_write_cHRM_fixed(png_structrp png_ptr, const png_xy *xy)
|
|
{
|
|
png_byte buf[32];
|
|
|
|
png_debug(1, "in png_write_cHRM");
|
|
|
|
/* Each value is saved in 1/100,000ths */
|
|
if (png_save_int_31(png_ptr, buf, xy->whitex) &&
|
|
png_save_int_31(png_ptr, buf + 4, xy->whitey) &&
|
|
png_save_int_31(png_ptr, buf + 8, xy->redx) &&
|
|
png_save_int_31(png_ptr, buf + 12, xy->redy) &&
|
|
png_save_int_31(png_ptr, buf + 16, xy->greenx) &&
|
|
png_save_int_31(png_ptr, buf + 20, xy->greeny) &&
|
|
png_save_int_31(png_ptr, buf + 24, xy->bluex) &&
|
|
png_save_int_31(png_ptr, buf + 28, xy->bluey))
|
|
png_write_complete_chunk(png_ptr, png_cHRM, buf, 32);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_tRNS_SUPPORTED
|
|
/* Write the tRNS chunk */
|
|
void /* PRIVATE */
|
|
png_write_tRNS(png_structrp png_ptr, png_const_bytep trans_alpha,
|
|
png_const_color_16p tran, int num_trans, int color_type)
|
|
{
|
|
png_byte buf[6];
|
|
|
|
png_debug(1, "in png_write_tRNS");
|
|
|
|
if (color_type == PNG_COLOR_TYPE_PALETTE)
|
|
{
|
|
affirm(num_trans > 0 && num_trans <= PNG_MAX_PALETTE_LENGTH);
|
|
{
|
|
# ifdef PNG_WRITE_INVERT_ALPHA_SUPPORTED
|
|
union
|
|
{
|
|
png_uint_32 u32[1];
|
|
png_byte b8[PNG_MAX_PALETTE_LENGTH];
|
|
} inverted_alpha;
|
|
|
|
/* Invert the alpha channel (in tRNS) if required */
|
|
if (png_ptr->write_invert_alpha)
|
|
{
|
|
int i;
|
|
|
|
memcpy(inverted_alpha.b8, trans_alpha, num_trans);
|
|
|
|
for (i=0; 4*i<num_trans; ++i)
|
|
inverted_alpha.u32[i] = ~inverted_alpha.u32[i];
|
|
|
|
trans_alpha = inverted_alpha.b8;
|
|
}
|
|
# endif /* WRITE_INVERT_ALPHA */
|
|
|
|
png_write_complete_chunk(png_ptr, png_tRNS, trans_alpha, num_trans);
|
|
}
|
|
}
|
|
|
|
else if (color_type == PNG_COLOR_TYPE_GRAY)
|
|
{
|
|
/* One 16 bit value */
|
|
affirm(tran->gray < (1 << png_ptr->bit_depth));
|
|
png_save_uint_16(buf, tran->gray);
|
|
png_write_complete_chunk(png_ptr, png_tRNS, buf, (png_size_t)2);
|
|
}
|
|
|
|
else if (color_type == PNG_COLOR_TYPE_RGB)
|
|
{
|
|
/* Three 16 bit values */
|
|
png_save_uint_16(buf, tran->red);
|
|
png_save_uint_16(buf + 2, tran->green);
|
|
png_save_uint_16(buf + 4, tran->blue);
|
|
affirm(png_ptr->bit_depth == 8 || (buf[0] | buf[2] | buf[4]) == 0);
|
|
png_write_complete_chunk(png_ptr, png_tRNS, buf, (png_size_t)6);
|
|
}
|
|
|
|
else /* Already checked in png_set_tRNS */
|
|
impossible("invalid tRNS");
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_bKGD_SUPPORTED
|
|
/* Write the background chunk */
|
|
void /* PRIVATE */
|
|
png_write_bKGD(png_structrp png_ptr, png_const_color_16p back, int color_type)
|
|
{
|
|
png_byte buf[6];
|
|
|
|
png_debug(1, "in png_write_bKGD");
|
|
|
|
if (color_type == PNG_COLOR_TYPE_PALETTE)
|
|
{
|
|
if (
|
|
# ifdef PNG_MNG_FEATURES_SUPPORTED
|
|
(png_ptr->num_palette != 0 ||
|
|
(png_ptr->mng_features_permitted & PNG_FLAG_MNG_EMPTY_PLTE) == 0) &&
|
|
# endif /* MNG_FEATURES */
|
|
back->index >= png_ptr->num_palette)
|
|
{
|
|
png_app_error(png_ptr, "Invalid background palette index");
|
|
return;
|
|
}
|
|
|
|
buf[0] = back->index;
|
|
png_write_complete_chunk(png_ptr, png_bKGD, buf, (png_size_t)1);
|
|
}
|
|
|
|
else if ((color_type & PNG_COLOR_MASK_COLOR) != 0)
|
|
{
|
|
png_save_uint_16(buf, back->red);
|
|
png_save_uint_16(buf + 2, back->green);
|
|
png_save_uint_16(buf + 4, back->blue);
|
|
#ifdef PNG_WRITE_16BIT_SUPPORTED
|
|
if (png_ptr->bit_depth == 8 && (buf[0] | buf[2] | buf[4]) != 0)
|
|
#else
|
|
if ((buf[0] | buf[2] | buf[4]) != 0)
|
|
#endif
|
|
{
|
|
png_app_error(png_ptr,
|
|
"Ignoring attempt to write 16-bit bKGD chunk when bit_depth is 8");
|
|
|
|
return;
|
|
}
|
|
|
|
png_write_complete_chunk(png_ptr, png_bKGD, buf, (png_size_t)6);
|
|
}
|
|
|
|
else
|
|
{
|
|
if (back->gray >= (1 << png_ptr->bit_depth))
|
|
{
|
|
png_app_error(png_ptr,
|
|
"Ignoring attempt to write bKGD chunk out-of-range for bit_depth");
|
|
|
|
return;
|
|
}
|
|
|
|
png_save_uint_16(buf, back->gray);
|
|
png_write_complete_chunk(png_ptr, png_bKGD, buf, (png_size_t)2);
|
|
}
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_hIST_SUPPORTED
|
|
/* Write the histogram */
|
|
void /* PRIVATE */
|
|
png_write_hIST(png_structrp png_ptr, png_const_uint_16p hist, int num_hist)
|
|
{
|
|
int i;
|
|
png_byte buf[3];
|
|
|
|
png_debug(1, "in png_write_hIST");
|
|
|
|
if (num_hist > (int)png_ptr->num_palette)
|
|
{
|
|
png_debug2(3, "num_hist = %d, num_palette = %d", num_hist,
|
|
png_ptr->num_palette);
|
|
|
|
png_warning(png_ptr, "Invalid number of histogram entries specified");
|
|
return;
|
|
}
|
|
|
|
png_write_chunk_header(png_ptr, png_hIST, (png_uint_32)(num_hist * 2));
|
|
|
|
for (i = 0; i < num_hist; i++)
|
|
{
|
|
png_save_uint_16(buf, hist[i]);
|
|
png_write_chunk_data(png_ptr, buf, (png_size_t)2);
|
|
}
|
|
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_tEXt_SUPPORTED
|
|
/* Write a tEXt chunk */
|
|
void /* PRIVATE */
|
|
png_write_tEXt(png_structrp png_ptr, png_const_charp key, png_const_charp text,
|
|
png_size_t text_len)
|
|
{
|
|
unsigned int key_len;
|
|
png_byte new_key[80];
|
|
|
|
png_debug(1, "in png_write_tEXt");
|
|
|
|
key_len = png_check_keyword(png_ptr, key, new_key);
|
|
|
|
if (key_len == 0)
|
|
{
|
|
png_chunk_report(png_ptr, "tEXt: invalid keyword", PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
if (text == NULL || *text == '\0')
|
|
text_len = 0;
|
|
|
|
else
|
|
text_len = strlen(text);
|
|
|
|
if (text_len > PNG_UINT_31_MAX - (key_len+1))
|
|
{
|
|
png_chunk_report(png_ptr, "tEXt: text too long", PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
/* Make sure we include the 0 after the key */
|
|
png_write_chunk_header(png_ptr, png_tEXt,
|
|
(png_uint_32)/*checked above*/(key_len + text_len + 1));
|
|
/*
|
|
* We leave it to the application to meet PNG-1.0 requirements on the
|
|
* contents of the text. PNG-1.0 through PNG-1.2 discourage the use of
|
|
* any non-Latin-1 characters except for NEWLINE. ISO PNG will forbid them.
|
|
* The NUL character is forbidden by PNG-1.0 through PNG-1.2 and ISO PNG.
|
|
*/
|
|
png_write_chunk_data(png_ptr, new_key, key_len + 1);
|
|
|
|
if (text_len != 0)
|
|
png_write_chunk_data(png_ptr, (png_const_bytep)text, text_len);
|
|
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_zTXt_SUPPORTED
|
|
/* Write a compressed text chunk */
|
|
void /* PRIVATE */
|
|
png_write_zTXt(png_structrp png_ptr, png_const_charp key, png_const_charp text,
|
|
int compression)
|
|
{
|
|
unsigned int key_len;
|
|
png_byte new_key[81];
|
|
|
|
png_debug(1, "in png_write_zTXt");
|
|
|
|
if (compression != PNG_TEXT_COMPRESSION_zTXt)
|
|
png_app_warning(png_ptr, "zTXt: invalid compression type ignored");
|
|
|
|
key_len = png_check_keyword(png_ptr, key, new_key);
|
|
|
|
if (key_len == 0)
|
|
{
|
|
png_chunk_report(png_ptr, "zTXt: invalid keyword", PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
/* Add the compression method and 1 for the keyword separator. */
|
|
++key_len;
|
|
new_key[key_len++] = PNG_COMPRESSION_TYPE_BASE;
|
|
|
|
if (png_compress_chunk_data(png_ptr, png_zTXt, key_len, text, strlen(text)))
|
|
{
|
|
png_write_chunk_header(png_ptr, png_zTXt,
|
|
key_len+png_length_compressed_chunk_data(png_ptr, key_len));
|
|
png_write_chunk_data(png_ptr, new_key, key_len);
|
|
png_write_compressed_chunk_data(png_ptr);
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
|
|
/* else chunk report already issued and ignored */
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_iTXt_SUPPORTED
|
|
/* Write an iTXt chunk */
|
|
void /* PRIVATE */
|
|
png_write_iTXt(png_structrp png_ptr, int compression, png_const_charp key,
|
|
png_const_charp lang, png_const_charp lang_key, png_const_charp text)
|
|
{
|
|
png_uint_32 key_len, prefix_len, data_len;
|
|
png_size_t lang_len, lang_key_len, text_len;
|
|
png_byte new_key[82]; /* 80 bytes for the key, 2 byte compression info */
|
|
|
|
png_debug(1, "in png_write_iTXt");
|
|
|
|
key_len = png_check_keyword(png_ptr, key, new_key);
|
|
|
|
if (key_len == 0)
|
|
{
|
|
png_chunk_report(png_ptr, "iTXt: invalid keyword", PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
debug(new_key[key_len] == 0);
|
|
++key_len; /* terminating 0 added by png_check_keyword */
|
|
|
|
/* Set the compression flag */
|
|
switch (compression)
|
|
{
|
|
case PNG_ITXT_COMPRESSION_NONE:
|
|
case PNG_TEXT_COMPRESSION_NONE:
|
|
compression = new_key[key_len++] = 0; /* no compression */
|
|
break;
|
|
|
|
case PNG_TEXT_COMPRESSION_zTXt:
|
|
case PNG_ITXT_COMPRESSION_zTXt:
|
|
compression = new_key[key_len++] = 1; /* compressed */
|
|
break;
|
|
|
|
default:
|
|
png_chunk_report(png_ptr, "iTXt: invalid compression",
|
|
PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
new_key[key_len++] = PNG_COMPRESSION_TYPE_BASE;
|
|
|
|
/* We leave it to the application to meet PNG-1.0 requirements on the
|
|
* contents of the text. PNG-1.0 through PNG-1.2 discourage the use of
|
|
* any non-Latin-1 characters except for NEWLINE (yes, this is really weird
|
|
* in an 'international' text string. ISO PNG, however, specifies that the
|
|
* text is UTF-8 and this *IS NOT YET CHECKED*, so invalid sequences may be
|
|
* present.
|
|
*
|
|
* The NUL character is forbidden by PNG-1.0 through PNG-1.2 and ISO PNG.
|
|
*
|
|
* TODO: validate the language tag correctly (see the spec.)
