[reactos.git] / reactos / dll / 3rdparty / libpng / docs / libpng-manual.txt
1 libpng-manual.txt - A description on how to use and modify libpng
3 libpng version 1.6.28 - January 5, 2017
4 Updated and distributed by Glenn Randers-Pehrson
5 <glennrp at users.sourceforge.net>
6 Copyright (c) 1998-2016 Glenn Randers-Pehrson
8 This document is released under the libpng license.
9 For conditions of distribution and use, see the disclaimer
10 and license in png.h
12 Based on:
14 libpng versions 0.97, January 1998, through 1.6.28 - January 5, 2017
15 Updated and distributed by Glenn Randers-Pehrson
16 Copyright (c) 1998-2016 Glenn Randers-Pehrson
18 libpng 1.0 beta 6 - version 0.96 - May 28, 1997
19 Updated and distributed by Andreas Dilger
20 Copyright (c) 1996, 1997 Andreas Dilger
22 libpng 1.0 beta 2 - version 0.88 - January 26, 1996
23 For conditions of distribution and use, see copyright
24 notice in png.h. Copyright (c) 1995, 1996 Guy Eric
25 Schalnat, Group 42, Inc.
27 Updated/rewritten per request in the libpng FAQ
28 Copyright (c) 1995, 1996 Frank J. T. Wojcik
29 December 18, 1995 & January 20, 1996
33 I. Introduction
34 II. Structures
35 III. Reading
36 IV. Writing
37 V. Simplified API
38 VI. Modifying/Customizing libpng
39 VII. MNG support
40 VIII. Changes to Libpng from version 0.88
41 IX. Changes to Libpng from version 1.0.x to 1.2.x
42 X. Changes to Libpng from version 1.0.x/1.2.x to 1.4.x
43 XI. Changes to Libpng from version 1.4.x to 1.5.x
44 XII. Changes to Libpng from version 1.5.x to 1.6.x
45 XIII. Detecting libpng
46 XIV. Source code repository
47 XV. Coding style
48 XVI. Y2K Compliance in libpng
50 I. Introduction
52 This file describes how to use and modify the PNG reference library
53 (known as libpng) for your own use. In addition to this
54 file, example.c is a good starting point for using the library, as
55 it is heavily commented and should include everything most people
56 will need. We assume that libpng is already installed; see the
57 INSTALL file for instructions on how to configure and install libpng.
59 For examples of libpng usage, see the files "example.c", "pngtest.c",
60 and the files in the "contrib" directory, all of which are included in
61 the libpng distribution.
63 Libpng was written as a companion to the PNG specification, as a way
64 of reducing the amount of time and effort it takes to support the PNG
65 file format in application programs.
67 The PNG specification (second edition), November 2003, is available as
68 a W3C Recommendation and as an ISO Standard (ISO/IEC 15948:2004 (E)) at
69 <http://www.w3.org/TR/2003/REC-PNG-20031110/
70 The W3C and ISO documents have identical technical content.
72 The PNG-1.2 specification is available at
73 <http://png-mng.sourceforge.net/pub/png/spec/1.2/>.
74 It is technically equivalent
75 to the PNG specification (second edition) but has some additional material.
77 The PNG-1.0 specification is available as RFC 2083
78 <http://png-mng.sourceforge.net/pub/png/spec/1.0/> and as a
79 W3C Recommendation <http://www.w3.org/TR/REC-png-961001>.
81 Some additional chunks are described in the special-purpose public chunks
82 documents at <http://www.libpng.org/pub/png/spec/register/>
84 Other information
85 about PNG, and the latest version of libpng, can be found at the PNG home
86 page, <http://www.libpng.org/pub/png/>.
88 Most users will not have to modify the library significantly; advanced
89 users may want to modify it more. All attempts were made to make it as
90 complete as possible, while keeping the code easy to understand.
91 Currently, this library only supports C. Support for other languages
92 is being considered.
94 Libpng has been designed to handle multiple sessions at one time,
95 to be easily modifiable, to be portable to the vast majority of
96 machines (ANSI, K&R, 16-, 32-, and 64-bit) available, and to be easy
97 to use. The ultimate goal of libpng is to promote the acceptance of
98 the PNG file format in whatever way possible. While there is still
99 work to be done (see the TODO file), libpng should cover the
100 majority of the needs of its users.
102 Libpng uses zlib for its compression and decompression of PNG files.
103 Further information about zlib, and the latest version of zlib, can
104 be found at the zlib home page, <http://zlib.net/>.
105 The zlib compression utility is a general purpose utility that is
106 useful for more than PNG files, and can be used without libpng.
107 See the documentation delivered with zlib for more details.
108 You can usually find the source files for the zlib utility wherever you
109 find the libpng source files.
111 Libpng is thread safe, provided the threads are using different
112 instances of the structures. Each thread should have its own
113 png_struct and png_info instances, and thus its own image.
114 Libpng does not protect itself against two threads using the
115 same instance of a structure.
117 II. Structures
119 There are two main structures that are important to libpng, png_struct
120 and png_info. Both are internal structures that are no longer exposed
121 in the libpng interface (as of libpng 1.5.0).
123 The png_info structure is designed to provide information about the
124 PNG file. At one time, the fields of png_info were intended to be
125 directly accessible to the user. However, this tended to cause problems
126 with applications using dynamically loaded libraries, and as a result
127 a set of interface functions for png_info (the png_get_*() and png_set_*()
128 functions) was developed, and direct access to the png_info fields was
129 deprecated..
131 The png_struct structure is the object used by the library to decode a
132 single image. As of 1.5.0 this structure is also not exposed.
134 Almost all libpng APIs require a pointer to a png_struct as the first argument.
135 Many (in particular the png_set and png_get APIs) also require a pointer
136 to png_info as the second argument. Some application visible macros
137 defined in png.h designed for basic data access (reading and writing
138 integers