|
|
*/
|
|
if (lang == NULL) lang = ""; /* empty language is valid */
|
|
lang_len = strlen(lang)+1U;
|
|
if (lang_key == NULL) lang_key = ""; /* may be empty */
|
|
lang_key_len = strlen(lang_key)+1U;
|
|
if (text == NULL) text = ""; /* may be empty */
|
|
|
|
if (lang_len > PNG_UINT_31_MAX-key_len ||
|
|
lang_key_len > PNG_UINT_31_MAX-key_len-lang_len)
|
|
{
|
|
png_chunk_report(png_ptr, "iTXt: prefix too long", PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
prefix_len = (png_uint_32)/*SAFE*/(key_len+lang_len+lang_key_len);
|
|
text_len = strlen(text); /* no trailing '\0' */
|
|
|
|
if (compression != 0)
|
|
{
|
|
if (png_compress_chunk_data(png_ptr, png_iTXt, prefix_len, text,
|
|
text_len))
|
|
data_len = png_length_compressed_chunk_data(png_ptr, prefix_len);
|
|
|
|
else
|
|
return; /* chunk report already issued and ignored */
|
|
}
|
|
|
|
else
|
|
{
|
|
if (text_len > PNG_UINT_31_MAX-prefix_len)
|
|
{
|
|
png_chunk_report(png_ptr, "iTXt: text too long",
|
|
PNG_CHUNK_WRITE_ERROR);
|
|
return;
|
|
}
|
|
|
|
data_len = (png_uint_32)/*SAFE*/text_len;
|
|
}
|
|
|
|
png_write_chunk_header(png_ptr, png_iTXt, prefix_len+data_len);
|
|
png_write_chunk_data(png_ptr, new_key, key_len);
|
|
png_write_chunk_data(png_ptr, lang, lang_len);
|
|
png_write_chunk_data(png_ptr, lang_key, lang_key_len);
|
|
|
|
if (compression != 0)
|
|
png_write_compressed_chunk_data(png_ptr);
|
|
|
|
else
|
|
png_write_chunk_data(png_ptr, text, data_len);
|
|
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
#endif /* WRITE_iTXt */
|
|
|
|
#if defined(PNG_WRITE_oFFs_SUPPORTED) ||\
|
|
defined(PNG_WRITE_pCAL_SUPPORTED)
|
|
/* PNG signed integers are saved in 32-bit 2's complement format. ANSI C-90
|
|
* defines a cast of a signed integer to an unsigned integer either to preserve
|
|
* the value, if it is positive, or to calculate:
|
|
*
|
|
* (UNSIGNED_MAX+1) + integer
|
|
*
|
|
* Where UNSIGNED_MAX is the appropriate maximum unsigned value, so when the
|
|
* negative integral value is added the result will be an unsigned value
|
|
* correspnding to the 2's complement representation.
|
|
*/
|
|
static int
|
|
save_int_32(png_structrp png_ptr, png_bytep buf, png_int_32 j)
|
|
{
|
|
png_uint_32 i = 0xFFFFFFFFU & (png_uint_32)/*SAFE & CORRECT*/j;
|
|
|
|
if (i != 0x80000000U/*value not permitted*/)
|
|
{
|
|
png_save_uint_32(buf, i);
|
|
return 1;
|
|
}
|
|
|
|
else
|
|
{
|
|
png_chunk_report(png_ptr, "invalid value in oFFS or pCAL",
|
|
PNG_CHUNK_WRITE_ERROR);
|
|
return 0;
|
|
}
|
|
}
|
|
#endif /* WRITE_oFFs || WRITE_pCAL */
|
|
|
|
#ifdef PNG_WRITE_oFFs_SUPPORTED
|
|
/* Write the oFFs chunk */
|
|
void /* PRIVATE */
|
|
png_write_oFFs(png_structrp png_ptr, png_int_32 x_offset, png_int_32 y_offset,
|
|
int unit_type)
|
|
{
|
|
png_byte buf[9];
|
|
|
|
png_debug(1, "in png_write_oFFs");
|
|
|
|
if (unit_type >= PNG_OFFSET_LAST)
|
|
png_warning(png_ptr, "Unrecognized unit type for oFFs chunk");
|
|
|
|
if (save_int_32(png_ptr, buf, x_offset) &&
|
|
save_int_32(png_ptr, buf + 4, y_offset))
|
|
{
|
|
/* unit type is 0 or 1, this has been checked already so the following
|
|
* is safe:
|
|
*/
|
|
buf[8] = unit_type != 0;
|
|
png_write_complete_chunk(png_ptr, png_oFFs, buf, (png_size_t)9);
|
|
}
|
|
}
|
|
#endif /* WRITE_oFFs */
|
|
|
|
#ifdef PNG_WRITE_pCAL_SUPPORTED
|
|
/* Write the pCAL chunk (described in the PNG extensions document) */
|
|
void /* PRIVATE */
|
|
png_write_pCAL(png_structrp png_ptr, png_charp purpose, png_int_32 X0,
|
|
png_int_32 X1, int type, int nparams, png_const_charp units,
|
|
png_charpp params)
|
|
{
|
|
png_uint_32 purpose_len;
|
|
size_t units_len;
|
|
png_byte buf[10];
|
|
png_byte new_purpose[80];
|
|
|
|
png_debug1(1, "in png_write_pCAL (%d parameters)", nparams);
|
|
|
|
if (type >= PNG_EQUATION_LAST)
|
|
png_error(png_ptr, "Unrecognized equation type for pCAL chunk");
|
|
|
|
purpose_len = png_check_keyword(png_ptr, purpose, new_purpose);
|
|
|
|
if (purpose_len == 0)
|
|
png_error(png_ptr, "pCAL: invalid keyword");
|
|
|
|
++purpose_len; /* terminator */
|
|
|
|
png_debug1(3, "pCAL purpose length = %d", (int)purpose_len);
|
|
units_len = strlen(units) + (nparams == 0 ? 0 : 1);
|
|
png_debug1(3, "pCAL units length = %d", (int)units_len);
|
|
|
|
if (save_int_32(png_ptr, buf, X0) &&
|
|
save_int_32(png_ptr, buf + 4, X1))
|
|
{
|
|
png_size_tp params_len = png_voidcast(png_size_tp,
|
|
png_malloc(png_ptr, nparams * sizeof (png_size_t)));
|
|
int i;
|
|
size_t total_len = purpose_len + units_len + 10;
|
|
|
|
/* Find the length of each parameter, making sure we don't count the
|
|
* null terminator for the last parameter.
|
|
*/
|
|
for (i = 0; i < nparams; i++)
|
|
{
|
|
params_len[i] = strlen(params[i]) + (i == nparams - 1 ? 0 : 1);
|
|
png_debug2(3, "pCAL parameter %d length = %lu", i,
|
|
(unsigned long)params_len[i]);
|
|
total_len += params_len[i];
|
|
}
|
|
|
|
png_debug1(3, "pCAL total length = %d", (int)total_len);
|
|
png_write_chunk_header(png_ptr, png_pCAL, (png_uint_32)total_len);
|
|
png_write_chunk_data(png_ptr, new_purpose, purpose_len);
|
|
buf[8] = png_check_byte(png_ptr, type);
|
|
buf[9] = png_check_byte(png_ptr, nparams);
|
|
png_write_chunk_data(png_ptr, buf, (png_size_t)10);
|
|
png_write_chunk_data(png_ptr, (png_const_bytep)units,
|
|
(png_size_t)units_len);
|
|
|
|
for (i = 0; i < nparams; i++)
|
|
png_write_chunk_data(png_ptr, (png_const_bytep)params[i],
|
|
params_len[i]);
|
|
|
|
png_free(png_ptr, params_len);
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
}
|
|
#endif /* WRITE_pCAL */
|
|
|
|
#ifdef PNG_WRITE_sCAL_SUPPORTED
|
|
/* Write the sCAL chunk */
|
|
void /* PRIVATE */
|
|
png_write_sCAL_s(png_structrp png_ptr, int unit, png_const_charp width,
|
|
png_const_charp height)
|
|
{
|
|
png_byte buf[64];
|
|
png_size_t wlen, hlen, total_len;
|
|
|
|
png_debug(1, "in png_write_sCAL_s");
|
|
|
|
wlen = strlen(width);
|
|
hlen = strlen(height);
|
|
total_len = wlen + hlen + 2;
|
|
|
|
if (total_len > 64)
|
|
{
|
|
png_warning(png_ptr, "Can't write sCAL (buffer too small)");
|
|
return;
|
|
}
|
|
|
|
buf[0] = png_check_byte(png_ptr, unit);
|
|
memcpy(buf + 1, width, wlen + 1); /* Append the '\0' here */
|
|
memcpy(buf + wlen + 2, height, hlen); /* Do NOT append the '\0' here */
|
|
|
|
png_debug1(3, "sCAL total length = %u", (unsigned int)total_len);
|
|
png_write_complete_chunk(png_ptr, png_sCAL, buf, total_len);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_pHYs_SUPPORTED
|
|
/* Write the pHYs chunk */
|
|
void /* PRIVATE */
|
|
png_write_pHYs(png_structrp png_ptr, png_uint_32 x_pixels_per_unit,
|
|
png_uint_32 y_pixels_per_unit,
|
|
int unit_type)
|
|
{
|
|
png_byte buf[9];
|
|
|
|
png_debug(1, "in png_write_pHYs");
|
|
|
|
if (unit_type >= PNG_RESOLUTION_LAST)
|
|
png_warning(png_ptr, "Unrecognized unit type for pHYs chunk");
|
|
|
|
png_save_uint_32(buf, x_pixels_per_unit);
|
|
png_save_uint_32(buf + 4, y_pixels_per_unit);
|
|
buf[8] = png_check_byte(png_ptr, unit_type);
|
|
|
|
png_write_complete_chunk(png_ptr, png_pHYs, buf, (png_size_t)9);
|
|
}
|
|
#endif
|
|
|
|
#ifdef PNG_WRITE_tIME_SUPPORTED
|
|
/* Write the tIME chunk. Use either png_convert_from_struct_tm()
|
|
* or png_convert_from_time_t(), or fill in the structure yourself.
|
|
*/
|
|
void /* PRIVATE */
|
|
png_write_tIME(png_structrp png_ptr, png_const_timep mod_time)
|
|
{
|
|
png_byte buf[7];
|
|
|
|
png_debug(1, "in png_write_tIME");
|
|
|
|
if (mod_time->month > 12 || mod_time->month < 1 ||
|
|
mod_time->day > 31 || mod_time->day < 1 ||
|
|
mod_time->hour > 23 || mod_time->second > 60)
|
|
{
|
|
png_warning(png_ptr, "Invalid time specified for tIME chunk");
|
|
return;
|
|
}
|
|
|
|
png_save_uint_16(buf, mod_time->year);
|
|
buf[2] = mod_time->month;
|
|
buf[3] = mod_time->day;
|
|
buf[4] = mod_time->hour;
|
|
buf[5] = mod_time->minute;
|
|
buf[6] = mod_time->second;
|
|
|
|
png_write_complete_chunk(png_ptr, png_tIME, buf, (png_size_t)7);
|
|
}
|
|
#endif
|
|
|
|
static void
|
|
png_end_IDAT(png_structrp png_ptr)
|
|
{
|
|
png_zlib_statep ps = png_ptr->zlib_state;
|
|
|
|
png_ptr->zowner = 0U; /* release the stream */
|
|
|
|
if (ps != NULL)
|
|
png_deflate_release(png_ptr, ps, 1/*check*/);
|
|
}
|
|
|
|
static void
|
|
png_write_IDAT(png_structrp png_ptr, int flush)
|
|
{
|
|
png_zlib_statep ps = png_ptr->zlib_state;
|
|
png_uint_32 IDAT_size;
|
|
|
|
/* Check for a correctly initialized list, the requirement that the end
|
|
* pointer is NULL means that the end of the list can be easily detected.
|
|
*/
|
|
affirm(ps != NULL && ps->s.end != NULL && *ps->s.end == NULL);
|
|
png_zlib_compress_validate(&png_ptr->zlib_state->s, 0/*in_use*/);
|
|
|
|
IDAT_size = png_ptr->IDAT_size;
|
|
if (IDAT_size == 0U)
|
|
{
|
|
switch (pz_get(ps, IDAT, png_level, PNG_DEFAULT_COMPRESSION_LEVEL))
|
|
{
|
|
case PNG_COMPRESSION_COMPAT: /* Legacy */
|
|
IDAT_size = 8192U;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_LOW_MEMORY:
|
|
case PNG_COMPRESSION_HIGH_SPEED:
|
|
case PNG_COMPRESSION_LOW:
|
|
/* png_compress uses PNG_ROW_BUFFER_SIZE buffers for the compressed
|
|
* data. Optimize to allocate only one of these:
|
|
*/
|
|
IDAT_size = PNG_ROW_BUFFER_SIZE;
|
|
break;
|
|
|
|
default:
|
|
case PNG_COMPRESSION_MEDIUM:
|
|
IDAT_size = PNG_ZBUF_SIZE;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_HIGH_READ_SPEED:
|
|
/* Assume the reader reads partial IDAT chunks (pretty much a
|
|
* requirement given that some PNG encoders produce just one IDAT)
|
|
*/
|
|
case PNG_COMPRESSION_HIGH:
|
|
/* This doesn't control the amount of memory allocated unless the
|
|
* PNG IDAT data really is this big.
|
|
*
|
|
* TODO: review handling out-of-memory from png_compress() by
|
|
* flushing an IDAT.
|
|
*/
|
|
IDAT_size = PNG_UINT_31_MAX;
|
|
break;
|
|
}
|
|
}
|
|
|
|
/* Write IDAT chunks while either 'flush' is true or there are at
|
|
* least png_ptr->IDAT_size bytes available to be written.
|
|
*/
|
|
for (;;)
|
|
{
|
|
png_uint_32 len = IDAT_size;
|
|
|
|
if (ps->s.overflow == 0U)
|
|
{
|
|
png_uint_32 avail = ps->s.len;
|
|
|
|
if (avail < len)
|
|
{
|
|
/* When end_of_image is true everything gets written, otherwise
|
|
* there must be at least IDAT_size bytes available.
|
|
*/
|
|
if (!flush)
|
|
return;
|
|
|
|
if (avail == 0U)
|
|
break;
|
|
|
|
len = avail;
|
|
}
|
|
}
|
|
|
|
png_write_chunk_header(png_ptr, png_IDAT, len);
|
|
|
|
/* Write bytes from the buffer list, adjusting {overflow,len} as they are
|
|
* written.
|
|
*/
|
|
do
|
|
{
|
|
png_compression_bufferp next = ps->s.list;
|
|
unsigned int avail = sizeof next->output;
|
|
unsigned int start = ps->s.start;
|
|
unsigned int written;
|
|
|
|
affirm(next != NULL);
|
|
|
|
if (next->next == NULL) /* end of list */
|
|
{
|
|
/* The z_stream should always be pointing into this output buffer,
|
|
* the buffer may not be full:
|
|
*/
|
|
debug(ps->s.zs.next_out + ps->s.zs.avail_out ==
|
|
next->output + sizeof next->output);
|
|
avail -= ps->s.zs.avail_out;
|
|
}
|
|
|
|
else /* not end of list */
|
|
debug((ps->s.zs.next_out < next->output ||
|
|
ps->s.zs.next_out > next->output + sizeof next->output) &&
|
|
(ps->s.overflow > 0 ||
|
|
ps->s.start + ps->s.len >= sizeof next->output));
|
|
|
|
/* First, if this is the very first IDAT (PNG_HAVE_IDAT not set)
|
|
* fix the Zlib CINFO field if required:
|
|
*/
|
|
if ((png_ptr->mode & PNG_HAVE_IDAT) == 0U &&
|
|
avail >= start+2U /* enough for the zlib header */)
|
|
{
|
|
debug(start == 0U);
|
|
fix_cinfo(ps, next->output+start, png_image_size(png_ptr));
|
|
}
|
|
|
|
else /* always expect to see at least 2 bytes: */
|
|
debug((png_ptr->mode & PNG_HAVE_IDAT) != 0U);
|
|
|
|
/* Set this now to prevent the above happening again second time round
|
|
* the loop:
|
|
*/
|
|
png_ptr->mode |= PNG_HAVE_IDAT;
|
|
|
|
if (avail <= start+len)
|
|
{
|
|
/* Write all of this buffer: */
|
|
affirm(avail > start); /* else overflow on the subtract */
|
|
written = avail-start;
|
|
png_write_chunk_data(png_ptr, next->output+start, written);
|
|
|
|
/* At the end there are no buffers in the list but the z_stream
|
|
* still points into the old (just released) buffer. This can
|
|
* happen when the old buffer is not full if the compressed bytes
|
|
* exactly match the IDAT length; it should always happen when
|
|
* end_of_image is set.
|
|
*/
|
|
ps->s.list = next->next;
|
|
|
|
if (next->next == NULL)
|
|
{
|
|
debug(avail == start+len);
|
|
ps->s.end = &ps->s.list;
|
|
ps->s.zs.next_out = NULL;
|
|
ps->s.zs.avail_out = 0U;
|
|
}
|
|
|
|
next->next = ps->stash;
|
|
ps->stash = next;
|
|
ps->s.start = 0U;
|
|
}
|
|
|
|
else /* write only part of this buffer */
|
|
{
|
|
written = len;
|
|
png_write_chunk_data(png_ptr, next->output+start, written);
|
|
ps->s.start = (unsigned int)/*SAFE*/(start + written);
|
|
}
|
|
|
|
/* 'written' bytes were written: */
|
|
len -= written;
|
|
|
|
if (written <= ps->s.len)
|
|
ps->s.len -= written;
|
|
|
|
else
|
|
{
|
|
affirm(ps->s.overflow > 0U);
|
|
--ps->s.overflow;
|
|
ps->s.len += 0x80000000U - written;
|
|
UNTESTED
|
|
}
|
|
}
|
|
while (len > 0U);
|
|
|
|
png_write_chunk_end(png_ptr);
|
|
}
|
|
|
|
/* avail == 0 && flush */
|
|
png_end_IDAT(png_ptr);
|
|
png_ptr->mode |= PNG_AFTER_IDAT;
|
|
}
|
|
|
|
/* This is is a convenience wrapper to handle IDAT compression; it takes a
|
|
* pointer to the input data and places no limit on the size of the output but
|
|
* is otherwise the same as png_compress(). It also handles the use of the
|
|
* stash (only used for IDAT compression.)
|
|
*/
|
|
static int
|
|
png_compress_IDAT_data(png_structrp png_ptr, png_zlib_statep ps,
|
|
png_zlib_compressp pz, png_const_voidp input, uInt input_len, int flush)
|
|
{
|
|
/* Delay initialize the z_stream. */
|
|
if (png_ptr->zowner != png_IDAT)
|
|
png_deflate_claim(png_ptr, png_IDAT, 0U);
|
|
|
|
affirm(png_ptr->zowner == png_IDAT && pz->end != NULL && *pz->end == NULL);
|
|
|
|
/* z_stream::{next,avail}_out are set by png_compress to point into the
|
|
* buffer list. next_in must be set here, avail_in comes from the input_len
|
|
* parameter:
|
|
*/
|
|
pz->zs.next_in = PNGZ_INPUT_CAST(png_voidcast(const Bytef*, input));
|
|
*pz->end = ps->stash; /* May be NULL */
|
|
ps->stash = NULL;
|
|
|
|
/* zlib buffers the output, the maximum amount of compressed data that can be
|
|
* produced here is governed by the amount of buffering.
|
|
*/
|
|
{
|
|
int ret = png_compress(pz, input_len, 0U/*unlimited*/, flush);
|
|
|
|
affirm(pz->end != NULL && ps->stash == NULL);
|
|
ps->stash = *pz->end; /* May be NULL */
|
|
*pz->end = NULL;
|
|
|
|
/* Z_FINISH should give Z_STREAM_END, everything else should give Z_OK, in
|
|
* either case all the input should have been consumed:
|
|
*/
|
|
implies(ret == Z_OK || ret == Z_FINISH, pz->zs.avail_in == 0U &&
|
|
(ret == Z_STREAM_END) == (flush == Z_FINISH));
|
|
pz->zs.next_in = NULL;
|
|
pz->zs.avail_in = 0U; /* safety */
|
|
png_zlib_compress_validate(pz, 0/*in_use*/);
|
|
|
|
return ret;
|
|
}
|
|
}
|
|
|
|
/* Compress some image data using the main png_zlib_compress. Write the result
|
|
* out if there is sufficient data.
|
|
*/
|
|
static void
|
|
png_compress_IDAT(png_structrp png_ptr, png_const_voidp input, uInt input_len,
|
|
int flush)
|
|
{
|
|
png_zlib_statep ps = png_ptr->zlib_state;
|
|
int ret = png_compress_IDAT_data(png_ptr, ps, &ps->s, input, input_len,
|
|
flush);
|
|
|
|
/* Check the return code. */
|
|
if (ret == Z_OK || ret == Z_STREAM_END)
|
|
png_write_IDAT(png_ptr, flush == Z_FINISH);
|
|
|
|
else /* ret != Z_OK && ret != Z_STREAM_END */
|
|
{
|
|
/* This is an error condition. It is fatal. */
|
|
png_end_IDAT(png_ptr);
|
|
png_zstream_error(&ps->s.zs, ret);
|
|
png_error(png_ptr, ps->s.zs.msg);
|
|
}
|
|
}
|
|
|
|
/* This is called at the end of every row to handle the required callbacks and
|
|
* advance png_struct::row_number and png_struct::pass.
|
|
*/
|
|
static void
|
|
png_write_end_row(png_structrp png_ptr, int flush)
|
|
{
|
|
png_uint_32 row_number = png_ptr->row_number;
|
|
unsigned int pass = png_ptr->pass;
|
|
|
|
debug(pass < 7U);
|
|
implies(flush == Z_FINISH, png_ptr->zowner == 0U);
|
|
|
|
/* API NOTE: the write callback is made before any changes to the row number
|
|
* or pass however, in 1.7.0, the zlib stream can be closed before the
|
|
* callback is made (this is new). The application flush function happens
|
|
* afterward as was the case before. In 1.7.0 this is solely determined by
|
|
* the order of the code that follows.
|
|
*/
|
|
if (png_ptr->write_row_fn != NULL)
|
|
png_ptr->write_row_fn(png_ptr, row_number, pass);
|
|
|
|
# ifdef PNG_WRITE_FLUSH_SUPPORTED
|
|
if (flush == Z_SYNC_FLUSH)
|
|
{
|
|
if (png_ptr->output_flush_fn != NULL)
|
|
png_ptr->output_flush_fn(png_ptr);
|
|
png_ptr->zlib_state->flush_rows = 0U;
|
|
}
|
|
# else /* !WRITE_FLUSH */
|
|
PNG_UNUSED(flush)
|
|
# endif /* !WRITE_FLUSH */
|
|
|
|
/* Finally advance to the next row/pass: */
|
|
if (png_ptr->interlaced == PNG_INTERLACE_NONE)
|
|
{
|
|
debug(row_number < png_ptr->height);
|
|
|
|
if (++row_number == png_ptr->height) /* last row */
|
|
{
|
|
row_number = 0U;
|
|
debug(flush == Z_FINISH);
|
|
png_ptr->pass = 7U;
|
|
}
|
|
}
|
|
|
|
# ifdef PNG_WRITE_INTERLACING_SUPPORTED
|
|
else /* interlaced */ if (png_ptr->do_interlace)
|
|
{
|
|
/* This gets called only for rows that are processed; i.e. rows that
|
|
* are in the pass of a pass which is itself in the output.
|
|
*/
|
|
debug(row_number < png_ptr->height &&
|
|
PNG_PASS_IN_IMAGE(png_ptr->width, png_ptr->height, pass) &&
|
|
pass <= PNG_LAST_PASS(png_ptr->width, png_ptr->height) &&
|
|
PNG_ROW_IN_INTERLACE_PASS(row_number, pass));
|
|
|
|
/* NOTE: the last row of the original image may not be in the pass, in
|
|
* this case the code which skipped the row must do the increment
|
|
* below! See 'interlace_row' in pngwrite.c and the code in
|
|
* png_write_png_rows below.
|
|
*
|
|
* In that case an earlier row will be the last one in the pass (if the
|
|
* pass is in the output), check this here:
|
|
*/
|
|
implies(pass == PNG_LAST_PASS(png_ptr->width, png_ptr->height) &&
|
|
PNG_LAST_PASS_ROW(row_number, pass, png_ptr->height),
|
|
flush == Z_FINISH);
|
|
|
|
if (++row_number == png_ptr->height) /* last row */
|
|
{
|
|
row_number = 0U;
|
|
png_ptr->pass = 0x7U & ++pass;
|
|
}
|
|
}
|
|
# endif /* WRITE_INTERLACING */
|
|
|
|
else /* application does interlace */
|
|
{
|
|
implies(png_ptr->height == 1U, pass != 6U);
|
|
debug(PNG_PASS_IN_IMAGE(png_ptr->width, png_ptr->height, pass) &&
|
|
row_number < PNG_PASS_ROWS(png_ptr->height, pass));
|
|
|
|
if (++row_number == PNG_PASS_ROWS(png_ptr->height, pass))
|
|
{
|
|
/* last row in this pass, next one may be empty. */
|
|
row_number = 0U;
|
|
|
|
do
|
|
++pass;
|
|
while (pass < 7U &&
|
|
!PNG_PASS_IN_IMAGE(png_ptr->width, png_ptr->height, pass));
|
|
|
|
implies(png_ptr->height == 1U, pass != 6U);
|
|
implies(pass == 7U, flush == Z_FINISH);
|
|
png_ptr->pass = 0x7U & pass;
|
|
}
|
|
}
|
|
|
|
png_ptr->row_number = row_number;
|
|
}
|
|
|
|
#ifdef PNG_WRITE_FLUSH_SUPPORTED
|
|
/* Flush the current output buffers now */
|
|
void PNGAPI
|
|
png_write_flush(png_structrp png_ptr)
|
|
{
|
|
png_debug(1, "in png_write_flush");
|
|
|
|
/* Force a flush at the end of the current row by setting 'flush_rows' to the
|
|
* maximum:
|
|
*/
|
|
if (png_ptr != NULL && png_ptr->zlib_state != NULL)
|
|
png_ptr->zlib_state->flush_rows = 0xEFFFFFFF;
|
|
}
|
|
|
|
/* Return the correct flush to use */
|
|
static int
|
|
row_flush(png_zlib_statep ps, unsigned int row_info_flags)
|
|
{
|
|
if (PNG_IDAT_END(row_info_flags))
|
|
return Z_FINISH;
|
|
|
|
else if ((row_info_flags & png_row_end) != 0 &&
|
|
++ps->flush_rows >= ps->flush_dist)
|
|
return Z_SYNC_FLUSH;
|
|
|
|
else
|
|
return Z_NO_FLUSH;
|
|
}
|
|
#else /* !WRITE_FLUSH */
|
|
# define row_flush(ps, ri) (PNG_IDAT_END(ri) ? Z_FINISH : Z_NO_FLUSH)
|
|
#endif /* !WRITE_FLUSH */
|
|
|
|
static void
|
|
write_filtered_row(png_structrp png_ptr, png_const_voidp filtered_row,
|
|
unsigned int row_bytes, unsigned int filter /*if at start of row*/,
|
|
int flush)
|
|
{
|
|
/* This handles writing a row that has been filtered, or did not need to be
|
|
* filtered. If the data row has a partial pixel it must have been handled
|
|
* correctly in the caller; filters generate a full 8 bits even if the pixel
|
|
* only has one significant bit!
|
|
*/
|
|
debug(row_bytes > 0);
|
|
affirm(row_bytes <= ZLIB_IO_MAX); /* I.e. it fits in a uInt */
|
|
|
|
if (filter < PNG_FILTER_VALUE_LAST) /* start of row */
|
|
{
|
|
png_byte buffer[1];
|
|
|
|
buffer[0] = PNG_BYTE(filter);
|
|
png_compress_IDAT(png_ptr, buffer, 1U/*len*/, Z_NO_FLUSH);
|
|
}
|
|
|
|
png_compress_IDAT(png_ptr, filtered_row, row_bytes, flush);
|
|
}
|
|
|
|
static void
|
|
write_unfiltered_rowbits(png_structrp png_ptr, png_const_bytep filtered_row,
|
|
unsigned int row_bits, png_byte filter /*if at start of row*/,
|
|
int flush)
|
|
{
|
|
/* Same as above, but it correctly clears the unused bits in a partial
|
|
* byte.
|
|
*/
|
|
const png_uint_32 row_bytes = row_bits >> 3;
|
|
|
|
debug(filter == PNG_FILTER_VALUE_NONE || filter == PNG_FILTER_VALUE_LAST);
|
|
|
|
if (row_bytes > 0U)
|
|
{
|
|
row_bits -= row_bytes << 3;
|
|
write_filtered_row(png_ptr, filtered_row, row_bytes, filter,
|
|
row_bits == 0U ? flush : Z_NO_FLUSH);
|
|
filter = PNG_FILTER_VALUE_LAST; /* written */
|
|
}
|
|
|
|
/* Handle a partial byte. */
|
|
if (row_bits > 0U)
|
|
{
|
|
png_byte buffer[1];
|
|
|
|
buffer[0] = PNG_BYTE(filtered_row[row_bytes] & ~(0xFFU >> row_bits));
|
|
write_filtered_row(png_ptr, buffer, 1U, filter, flush);
|
|
}
|
|
}
|
|
|
|
#ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
static void
|
|
filter_block_singlebyte(unsigned int row_bytes, png_bytep sub_row,
|
|
png_bytep up_row, png_bytep avg_row, png_bytep paeth_row,
|
|
png_const_bytep row, png_const_bytep prev_row, png_bytep prev_pixels)
|
|
{
|
|
/* Calculate rows for all four filters where the input has one byte per pixel
|
|
* (more accurately per filter-unit).
|
|
*/
|
|
png_byte a = prev_pixels[0];
|
|
png_byte c = prev_pixels[1];
|
|
|
|
while (row_bytes-- > 0U)
|
|
{
|
|
const png_byte x = *row++;
|
|
const png_byte b = prev_row == NULL ? 0U : *prev_row++;
|
|
|
|
/* Calculate each filtered byte in turn: */
|
|
if (sub_row != NULL) *sub_row++ = 0xFFU & (x - a);
|
|
if (up_row != NULL) *up_row++ = 0xFFU & (x - b);
|
|
if (avg_row != NULL) *avg_row++ = 0xFFU & (x - (a+b)/2U);
|
|
|
|
/* Paeth is a little more difficult: */
|
|
if (paeth_row != NULL)
|
|
{
|
|
int pa = b-c; /* a+b-c - a */
|
|
int pb = a-c; /* a+b-c - b */
|
|
int pc = pa+pb; /* a+b-c - c = b-c + a-c */
|
|
png_byte p = a;
|
|
|
|
pa = abs(pa);
|
|
pb = abs(pb);
|
|
if (pa > pb) pa = pb, p = b;
|
|
if (pa > abs(pc)) p = c;
|
|
|
|
*paeth_row++ = 0xFFU & (x - p);
|
|
}
|
|
|
|
/* And set a and c for the next pixel: */
|
|
a = x;
|
|
c = b;
|
|
}
|
|
|
|
/* Store a and c for the next block: */
|
|
prev_pixels[0] = a;
|
|
prev_pixels[1] = c;
|
|
}
|
|
|
|
static void
|
|
filter_block_multibyte(unsigned int row_bytes,
|
|
const unsigned int bpp, png_bytep sub_row, png_bytep up_row,
|
|
png_bytep avg_row, png_bytep paeth_row, png_const_bytep row,
|
|
png_const_bytep prev_row, png_bytep prev_pixels)
|
|
{
|
|
/* Calculate rows for all four filters, the input is a block of bytes such
|
|
* that row_bytes is a multiple of bpp. bpp can be 2, 3, 4, 6 or 8.
|
|
* prev_pixels will be updated to the last pixels processed.
|
|
*/
|
|
while (row_bytes >= bpp)
|
|
{
|
|
unsigned int i;
|
|
|
|
for (i=0; i<bpp; ++i)
|
|
{
|
|
const png_byte a = prev_pixels[i];
|
|
const png_byte c = prev_pixels[i+bpp];
|
|
const png_byte b = prev_row == NULL ? 0U : *prev_row++;
|
|
const png_byte x = *row++;
|
|
|
|
/* Save for the next pixel: */
|
|
prev_pixels[i] = x;
|
|
prev_pixels[i+bpp] = b;
|
|
|
|
/* Calculate each filtered byte in turn: */
|
|
if (sub_row != NULL) *sub_row++ = 0xFFU & (x - a);
|
|
if (up_row != NULL) *up_row++ = 0xFFU & (x - b);
|
|
if (avg_row != NULL) *avg_row++ = 0xFFU & (x - (a+b)/2U);
|
|
|
|
/* Paeth is a little more difficult: */
|
|
if (paeth_row != NULL)
|
|
{
|
|
int pa = b-c; /* a+b-c - a */
|
|
int pb = a-c; /* a+b-c - b */
|
|
int pc = pa+pb; /* a+b-c - c = b-c + a-c */
|
|
png_byte p = a;
|
|
|
|
pa = abs(pa);
|
|
pb = abs(pb);
|
|
if (pa > pb) pa = pb, p = b;
|
|
if (pa > abs(pc)) p = c;
|
|
|
|
*paeth_row++ = 0xFFU & (x - p);
|
|
}
|
|
}
|
|
|
|
row_bytes -= i;
|
|
}
|
|
}
|
|
|
|
static void
|
|
filter_block(png_const_bytep prev_row, png_bytep prev_pixels,
|
|
png_const_bytep unfiltered_row, unsigned int row_bits,
|
|
const unsigned int bpp, png_bytep sub_row, png_bytep up_row,
|
|
png_bytep avg_row, png_bytep paeth_row)
|
|
{
|
|
const unsigned int row_bytes = row_bits >> 3; /* complete bytes */
|
|
|
|
if (bpp <= 8U)
|
|
{
|
|
/* There may be a partial byte at the end. */
|
|
if (row_bytes > 0)
|
|
filter_block_singlebyte(row_bytes, sub_row, up_row, avg_row, paeth_row,
|
|
unfiltered_row, prev_row, prev_pixels);
|
|
|
|
/* The partial byte must be handled correctly here; both the previous row
|
|
* value and the current value need to have non-present bits cleared.
|
|
*/
|
|
if ((row_bits & 7U) != 0)
|
|
{
|
|
const png_byte mask = PNG_BYTE(~(0xFFU >> (row_bits & 7U)));
|
|
png_byte buffer[2];
|
|
|
|
buffer[0] = unfiltered_row[row_bytes] & mask;
|
|
|
|
if (prev_row != NULL)
|
|
buffer[1U] = prev_row[row_bytes] & mask;
|
|
|
|
else
|
|
buffer[1U] = 0U;
|
|
|
|
filter_block_singlebyte(1U,
|
|
sub_row == NULL ? NULL : sub_row+row_bytes,
|
|
up_row == NULL ? NULL : up_row+row_bytes,
|
|
avg_row == NULL ? NULL : avg_row+row_bytes,
|
|
paeth_row == NULL ? NULL : paeth_row+row_bytes,
|
|
buffer, buffer+1U, prev_pixels);
|
|
}
|
|
}
|
|
|
|
else
|
|
filter_block_multibyte(row_bytes, bpp >> 3,
|
|
sub_row, up_row, avg_row, paeth_row,
|
|
unfiltered_row, prev_row, prev_pixels);
|
|
}
|
|
|
|
static void
|
|
filter_row(png_structrp png_ptr, png_const_bytep prev_row,
|
|
png_bytep prev_pixels, png_const_bytep unfiltered_row,
|
|
unsigned int row_bits, unsigned int bpp, unsigned int filter,
|
|
int start_of_row, int flush)
|
|
{
|
|
/* filters_to_try identifies a single filter and it is not PNG_FILTER_NONE.
|
|
*/
|
|
png_byte filtered_row[PNG_ROW_BUFFER_SIZE];
|
|
|
|
affirm((row_bits+7U) >> 3 <= PNG_ROW_BUFFER_SIZE &&
|
|
filter >= PNG_FILTER_VALUE_SUB && filter <= PNG_FILTER_VALUE_PAETH);
|
|
debug((row_bits % bpp) == 0U);
|
|
|
|
filter_block(prev_row, prev_pixels, unfiltered_row, row_bits, bpp,
|
|
filter == PNG_FILTER_VALUE_SUB ? filtered_row : NULL,
|
|
filter == PNG_FILTER_VALUE_UP ? filtered_row : NULL,
|
|
filter == PNG_FILTER_VALUE_AVG ? filtered_row : NULL,
|
|
filter == PNG_FILTER_VALUE_PAETH ? filtered_row : NULL);
|
|
|
|
write_filtered_row(png_ptr, filtered_row, (row_bits+7U)>>3,
|
|
start_of_row ? filter : PNG_FILTER_VALUE_LAST, flush);
|
|
}
|
|
|
|
/* Allow the application to select one or more row filters to use. */
|
|
static png_int_32
|
|
set_filter(png_zlib_statep ps, unsigned int filtersIn)
|
|
{
|
|
/* Notice that PNG_NO_FILTERS is 0 and passes this test; this is OK because
|
|
* filters then gets set to PNG_FILTER_NONE, as is required.
|
|
*
|
|
* The argument to this routine is actually an (int), but conversion to
|
|
* (unsigned int) is safe because it leaves the top bits set which results in
|
|
* PNG_EDOM below.
|
|
*/
|
|
if (filtersIn < PNG_FILTER_NONE)
|
|
filtersIn = PNG_FILTER_MASK(filtersIn);
|
|
|
|
/* PNG_ALL_FILTERS is a constant, unfortunately it is nominally signed, for
|
|
* historical reasons, hence the PNG_BIC_MASK here.
|
|
*/
|
|
if ((filtersIn & PNG_BIC_MASK(PNG_ALL_FILTERS)) == 0U)
|
|
{
|
|
# ifndef PNG_SELECT_FILTER_SUPPORTED
|
|
filtersIn &= -filtersIn; /* Use lowest set bit */
|
|
# endif /* !SELECT_FILTER */
|
|
|
|
return ps->filter_mask = filtersIn & PNG_ALL_FILTERS;
|
|
}
|
|
|
|
else /* Out-of-range filtersIn: */
|
|
return PNG_EDOM;
|
|
}
|
|
#endif /* WRITE_FILTER */
|
|
|
|
#ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
void /* PRIVATE */
|
|
png_write_start_IDAT(png_structrp png_ptr)
|
|
{
|
|
png_zlib_statep ps = get_zlib_state(png_ptr);
|
|
int png_level;
|
|
|
|
/* Set up the IDAT compression state. Expect the state to have been released
|
|
* by the previous owner, but it doesn't much matter if there was an error.
|
|
* Note that the stream is not claimed yet.
|
|
*/
|
|
debug(png_ptr->zowner == 0U);
|
|
|
|
/* This sets the buffer limits and write_row_size, which is used below. */
|
|
png_zlib_state_set_buffer_limits(png_ptr, ps);
|
|
|
|
/* Now default the filter mask if it hasn't been set already: */
|
|
png_level = pz_get(ps, IDAT, png_level, PNG_DEFAULT_COMPRESSION_LEVEL);
|
|
|
|
if (ps->filter_mask == 0)
|
|
{
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
/* If the bit depth is less than 8, so pixels are not byte aligned, PNG
|
|
* filtering hardly ever helps because there is no correlation between
|
|
* the bytes on which the filter works and the actual pixel values.
|
|
* Note that GIF is a whole lot better at this because it uses LZW to
|
|
* compress a bit-stream, not a byte stream as in the deflate
|
|
* implementation of LZ77.
|
|
*
|
|
* If the row size is less than 256 bytes filter selection algorithms
|
|
* are flakey because the restricted range of codes in each row can
|
|
* lead to poor selection of filters, particularly if the bytes in the
|
|
* image are themselves limited. (This happens when a low bit-depth
|
|
* image is encoded with 8-bit channels.)
|
|
*
|
|
* By experiment with the test set of images the breakpoint between
|
|
* not filtering and filtering based on which gives best compression by
|
|
* row size is as follows:
|
|
*
|
|
* NONE FAST ALL
|
|
* PAL <=anything [even 8-bit palette images larger if filtered]
|
|
* G<8 <=anything [low bit depth gray images]
|
|
* G8 <=16 [+~1%] >16
|
|
* G16 <=128 [+~1%] >128
|
|
* GA8 <=64 [+~1%] >64
|
|
* GA16 <=anything [always better without filtering!]
|
|
* RGB8 <=32 [+0-2%(1)] >32
|
|
* RGB16 <=1024 [+~1%] >1024
|
|
* RGBA8 <=64 [+~~1%] >64
|
|
* RGBA16 <=128 {+~0.5%] >128
|
|
*
|
|
* (1) The largest 24-bit RGB image (RGB8) faired better, by 1.3%,
|
|
* with 'fast' filters. This is assumed to be random.
|
|
*
|
|
* Aggregated across all color types and bit depths the breakpoint for
|
|
* filtering is >16 bytes, but the size increase only exceeds 0.5% for
|
|
* images with rows between 64 and 128 bytes, hence the choices below.
|
|
*
|
|
* Across all the test images that change (not including selecting just
|
|
* the 'fast' filters by default) does not change the compressed size
|
|
* significantly (+0.06% across the whole test set), however it does
|
|
* substantially increase the number of images without filtering.
|
|
*
|
|
* Using just none and sub filters results in overall compressed sizes
|
|
* somewhere around the geometric mean of no filtering and 'fast'.
|
|
*
|
|
* The image size also plays a part. Filtering is not an advantage for
|
|
* images of size <= 512 bytes. This is also reflected below.
|
|
*
|
|
* NOTE: the libpng 1.6 (and earlier) algorithm seems to work
|
|
* because it biases the byte codes in the output towards 0 and 255.
|
|
* Zlib doesn't care what the codes are, but Huffman encoding always
|
|
* benefits from a biased distribution and the filters themselves were
|
|
* designed to produce values in this range.
|
|
*
|
|
* In a raw comparison with the legacy code selection of specific sets
|
|
* of filters always increased the compressed size of the test set, as
|
|
* follows:
|
|
*
|
|
* PNG_ALL_FILTERS: +0.26%
|
|
* PNG_FAST_FILTERS: +1.9%
|
|
* NONE+SUB: +5.8%
|
|
* PNG_NO_FILTERS: +14%
|
|
*
|
|
* This mainly proves that a static selection of filters (without
|
|
* considering the PNG format) is always worse than the legacy
|
|
* algorithm below.
|
|
*
|
|
* NOTE: ps->filter_mask must be set to a mask value, not a simple
|
|
* PNG_FILTER_VALUE_ number.
|
|
*/
|
|
if (ps->write_row_size == 0U /* row cannot be buffered */)
|
|
ps->filter_mask = PNG_FILTER_NONE;
|
|
|
|
else if (png_level == PNG_COMPRESSION_COMPAT/* Legacy */)
|
|
{
|
|
if (png_ptr->color_type == PNG_COLOR_TYPE_PALETTE ||
|
|
png_ptr->bit_depth < 8U)
|
|
ps->filter_mask = PNG_FILTER_NONE;
|
|
|
|
else
|
|
ps->filter_mask = PNG_ALL_FILTERS;
|
|
}
|
|
|
|
/* NOTE: overall with the following size tests (row and image size) the
|
|
* test set of images end up 0.06% larger, however some color types are
|
|
* smaller and some larger; the differences are minute. If the test is
|
|
* <=128 (which means <=129 bytes per row with the filter byte) the
|
|
* resultant inclusion of 32x32 RGBA images results in significantly
|
|
* increased compressed size.
|
|
*
|
|
* The test on png_level captures the following settings:
|
|
*
|
|
* PNG_COMPRESSION_LOW_MEMORY
|
|
* PNG_COMPRESSION_HIGH_SPEED
|
|
* PNG_COMPRESSION_HIGH_READ_SPEED
|
|
*
|
|
* NOTE: this relies on the exact values in png.h!
|
|
*/
|
|
else if (png_level <= PNG_COMPRESSION_HIGH_READ_SPEED
|
|
|| png_ptr->color_type == PNG_COLOR_TYPE_PALETTE
|
|
|| png_ptr->bit_depth < 8U
|
|
|| ps->write_row_size/*does not include filter*/ < 128U
|
|
|| png_image_size(png_ptr) <= 512U)
|
|
ps->filter_mask = PNG_FILTER_NONE;
|
|
|
|
/* ELSE: there are at least 128 bytes in every row and the pixels
|
|
* are multiples of a byte.
|
|
*/
|
|
else switch (png_level)
|
|
{
|
|
default: /* For GCC */
|
|
case PNG_COMPRESSION_LOW:
|
|
ps->filter_mask = PNG_FILTER_NONE+PNG_FILTER_SUB;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_MEDIUM:
|
|
ps->filter_mask = PNG_FAST_FILTERS;
|
|
break;
|
|
|
|
case PNG_COMPRESSION_HIGH:
|
|
ps->filter_mask = PNG_ALL_FILTERS;
|
|
break;
|
|
}
|
|
# else /* !SELECT_FILTER */
|
|
ps->filter_mask = PNG_FILTER_NONE;
|
|
# endif /* !SELECT_FILTER */
|
|
}
|
|
}
|
|
|
|
static png_byte
|
|
png_write_start_row(png_zlib_statep ps, int start_of_pass, int no_previous_row)
|
|
/* Called at the start of a row to set up anything required for filter
|
|
* handling in the row. Sets png_zlib_state::filters to a single filter.
|
|
*/
|
|
{
|
|
unsigned int mask = ps->filter_mask;
|
|
|
|
/* If we see a previous-row filter in mask and png_zlib_state::save_row is
|
|
* still unset set it. This means that the first time a previous-row filter
|
|
* is seen row-saving gets turned on.
|
|
*/
|
|
if (ps->save_row == SAVE_ROW_UNSET && (mask & PREVIOUS_ROW_FILTERS) != 0U)
|
|
ps->save_row = SAVE_ROW_DEFAULT;
|
|
|
|
if ((no_previous_row /* row not stored */ && !start_of_pass) ||
|
|
ps->save_row == SAVE_ROW_OFF /* disabled by app */ ||
|
|
ps->write_row_size == 0U /* row too large to buffer */)
|
|
mask &= PNG_BIC_MASK(PREVIOUS_ROW_FILTERS);
|
|
|
|
/* On the first row of a pass Paeth is equivalent to sub and up is equivalent
|
|
* to none, so try to simplify the mask in in this case.
|
|
*/
|
|
else if (start_of_pass) {
|
|
# define MATCH(flags) ((mask & (flags)) == (flags))
|
|
if (MATCH(PNG_FILTER_NONE|PNG_FILTER_UP))
|
|
mask &= PNG_BIC_MASK(PNG_FILTER_UP);
|
|
|
|
if (MATCH(PNG_FILTER_SUB|PNG_FILTER_PAETH))
|
|
mask &= PNG_BIC_MASK(PNG_FILTER_PAETH);
|
|
# undef MATCH
|
|
}
|
|
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
if ((mask & (mask-1U)) == 0U /* single bit set */ ||
|
|
ps->write_row_size == 0U /* row cannot be buffered */)
|
|
# endif /* SELECT_FILTER */
|
|
/* Convert the lowest set bit into the corresponding value. If no bits
|
|
* are set select NONE. After this switch statement the value of
|
|
* ps->filters is guaranteed to just be a single filter.
|
|
*/
|
|
switch (mask & -mask)
|
|
{
|
|
default: mask = PNG_FILTER_VALUE_NONE; break;
|
|
case PNG_FILTER_SUB: mask = PNG_FILTER_VALUE_SUB; break;
|
|
case PNG_FILTER_UP: mask = PNG_FILTER_VALUE_UP; break;
|
|
case PNG_FILTER_AVG: mask = PNG_FILTER_VALUE_AVG; break;
|
|
case PNG_FILTER_PAETH: mask = PNG_FILTER_VALUE_PAETH; break;
|
|
}
|
|
|
|
return ps->filters = PNG_BYTE(mask);
|
|
}
|
|
|
|
static png_bytep
|
|
allocate_row(png_structrp png_ptr, png_const_bytep data, png_alloc_size_t size)
|
|
/* Utility to allocate and save some row bytes. If the result is NULL the
|
|
* allocation failed and the png_zlib_struct will have been updated to
|
|
* prevent further allocation attempts.
|
|
*/
|
|
{
|
|
const png_zlib_statep ps = png_ptr->zlib_state;
|
|
png_bytep buffer;
|
|
|
|
debug(ps->write_row_size > 0U);
|
|
|
|
/* OOM is handled silently, as is the case where the row is too large to
|
|
* buffer.
|
|
*/
|
|
buffer = png_voidcast(png_bytep,
|
|
png_malloc_base(png_ptr, ps->write_row_size));
|
|
|
|
/* Setting write_row_size to 0 switches on the code for handling a row that
|
|
* is too large to buffer. This will kick in next time round, i.e. on the
|
|
* next row.
|
|
*/
|
|
if (buffer == NULL)
|
|
ps->write_row_size = 0U;
|
|
|
|
else
|
|
memcpy(buffer, data, size);
|
|
|
|
return buffer;
|
|
}
|
|
#endif /* WRITE_FILTER */
|
|
|
|
#ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
/* Bit set operations. Not in ANSI C-90 but commonly available in highly
|
|
* optimized versions, hence the ifndef. These operations just work on bitsets
|
|
* of size 256. The second argument (the code index) may be evaluated multiple
|
|
* times.
|
|
*/
|
|
#ifndef PNG_CODE_SET /* Can be set in pngpriv.h */
|
|
typedef png_uint_32 png_codeset[8];
|
|
# define PNG_CODE_MASK(i) (((png_uint_32)1U) << ((i) & 0x1FU))
|
|
# define PNG_CODE_IS_SET(c,i) (((c)[(i) >> 5] & PNG_CODE_MASK(i)))
|
|
# define PNG_CODE_SET(c,i) (((c)[(i) >> 5] |= PNG_CODE_MASK(i)))
|
|
# define PNG_CODE_CLEAR(c,i) (((c)[(i) >> 5] &= ~PNG_CODE_MASK(i)))
|
|
#endif /* !PNG_CODE_SET */
|
|
|
|
typedef struct filter_selector
|
|
{
|
|
/* Persistent filter selection information (stored across row boundaries).
|
|
* A code is not considered if it last occured more than 'window' bytes ago.
|
|
* The deflate algorithm means that 'window' cannot exceed 8453377, however
|
|
* practical versions may be far less. When 'distance' reaches 'window' any
|
|
* code where:
|
|
*
|
|
* distance - code_distance[code] > window
|
|
*
|
|
* at the end of a row 'code' is removed from codeset. Otherwise
|
|
* (rearranging the above):
|
|
*
|
|
* distance - window <= code_distance[code]
|
|
*
|
|
* and so the distances of the still active codes can be reduced:
|
|
*
|
|
* code_distance[code] -= distance-window
|
|
* distance = window
|
|
*
|
|
* This prevents any wrap of 'distance' on a row which is shorter than
|
|
* 2^32-window.
|
|
*
|
|
* However when then row is 2^32-window or more bytes long (the row can be up
|
|
* to just under 2^34 bytes long) this algorithm doesn't work; 'distance'
|
|
* will overflow in the middle of the row and all codes are relevant. This
|
|
* is handled below simply by reseting the set of present codes at the start
|
|
* of the row and ignoring the overflow.
|
|
*/
|
|
unsigned int code_count; /* Number of distinct codes seen */
|
|
int png_level; /* Cached compression level */
|
|
png_uint_32 filter_select_max_width;
|
|
/* The maximum number of pixels which can be fitted in the window without
|
|
* filling the entire window (i.e. the maximum number that can be fitted
|
|
* in (window-1) bytes).
|
|
*/
|
|
png_uint_32 sum_bias[PNG_FILTER_VALUE_LAST];
|
|
/* For each filter a measure of its cost in the filter sum calculation.
|
|
* This allows filter selection based on the sum-of-absolute-dfferences
|
|
* method to be biased to favour particular filters. There was no such
|
|
* bias before 1.7 and the filter byte was ignored.
|
|
*/
|
|
png_uint_32 distance; /* Distance from beginning */
|
|
png_codeset codeset; /* Set of seen codes */
|
|
png_uint_32 code_distance[256]; /* Distance at last occurence */
|
|
} filter_selector;
|
|
|
|
static const filter_selector *
|
|
png_start_filter_select(png_zlib_statep ps, unsigned int bpp)
|
|
{
|
|
# define png_ptr ps_png_ptr(ps)
|
|
filter_selector *fs = ps->selector;
|
|
|
|
if (fs == NULL)
|
|
{
|
|
fs = png_voidcast(filter_selector*, png_malloc_base(png_ptr, sizeof *fs));
|
|
|
|
if (fs != NULL)
|
|
{
|
|
png_uint_32 window = ps->filter_select_window;
|
|
fs->png_level = pz_get(ps, IDAT, png_level,
|
|
PNG_DEFAULT_COMPRESSION_LEVEL);
|
|
|
|
/* Delay initialize this here: */
|
|
if (window < 3U || window > PNG_FILTER_SELECT_WINDOW_MAX)
|
|
ps->filter_select_window = window = PNG_FILTER_SELECT_WINDOW_MAX;
|
|
|
|
fs->code_count = 0;
|
|
|
|
switch (fs->png_level)
|
|
{
|
|
default:
|
|
/* TODO: investigate other settings */
|
|
{
|
|
unsigned int f;
|
|
|
|
for (f=0; f<PNG_FILTER_VALUE_LAST; ++f)
|
|
fs->sum_bias[f] = f;
|
|
}
|
|
ps->filter_select_threshold = 64U; /* 6bit RGB */
|
|
ps->filter_select_threshold2 = 50U; /* TODO: experiment! */
|
|
break;
|
|
|
|
case PNG_COMPRESSION_COMPAT: /* Legacy */
|
|
memset(fs->sum_bias, 0U, sizeof fs->sum_bias);
|
|
ps->filter_select_threshold = 1U; /* disabled */
|
|
ps->filter_select_threshold2 = 1U;
|
|
break;
|
|
}
|
|
|
|
/* This is the maximum row width, in pixels, of a row which fits and
|
|
* leaves 1 byte free in the window. For any bigger row filter
|
|
* selection ignores the previous rows.
|
|
*/
|
|
fs->filter_select_max_width = ((window-2U/*filter+last byte*/)*8U)/bpp;
|
|
fs->distance = 0U;
|
|
memset(fs->codeset, 0U, sizeof fs->codeset);
|
|
/* fs->code_distance is left uninitialized because fs->codeset says
|
|
* whether or not each entry has been initialized.
|
|
*/
|
|
ps->selector = fs;
|
|
}
|
|
|
|
else
|
|
ps->write_row_size = 0U; /* OOM */
|
|
}
|
|
# undef png_ptr
|
|
|
|
return fs;
|
|
}
|
|
|
|
typedef struct
|
|
{
|
|
/* Per-filter data. This remains separate from the above until the filter
|
|
* selection has been made. It reflects the above however the codeset only
|
|
* records codes present in this row.
|
|
*
|
|
* The 'sum' fields are the sum of the absolute deviation of each code from
|
|
* 0, the algorithm from 1.6 and earlier. In other words:
|
|
*
|
|
* if (code >= 128)
|
|
* sum += code;
|
|
* else
|
|
* sum += 256-code;
|
|
*/
|
|
unsigned int code_count; /* Number of distinct codes seen in row */
|
|
unsigned int new_code_count; /* Number of new codes seen in row */
|
|
png_uint_32 sum_low; /* Low 31 bits of code sum */
|
|
png_uint_32 sum_high; /* High 32 bits of code sum */
|
|
png_codeset codeset; /* Set of codes seen in this row */
|
|
png_uint_32 code_distance[256]; /* Distance at last occurence in this row */
|
|
} filter_data;
|
|
|
|
static void
|
|
filter_data_init(filter_data *fd, png_uint_32 distance, unsigned int filter,
|
|
unsigned int code_is_set, png_uint_32 bias)
|
|
{
|
|
fd->code_count = 1U;
|
|
fd->new_code_count = !code_is_set;
|
|
fd->sum_low = bias;
|
|
fd->sum_high = 0U;
|
|
memset(&fd->codeset, 0U, sizeof fd->codeset);
|
|
PNG_CODE_SET(fd->codeset, filter);
|
|
fd->code_distance[filter] = distance;
|
|
}
|
|
|
|
static void
|
|
add_code(const filter_selector *fs, filter_data *fd, png_uint_32 distance,
|
|
unsigned int code)
|
|
{
|
|
if (!PNG_CODE_IS_SET(fd->codeset, code))
|
|
{
|
|
PNG_CODE_SET(fd->codeset, code);
|
|
++(fd->code_count);
|
|
fd->code_distance[code] = distance;
|
|
if (!PNG_CODE_IS_SET(fs->codeset, code))
|
|
++(fd->new_code_count);
|
|
}
|
|
|
|
{
|
|
png_uint_32 low = fd->sum_low;
|
|
|
|
if (code <= 128U)
|
|
low += code;
|
|
|
|
else
|
|
low += 256U-code;
|
|
|
|
/* Handle overflow into the top bit: */
|
|
if (low & 0x80000000U)
|
|
fd->sum_low = low & 0x7FFFFFFFU, ++fd->sum_high;
|
|
|
|
else
|
|
fd->sum_low = low;
|
|
}
|
|
}
|
|
|
|
static png_byte
|
|
filter_data_select(png_zlib_statep ps, filter_data fd[PNG_FILTER_VALUE_LAST],
|
|
unsigned int filter, png_uint_32 distance, png_uint_32 w)
|
|
{
|
|
# define png_ptr ps_png_ptr(ps)
|
|
/* Choose how to do this depending on the row and window size. */
|
|
filter_selector *fs = ps->selector;
|
|
png_uint_32 window = ps->filter_select_window;
|
|
|
|
affirm(fs != NULL);
|
|
|
|
/* Check the width against the maximum number of pixels that can fit in a
|
|
* window without filling it:
|
|
*/
|
|
if (w > fs->filter_select_max_width)
|
|
{
|
|
/* The cache is not used */
|
|
fs->distance = 0U; /* for next row */
|
|
fs->code_count = 0U;
|
|
memset(fs->codeset, 0U, sizeof fs->codeset);
|
|
}
|
|
|
|
else
|
|
{
|
|
/* Merge the two code sets, discounting codes that last occurred before
|
|
* the start of the window.
|
|
*/
|
|
png_uint_32 adjust, code_count;
|
|
unsigned int code;
|
|
|
|
/* filter_selector::distance is the distance of the first byte in the row
|
|
* (the filter byte), but 'distance' can wrap on long rows. The above
|
|
* test is meant to exclude the wrap case by excluding any case where the
|
|
* row has as many bytes as the window, so:
|
|
*/
|
|
affirm(distance > fs->distance && distance - fs->distance < window);
|
|
|
|
/* Set 'adjust' to the current distance of the start of the window. I.e:
|
|
*
|
|
* +---------------+--------+
|
|
* | before window | window | future data
|
|
* +---------------+--------+
|
|
* A A
|
|
* | |
|
|
* adjust + + distance
|
|
*
|
|
* If the window isn't full yet 'adjust' will be zero, otherwise all the
|
|
* distances will be reduced by 'adjust' so that the first byte of the
|
|
* window has distance 0.
|
|
*/
|
|
if (distance > window)
|
|
adjust = distance-window;
|
|
|
|
else
|
|
adjust = 0;
|
|
|
|
|
|
/* This may be decreased below if some old codes only occured before the
|
|
* start of the window.
|
|
*/
|
|
code_count = fs->code_count + fd->new_code_count;
|
|
|
|
for (code=0U; code<256U; ++code)
|
|
{
|
|
if (PNG_CODE_IS_SET(fd[filter].codeset, code))
|
|
{
|
|
PNG_CODE_SET(fs->codeset, code);
|
|
debug(fd[filter].code_distance[code] >= adjust);
|
|
fs->code_distance[code] = fd[filter].code_distance[code] - adjust;
|
|
}
|
|
|
|
else if (PNG_CODE_IS_SET(fs->codeset, code) && adjust > 0)
|
|
{
|
|
/* The code did not occur in this row, the old distance may now be
|
|
* outside the window (because adjust is non-zero).
|
|
*/
|
|
const png_uint_32 d = fs->code_distance[code];
|
|
|
|
if (d >= adjust)
|
|
fs->code_distance[code] = d-adjust;
|
|
|
|
else
|
|
PNG_CODE_CLEAR(fs->codeset, code), --code_count;
|
|
}
|
|
}
|
|
|
|
fs->code_count = code_count;
|
|
fs->distance = distance - adjust; /* I.e. either distance or window! */
|
|
}
|
|
|
|
return ps->filters = PNG_BYTE(filter);
|
|
# undef png_ptr
|
|
}
|
|
|
|
static png_byte
|
|
select_filter(png_zlib_statep ps, png_const_bytep row,
|
|
png_const_bytep prev, unsigned int bpp, png_uint_32 width, int start_of_pass)
|
|
/* Select a filter from the list provided by png_write_start_row. */
|
|
{
|
|
png_byte filters = png_write_start_row(ps, start_of_pass, prev == NULL);
|
|
|
|
# define png_ptr ps_png_ptr(ps)
|
|
if (filters >= PNG_FILTER_NONE) /* multiple filters to test */
|
|
{
|
|
const png_uint_32 max_pixels = ps->row_buffer_max_pixels;
|
|
const png_uint_32 block_pixels = ps->row_buffer_max_aligned_pixels;
|
|
const filter_selector *fs = ps->selector;
|
|
png_uint_32 pixels_to_go = width;
|
|
png_uint_32 distance;
|
|
unsigned int bits_at_end = 0U;
|
|
png_byte prev_pixels[4*2*2]; /* 2 pixels up to 4x2-bytes each */
|
|
filter_data fd[PNG_FILTER_VALUE_LAST];
|
|
|
|
debug((filters & (filters-1)) != 0U); /* Expect more than one bit! */
|
|
|
|
if (fs == NULL)
|
|
{
|
|
/* Delay initialize with a quiet OOM handler */
|
|
fs = png_start_filter_select(ps, bpp);
|
|
if (fs == NULL)
|
|
{
|
|
ps->filters = PNG_FILTER_VALUE_NONE;
|
|
return PNG_FILTER_VALUE_NONE;
|
|
}
|
|
}
|
|
|
|
/* If PNG_FILTER_NONE is in the list check it first. */
|
|
if (filters & PNG_FILTER_NONE)
|
|
{
|
|
png_const_bytep rp = row;
|
|
png_uint_32 w = width;
|
|
|
|
distance = fs->distance;
|
|
filter_data_init(fd+PNG_FILTER_VALUE_NONE, distance++,
|
|
PNG_FILTER_VALUE_NONE,
|
|
PNG_CODE_IS_SET(fs->codeset, PNG_FILTER_VALUE_NONE),
|
|
fs->sum_bias[PNG_FILTER_VALUE_NONE]);
|
|
|
|
if (bpp >= 8) /* complete bytes */
|
|
{
|
|
const unsigned int bytes = bpp/8U;
|
|
|
|
while (w > 0)
|
|
{
|
|
unsigned int b;
|
|
for (b=0; b<bytes; ++b)
|
|
add_code(fs, fd+PNG_FILTER_VALUE_NONE, distance++, *rp++);
|
|
--w;
|
|
}
|
|
}
|
|
|
|
else /* multiple pixels per byte */
|
|
{
|
|
const unsigned int ppb = 8U/bpp;
|
|
|
|
debug(ppb * bpp == 8U); /* Expect bpp to be a power of 2 */
|
|
|
|
while (w >= ppb)
|
|
{
|
|
add_code(fs, fd+PNG_FILTER_VALUE_NONE, distance++, *rp++);
|
|
w -= ppb;
|
|
}
|
|
|
|
if (w > 0) /* partial byte at end */
|
|
add_code(fs, fd+PNG_FILTER_VALUE_NONE, distance++,
|
|
*rp & (0xFFU >> (w*bpp) /* zero unused bits */));
|
|
}
|
|
|
|
/* For PNG data with a small number of codes it is worth skipping the
|
|
* filtering because it almost always increases the code count
|
|
* significantly. This is controlled by
|
|
* png_zlib_state::filter_select_threshold and causes an early return
|
|
* here.
|
|
*/
|
|
if (fd[PNG_FILTER_VALUE_NONE].new_code_count +
|
|
fs->code_count < ps->filter_select_threshold)
|
|
return filter_data_select(ps, fd, PNG_FILTER_VALUE_NONE, distance,
|
|
width);
|
|
} /* PNG_FILTER_NONE */
|
|
|
|
memset(prev_pixels, 0U, sizeof prev_pixels);
|
|
distance = fs->distance;
|
|
|
|
{
|
|
unsigned int i;
|
|
|
|
for (i=PNG_FILTER_VALUE_NONE+1U; i<PNG_FILTER_VALUE_LAST; ++i)
|
|
if (PNG_FILTER_MASK(i) & filters)
|
|
filter_data_init(fd+i, distance, i,
|
|
PNG_CODE_IS_SET(fs->codeset, i), fs->sum_bias[i]);
|
|
}
|
|
|
|
++distance;
|
|
|
|
while (pixels_to_go || bits_at_end)
|
|
{
|
|
unsigned int bits, i;
|
|
union
|
|
{
|
|
PNG_ROW_BUFFER_ALIGN_TYPE force_buffer_alignment;
|
|
png_byte row[4][PNG_ROW_BUFFER_SIZE];
|
|
} filtered;
|
|
union
|
|
{
|
|
PNG_ROW_BUFFER_ALIGN_TYPE force_buffer_alignment;
|
|
png_byte byte;
|
|
} last;
|
|
|
|
if (pixels_to_go)
|
|
{
|
|
if (pixels_to_go > max_pixels)
|
|
{
|
|
/* Maintain alignment by consuming on block_pixels at once */
|
|
bits = block_pixels * bpp;
|
|
pixels_to_go -= block_pixels; /* May be 0 */
|
|
}
|
|
|
|
else
|
|
{
|
|
bits = pixels_to_go * bpp;
|
|
bits_at_end = bits & 0x7U;
|
|
bits -= bits_at_end;
|
|
pixels_to_go = 0U; /* +bits_at_end */
|
|
}
|
|
}
|
|
|
|
else /* incomplete byte at the end of the pixel */
|
|
{
|
|
/* Make sure the unused bits are cleared (to zero, although this is
|
|
* an arbitrary choice):
|
|
*/
|
|
last.byte = PNG_BYTE(*row & ~(0xFFU >> bits_at_end));
|
|
row = &last.byte;
|
|
bits = bits_at_end;
|
|
bits_at_end = 0U;
|
|
}
|
|
|
|
filter_block(prev, prev_pixels, row, bits, bpp,
|
|
filtered.row[0/*sub*/], filtered.row[1/*up*/],
|
|
filtered.row[2/*avg*/], filtered.row[3/*Paeth*/]);
|
|
|
|
/* A block of (bits+7)/8 bytes is now available to process. */
|
|
for (i=0; 8U*i < bits; ++i, ++distance)
|
|
{
|
|
unsigned int f;
|
|
|
|
for (f=PNG_FILTER_VALUE_NONE+1U; f<PNG_FILTER_VALUE_LAST; ++f)
|
|
if (PNG_FILTER_MASK(f) & filters)
|
|
add_code(fs, fd+f, distance, filtered.row[f-1U][i]);
|
|
}
|
|
|
|
if (prev != NULL)
|
|
prev += bits >> 3;
|
|
|
|
row += bits >> 3;
|
|
}
|
|
|
|
/* Now look at the candidate filters, including 'none' and select the
|
|
* best. We know that 'none' increases the code count beyond the
|
|
* threshold, so if the old code count is below the threshold and there is
|
|
* a filter which does not increase the code count select it; doing so
|
|
* should do no harm to the overall compression.
|
|
*/
|
|
if (fs->code_count < ps->filter_select_threshold)
|
|
{
|
|
unsigned int f, min_new_count = 257U, min_f = PNG_FILTER_VALUE_NONE;
|
|
|
|
for (f=PNG_FILTER_VALUE_NONE+1U; f<PNG_FILTER_VALUE_LAST; ++f)
|
|
if ((PNG_FILTER_MASK(f) & filters) != 0)
|
|
{
|
|
unsigned int new_code_count = fd[f].new_code_count;
|
|
|
|
if (new_code_count == 0U)
|
|
return filter_data_select(ps, fd, f, distance, width);
|
|
|
|
else if (new_code_count < min_new_count)
|
|
min_new_count = new_code_count, min_f = f;
|
|
}
|
|
|
|
/* Use the second threshold to decide whether to select the best filter
|
|
* on this basis alone:
|
|
*/
|
|
if (min_f != PNG_FILTER_VALUE_NONE &&
|
|
fs->code_count + min_new_count < ps->filter_select_threshold2)
|
|
return filter_data_select(ps, fd, min_f, distance, width);
|
|
}
|
|
|
|
/* Now fall back to the libpng 1.6 and earlier algorithm. This favours
|
|
* the filter which produces least deviation in the codes from 0. When
|
|
* this works it does so by reducing the distribution of code values. The
|
|
* filters implicitly encode the difference between a predictor based on
|
|
* adjacent values, the assumption is that this will result in values
|
|
* close to 0.
|
|
*/
|
|
{
|
|
png_uint_32 high = -1;
|
|
png_uint_32 low = -1;
|
|
unsigned int min_f = 0 /*unset, but safe*/;
|
|
unsigned int f;
|
|
|
|
for (f=PNG_FILTER_VALUE_NONE; f<PNG_FILTER_VALUE_LAST; ++f)
|
|
if ((PNG_FILTER_MASK(f) & filters) != 0 &&
|
|
(fd[f].sum_high < high ||
|
|
(fd[f].sum_high == high && fd[f].sum_low < low)))
|
|
{
|
|
high = fd[f].sum_high;
|
|
low = fd[f].sum_low;
|
|
|
|
if (low & 0x80000000U)
|
|
{
|
|
low &= 0x7FFFFFFFU, --high;
|
|
if (high & 0x80000000U)
|
|
low = high = 0U;
|
|
}
|
|
|
|
min_f = f;
|
|
}
|
|
|
|
return filter_data_select(ps, fd, min_f, distance, width);
|
|
}
|
|
}
|
|
|
|
debug(filters < PNG_FILTER_VALUE_LAST);
|
|
return ps->filters = filters;
|
|
# undef png_ptr
|
|
}
|
|
#else /* !SELECT_FILTER */
|
|
/* Filter selection not being done, just call png_write_start_row: */
|
|
# define select_filter(ps, rp, pp, bpp, width, start_of_pass)\
|
|
png_write_start_row((ps), (start_of_pass), (pp) == NULL)
|
|
#endif /* !SELECT_FILTER */
|
|
|
|
/* This is the common function to write multiple rows of PNG data. The data is
|
|
* in the relevant PNG format but has had no filtering done.
|
|
*/
|
|
void /* PRIVATE */
|
|
png_write_png_rows(png_structrp png_ptr, png_const_bytep *rows,
|
|
png_uint_32 num_rows)
|
|
{
|
|
const png_zlib_statep ps = png_ptr->zlib_state;
|
|
const unsigned int bpp = png_ptr->row_output_pixel_depth;
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
png_const_bytep previous_row = ps->previous_write_row;
|
|
# else /* !WRITE_FILTER */
|
|
/* These are constant in the no-filer case: */
|
|
const png_byte filter = PNG_FILTER_VALUE_NONE;
|
|
const png_uint_32 max_pixels = ps->zlib_max_pixels;
|
|
const png_uint_32 block_pixels = ps->zlib_max_aligned_pixels;
|
|
# endif /* !WRITE_FILTER */
|
|
/* Write the given rows handling the png_compress_IDAT argument limitations
|
|
* (uInt) and any valid row width.
|
|
*/
|
|
png_uint_32 last_row_in_pass = 0U; /* Actual last, not last+1! */
|
|
png_uint_32 pixels_in_pass = 0U;
|
|
unsigned int first_row_in_pass = 0U; /* For do_interlace */
|
|
unsigned int pixels_at_end = 0U; /* for a partial byte at the end */
|
|
unsigned int base_info_flags = png_row_end;
|
|
int pass = -1; /* Invalid: force calculation first time round */
|
|
|
|
debug(png_ptr->row_output_pixel_depth == PNG_PIXEL_DEPTH(*png_ptr));
|
|
|
|
while (num_rows-- > 0U)
|
|
{
|
|
if (png_ptr->pass != pass)
|
|
{
|
|
/* Recalcuate the row bytes and partial bits */
|
|
pass = png_ptr->pass;
|
|
pixels_in_pass = png_ptr->width;
|
|
|
|
if (png_ptr->interlaced == PNG_INTERLACE_NONE)
|
|
{
|
|
debug(pass == 0);
|
|
last_row_in_pass = png_ptr->height - 1U;
|
|
base_info_flags |= png_pass_last; /* there is only one */
|
|
}
|
|
|
|
else
|
|
{
|
|
const png_uint_32 height = png_ptr->height;
|
|
|
|
last_row_in_pass = PNG_PASS_ROWS(height, pass);
|
|
debug(pass >= 0 && pass < 7);
|
|
|
|
# ifdef PNG_WRITE_INTERLACING_SUPPORTED
|
|
if (png_ptr->do_interlace)
|
|
{
|
|
/* libpng is doing the interlace handling, the row number is
|
|
* actually the row in the image.
|
|
*
|
|
* This overflows when the PNG height is such that the are no
|
|
* rows in this pass. This does not matter; because there are
|
|
* no rows the value doesn't get used.
|
|
*/
|
|
last_row_in_pass =
|
|
PNG_ROW_FROM_PASS_ROW(last_row_in_pass-1U, pass);
|
|
first_row_in_pass = PNG_PASS_START_ROW(pass);
|
|
}
|
|
|
|
else /* Application handles the interlace */
|
|
# endif /* WRITE_INTERLACING */
|
|
{
|
|
/* The row does exist, so this works without checking the column
|
|
* count.
|
|
*/
|
|
debug(last_row_in_pass > 0U);
|
|
last_row_in_pass -= 1U;
|
|
}
|
|
|
|
if (pass == PNG_LAST_PASS(pixels_in_pass/*PNG width*/, height))
|
|
base_info_flags |= png_pass_last;
|
|
|
|
/* Finally, adjust pixels_in_pass for the interlacing (skip the
|
|
* final pass; it is full width).
|
|
*/
|
|
if (pass < 6)
|
|
pixels_in_pass = PNG_PASS_COLS(pixels_in_pass, pass);
|
|
}
|
|
|
|
/* Mask out the bits in a partial byte. */
|
|
pixels_at_end = pixels_in_pass & PNG_ADDOF(bpp);
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* Reset the previous_row pointer correctly; NULL at the start of
|
|
* the pass. If row_number is not 0 then a previous write_rows was
|
|
* interrupted in mid-pass and any required buffer should be in
|
|
* previous_write_row (set in the initializer).
|
|
*/
|
|
if (png_ptr->row_number == first_row_in_pass)
|
|
previous_row = NULL;
|
|
# endif /* WRITE_FILTER */
|
|
}
|
|
|
|
# ifdef PNG_WRITE_INTERLACING_SUPPORTED
|
|
/* When libpng is handling the interlace we see rows that must be
|
|
* skipped.
|
|
*/
|
|
if (!png_ptr->do_interlace ||
|
|
PNG_ROW_IN_INTERLACE_PASS(png_ptr->row_number, pass))
|
|
# endif /* WRITE_INTERLACING */
|
|
{
|
|
const unsigned int row_info_flags = base_info_flags |
|
|
(png_ptr->row_number ==
|
|
first_row_in_pass ? png_pass_first_row : 0) |
|
|
(png_ptr->row_number == last_row_in_pass ? png_pass_last_row : 0);
|
|
const int flush = row_flush(ps, row_info_flags);
|
|
png_const_bytep row = *rows;
|
|
png_uint_32 pixels_to_go = pixels_in_pass;
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* The filter can change each time round. Call png_write_start_row
|
|
* to resolve any changes. Note that when this function is used to
|
|
* do filter selection from png_write_png_data on the first row
|
|
* png_write_start_row will get called twice.
|
|
*/
|
|
const png_byte filter = select_filter(ps, row, previous_row, bpp,
|
|
pixels_in_pass, png_ptr->row_number == first_row_in_pass);
|
|
const png_uint_32 max_pixels = filter == PNG_FILTER_VALUE_NONE ?
|
|
ps->zlib_max_pixels : ps->row_buffer_max_pixels;
|
|
const png_uint_32 block_pixels = filter == PNG_FILTER_VALUE_NONE ?
|
|
ps->zlib_max_aligned_pixels : ps->row_buffer_max_aligned_pixels;
|
|
|
|
/* The row handling uses png_compress_IDAT directly if there is no
|
|
* filter to be applied, otherwise it uses filter_row.
|
|
*/
|
|
if (filter != PNG_FILTER_VALUE_NONE)
|
|
{
|
|
int start_of_row = 1;
|
|
png_byte prev_pixels[4*2*2]; /* 2 pixels up to 4x2-bytes each */
|
|
|
|
memset(prev_pixels, 0U, sizeof prev_pixels);
|
|
|
|
while (pixels_to_go > max_pixels)
|
|
{
|
|
/* Write a block at once to maintain alignment */
|
|
filter_row(png_ptr, previous_row, prev_pixels, row,
|
|
bpp * block_pixels, bpp, filter, start_of_row,
|
|
Z_NO_FLUSH);
|
|
|
|
if (previous_row != NULL)
|
|
previous_row += (block_pixels * bpp) >> 3;
|
|
|
|
row += (block_pixels * bpp) >> 3;
|
|
pixels_to_go -= block_pixels;
|
|
start_of_row = 0;
|
|
}
|
|
|
|
/* The filter code handles the partial byte at the end correctly,
|
|
* so this is all that is required:
|
|
*/
|
|
if (pixels_to_go > 0)
|
|
filter_row(png_ptr, previous_row, prev_pixels, row,
|
|
bpp * pixels_to_go, bpp, filter, start_of_row, flush);
|
|
}
|
|
|
|
else
|
|
# endif /* WRITE_FILTER */
|
|
|
|
{
|
|
/* The no-filter case. */
|
|
const uInt block_bytes = (uInt)/*SAFE*/(
|
|
bpp <= 8U ?
|
|
block_pixels >> PNG_SHIFTOF(bpp) :
|
|
block_pixels * (bpp >> 3));
|
|
|
|
/* png_write_start_IDAT guarantees this, but double check for
|
|
* overflow above in debug:
|
|
*/
|
|
debug((block_bytes & (PNG_ROW_BUFFER_BYTE_ALIGN-1U)) == 0U);
|
|
|
|
/* The filter has to be written here: */
|
|
png_compress_IDAT(png_ptr, &filter, 1U/*len*/, Z_NO_FLUSH);
|
|
|
|
/* Process blocks of pixels up to the limit. */
|
|
while (pixels_to_go > max_pixels)
|
|
{
|
|
png_compress_IDAT(png_ptr, row, block_bytes, Z_NO_FLUSH);
|
|
row += block_bytes;
|
|
pixels_to_go -= block_pixels;
|
|
}
|
|
|
|
/* Now compress the remainder; pixels_to_go <= max_pixels so it will
|
|
* fit in a uInt.
|
|
*/
|
|
{
|
|
const png_uint_32 remainder =
|
|
bpp <= 8U
|
|
? (pixels_to_go-pixels_at_end) >> PNG_SHIFTOF(bpp)
|
|
: (pixels_to_go-pixels_at_end) * (bpp >> 3);
|
|
|
|
if (remainder > 0U)
|
|
png_compress_IDAT(png_ptr, row, remainder,
|
|
pixels_at_end > 0U ? Z_NO_FLUSH : flush);
|
|
|
|
else
|
|
debug(pixels_at_end > 0U);
|
|
|
|
if (pixels_at_end > 0U)
|
|
{
|
|
/* There is a final partial byte. This is PNG format so the
|
|
* left-most bits are the most significant.
|
|
*/
|
|
const png_byte last = PNG_BYTE(row[remainder] &
|
|
~(0xFFU >> (pixels_at_end * bpp)));
|
|
|
|
png_compress_IDAT(png_ptr, &last, 1U, flush);
|
|
}
|
|
}
|
|
}
|
|
|
|
png_write_end_row(png_ptr, flush);
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
previous_row = *rows;
|
|
# endif /* WRITE_FILTER */
|
|
# undef HANDLE
|
|
} /* row in pass */
|
|
|
|
# ifdef PNG_WRITE_INTERLACING_SUPPORTED
|
|
else /* row not in pass; just skip it */
|
|
{
|
|
if (++png_ptr->row_number >= png_ptr->height)
|
|
{
|
|
debug(png_ptr->row_number == png_ptr->height);
|
|
|
|
png_ptr->row_number = 0U;
|
|
png_ptr->pass = 0x7U & (pass+1U);
|
|
}
|
|
}
|
|
# endif /* WRITE_INTERLACING */
|
|
|
|
++rows;
|
|
} /* while num_rows */
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* previous_row must be copied back unless we don't need it because the
|
|
* next row is the first one in the pass (this relies on png_write_end_row
|
|
* setting row_number to 0 at the end!)
|
|
*/
|
|
if (png_ptr->row_number != 0U && previous_row != NULL && SAVE_ROW(ps) &&
|
|
ps->previous_write_row != previous_row/*all rows skipped*/)
|
|
{
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
/* We might be able to avoid any copy. */
|
|
if (ps->current_write_row == previous_row)
|
|
{
|
|
png_bytep old = ps->previous_write_row;
|
|
ps->previous_write_row = ps->current_write_row;
|
|
ps->current_write_row = old; /* may be NULL */
|
|
}
|
|
|
|
else
|
|
# endif /* SELECT_FILTER */
|
|
|
|
if (ps->previous_write_row != NULL)
|
|
memcpy(ps->previous_write_row, previous_row,
|
|
png_calc_rowbytes(png_ptr, bpp, pixels_in_pass));
|
|
|
|
else
|
|
ps->previous_write_row = allocate_row(png_ptr, previous_row,
|
|
png_calc_rowbytes(png_ptr, bpp, pixels_in_pass));
|
|
}
|
|
# endif /* WRITE_FILTER */
|
|
}
|
|
|
|
#ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
/* This filters the row, chooses which filter to use, if it has not already
|
|
* been specified by the application, and then writes the row out with the
|
|
* chosen filter.
|
|
*/
|
|
static void
|
|
write_png_data(png_structrp png_ptr, png_const_bytep prev_row,
|
|
png_bytep prev_pixels, png_const_bytep unfiltered_row, png_uint_32 x,
|
|
unsigned int row_bits, unsigned int row_info_flags)
|
|
/* This filters the row appropriately and returns an updated prev_row
|
|
* (updated for 'x').
|
|
*/
|
|
{
|
|
const png_zlib_statep ps = png_ptr->zlib_state;
|
|
const unsigned int bpp = png_ptr->row_output_pixel_depth;
|
|
const int flush = row_flush(ps, row_info_flags);
|
|
const png_byte filter = ps->filters; /* just one */
|
|
|
|
/* These invariants are expected from the caller: */
|
|
affirm(row_bits <= 8U*PNG_ROW_BUFFER_SIZE);
|
|
debug(filter < PNG_FILTER_VALUE_LAST/*sic: last+1*/);
|
|
|
|
/* Now choose the correct filter implementation according to the number of
|
|
* filters in the filters_to_try list. The prev_row parameter is made
|
|
* NULL on the first row because it is uninitialized at that point.
|
|
*/
|
|
if (filter == PNG_FILTER_VALUE_NONE)
|
|
write_unfiltered_rowbits(png_ptr, unfiltered_row, row_bits,
|
|
x == 0 ? PNG_FILTER_VALUE_NONE : PNG_FILTER_VALUE_LAST, flush);
|
|
|
|
else
|
|
filter_row(png_ptr,
|
|
(row_info_flags & png_pass_first_row) ? NULL : prev_row,
|
|
prev_pixels, unfiltered_row, row_bits, bpp, filter, x == 0, flush);
|
|
|
|
/* Handle end of row: */
|
|
if ((row_info_flags & png_row_end) != 0)
|
|
png_write_end_row(png_ptr, flush);
|
|
}
|
|
|
|
void /* PRIVATE */
|
|
png_write_png_data(png_structrp png_ptr, png_bytep prev_pixels,
|
|
png_const_bytep unfiltered_row, png_uint_32 x,
|
|
unsigned int width/*pixels*/, unsigned int row_info_flags)
|
|
{
|
|
const png_zlib_statep ps = png_ptr->zlib_state;
|
|
|
|
affirm(ps != NULL);
|
|
|
|
{
|
|
const unsigned int bpp = png_ptr->row_output_pixel_depth;
|
|
const unsigned int row_bits = width * bpp;
|
|
png_bytep prev_row = ps->previous_write_row;
|
|
|
|
debug(bpp <= 64U && width <= 65535U &&
|
|
width < 65535U/bpp); /* Expensive: only matters on 16-bit */
|
|
|
|
/* This is called once before starting a new row here, but below it is
|
|
* only called once between starting a new list of rows.
|
|
*/
|
|
if (x == 0)
|
|
png_write_start_row(ps, (row_info_flags & png_pass_first_row) != 0,
|
|
prev_row == NULL);
|
|
|
|
/* If filter selection is required the filter will have at least one mask
|
|
* bit set.
|
|
*/
|
|
# ifdef PNG_SELECT_FILTER_SUPPORTED
|
|
if (ps->filters >= PNG_FILTER_NONE/*lowest mask bit*/)
|
|
{
|
|
/* If the entire row is passed in the input process it via
|
|
* immediately, otherwise the row must be buffered for later
|
|
* analysis.
|
|
*/
|
|
png_const_bytep row;
|
|
|
|
if (x > 0 || (row_info_flags & png_row_end) == 0)
|
|
{
|
|
/* The row must be saved for later. */
|
|
png_bytep buffer = ps->current_write_row;
|
|
|
|
/* png_write_start row should always check this: */
|
|
debug(ps->write_row_size > 0U);
|
|
|
|
if (buffer != NULL)
|
|
memcpy(buffer + png_calc_rowbytes(png_ptr, bpp, x),
|
|
unfiltered_row, (row_bits + 7U) >> 3);
|
|
|
|
|
|
else if (x == 0U)
|
|
ps->current_write_row = buffer = allocate_row(png_ptr,
|
|
unfiltered_row, (row_bits + 7U) >> 3);
|
|
|
|
row = buffer;
|
|
}
|
|
|
|
else
|
|
row = unfiltered_row;
|
|
|
|
if (row != NULL) /* else out of memory */
|
|
{
|
|
/* At row end, process the save buffer. */
|
|
if ((row_info_flags & png_row_end) != 0)
|
|
png_write_png_rows(png_ptr, &row, 1U);
|
|
|
|
/* Early return to skip the single-filter code */
|
|
return;
|
|
}
|
|
|
|
/* Caching the row failed, so process the row using the lowest set
|
|
* filter. The allocation error should only ever happen at the
|
|
* start of the row. If this goes wrong the output will have been
|
|
* damaged.
|
|
*/
|
|
affirm(x == 0U);
|
|
}
|
|
# endif /* SELECT_FILTER */
|
|
|
|
/* prev_row is either NULL or the position in the previous row buffer */
|
|
if (prev_row != NULL && x > 0)
|
|
prev_row += png_calc_rowbytes(png_ptr, bpp, x);
|
|
|
|
/* This is the single filter case (no selection): */
|
|
write_png_data(png_ptr, prev_row, prev_pixels, unfiltered_row, x,
|
|
row_bits, row_info_flags);
|
|
|
|
/* Copy the current row into the previous row buffer, if available, unless
|
|
* this is the last row in the pass, when there is no point. Note that
|
|
* write_previous_row may have garbage in a partial byte at the end as a
|
|
* result of this memcpy.
|
|
*/
|
|
if (!(row_info_flags & png_pass_last_row) && SAVE_ROW(ps)) {
|
|
if (prev_row != NULL)
|
|
memcpy(prev_row, unfiltered_row, (row_bits + 7U) >> 3);
|
|
|
|
/* NOTE: if the application sets png_zlib_state::save_row in a callback
|
|
* it isn't possible to do the save until the next row. allocate_row
|
|
* handles OOM silently by turning off the save.
|
|
*/
|
|
else if (x == 0) /* can allocate the save buffer */
|
|
ps->previous_write_row =
|
|
allocate_row(png_ptr, unfiltered_row, (row_bits + 7U) >> 3);
|
|
}
|
|
}
|
|
}
|
|
#else /* !WRITE_FILTER */
|
|
void /* PRIVATE */
|
|
png_write_start_IDAT(png_structrp png_ptr)
|
|
{
|
|
png_zlib_statep ps = get_zlib_state(png_ptr);
|
|
|
|
/* Set up the IDAT compression state. Expect the state to have been released
|
|
* by the previous owner, but it doesn't much matter if there was an error.
|
|
* Note that the stream is not claimed yet.
|
|
*/
|
|
debug(png_ptr->zowner == 0U);
|
|
|
|
/* This sets the buffer limits and write_row_size, which is used below. */
|
|
png_zlib_state_set_buffer_limits(png_ptr, ps);
|
|
}
|
|
|
|
void /* PRIVATE */
|
|
png_write_png_data(png_structrp png_ptr, png_bytep prev_pixels,
|
|
png_const_bytep unfiltered_row, png_uint_32 x,
|
|
unsigned int width/*pixels*/, unsigned int row_info_flags)
|
|
{
|
|
const unsigned int bpp = png_ptr->row_output_pixel_depth;
|
|
int flush;
|
|
png_uint_32 row_bits;
|
|
|
|
row_bits = width;
|
|
row_bits *= bpp;
|
|
/* These invariants are expected from the caller: */
|
|
affirm(width < 65536U && bpp <= 64U && width < 65536U/bpp &&
|
|
row_bits <= 8U*PNG_ROW_BUFFER_SIZE);
|
|
|
|
affirm(png_ptr->zlib_state != NULL);
|
|
flush = row_flush(png_ptr->zlib_state, row_info_flags);
|
|
|
|
write_unfiltered_rowbits(png_ptr, unfiltered_row, row_bits,
|
|
x == 0 ? PNG_FILTER_VALUE_NONE : PNG_FILTER_VALUE_LAST, flush);
|
|
|
|
PNG_UNUSED(prev_pixels);
|
|
|
|
/* Handle end of row: */
|
|
if ((row_info_flags & png_row_end) != 0)
|
|
png_write_end_row(png_ptr, flush);
|
|
}
|
|
#endif /* !WRITE_FILTER */
|
|
|
|
png_int_32 /* PRIVATE */
|
|
png_write_setting(png_structrp png_ptr, png_uint_32 setting,
|
|
png_uint_32 parameter, png_int_32 value)
|
|
{
|
|
/* Caller checks the arguments for basic validity */
|
|
int only_get = (setting & PNG_SF_GET) != 0U;
|
|
|
|
setting &= ~PNG_SF_GET;
|
|
|
|
switch (setting)
|
|
{
|
|
/* Settings in png_struct: */
|
|
case PNG_SW_IDAT_size:
|
|
if (parameter > 0 && parameter <= PNG_UINT_31_MAX)
|
|
{
|
|
if (!only_get)
|
|
png_ptr->IDAT_size = parameter;
|
|
|
|
return 0; /* set ok */
|
|
}
|
|
|
|
else
|
|
return PNG_EINVAL;
|
|
|
|
/* Settings in zlib_state: */
|
|
case PNG_SW_COMPRESS_png_level:
|
|
return compression_setting(png_ptr, parameter, png_level, value,
|
|
only_get);
|
|
|
|
# ifdef PNG_WRITE_CUSTOMIZE_COMPRESSION_SUPPORTED
|
|
case PNG_SW_COMPRESS_zlib_level:
|
|
return compression_setting(png_ptr, parameter, level, value,
|
|
only_get);
|
|
|
|
case PNG_SW_COMPRESS_windowBits:
|
|
return compression_setting(png_ptr, parameter, windowBits, value,
|
|
only_get);
|
|
|
|
case PNG_SW_COMPRESS_memLevel:
|
|
return compression_setting(png_ptr, parameter, memLevel, value,
|
|
only_get);
|
|
|
|
case PNG_SW_COMPRESS_strategy:
|
|
return compression_setting(png_ptr, parameter, strategy, value,
|
|
only_get);
|
|
|
|
case PNG_SW_COMPRESS_method:
|
|
if (value != 8) /* Only supported method */
|
|
return PNG_EINVAL;
|
|
return 8; /* old method */
|
|
# endif /* WRITE_CUSTOMIZE_COMPRESSION */
|
|
|
|
# ifdef PNG_WRITE_FILTER_SUPPORTED
|
|
case PNG_SW_COMPRESS_filters:
|
|
/* The method must match that in the IHDR: */
|
|
if (parameter == png_ptr->filter_method)
|
|
{
|
|
if (!only_get)
|
|
return set_filter(get_zlib_state(png_ptr), value);
|
|
|
|
else if (png_ptr->zlib_state != NULL &&
|
|
png_ptr->zlib_state->filter_mask != 0U/*unset*/)
|
|
return png_ptr->zlib_state->filter_mask;
|
|
|
|
else
|
|
return PNG_UNSET;
|
|
}
|
|
|
|
else /* Invalid filter method */
|
|
return PNG_EINVAL;
|
|
|
|
case PNG_SW_COMPRESS_row_buffers:
|
|
/* New in 1.7.0: direct control of the buffering. */
|
|
switch (parameter)
|
|
{
|
|
case 0:
|
|
if (!only_get)
|
|
get_zlib_state(png_ptr)->save_row = SAVE_ROW_OFF;
|
|
return 0;
|
|
|
|
case 1:
|
|
if (!only_get)
|
|
get_zlib_state(png_ptr)->save_row = SAVE_ROW_ON;
|
|
return 1;
|
|
|
|
default:
|
|
return PNG_ENOSYS; /* no support for bigger values */
|
|
}
|
|
# endif /* WRITE_FILTER */
|
|
|
|
# ifdef PNG_WRITE_FLUSH_SUPPORTED
|
|
case PNG_SW_FLUSH:
|
|
/* Set the automatic flush interval or 0 to turn flushing off */
|
|
if (!only_get)
|
|
get_zlib_state(png_ptr)->flush_dist =
|
|
value <= 0 ? 0xEFFFFFFFU : (png_uint_32)/*SAFE*/value;
|
|
|
|
return 0;
|
|
# endif /* WRITE_FLUSH */
|
|
|
|
# ifdef PNG_WRITE_CHECK_FOR_INVALID_INDEX_SUPPORTED
|
|
case PNG_SRW_CHECK_FOR_INVALID_INDEX:
|
|
/* The 'enabled' value is a FORTRAN style three-state: */
|
|
if (value > 0)
|
|
png_ptr->palette_index_check = PNG_PALETTE_CHECK_ON;
|
|
|
|
else if (value < 0)
|
|
png_ptr->palette_index_check = PNG_PALETTE_CHECK_OFF;
|
|
|
|
else
|
|
png_ptr->palette_index_check = PNG_PALETTE_CHECK_DEFAULT;
|
|
|
|
return 0;
|
|
# endif /* WRITE_CHECK_FOR_INVALID_INDEX */
|
|
|
|
# ifdef PNG_BENIGN_WRITE_ERRORS_SUPPORTED
|
|
case PNG_SRW_ERROR_HANDLING:
|
|
/* The parameter is a bit mask of what to set, the value is what to
|
|
* set it to. PNG_IDAT_ERRORS is ignored on write.
|
|
*/
|
|
if (value >= PNG_IGNORE && value <= PNG_ERROR &&
|
|
parameter <= PNG_ALL_ERRORS)
|
|
{
|
|
if ((parameter & PNG_BENIGN_ERRORS) != 0U)
|
|
png_ptr->benign_error_action = value & 0x3U;
|
|
|
|
if ((parameter & PNG_APP_WARNINGS) != 0U)
|
|
png_ptr->app_warning_action = value & 0x3U;
|
|
|
|
if ((parameter & PNG_APP_ERRORS) != 0U)
|
|
png_ptr->app_error_action = value & 0x3U;
|
|
|
|
return 0;
|
|
}
|
|
|
|
return PNG_EINVAL;
|
|
# endif /* BENIGN_WRITE_ERRORS */
|
|
|
|
default:
|
|
return PNG_ENOSYS; /* not supported (whatever it is) */
|
|
}
|
|
}
|
|
#endif /* WRITE */
|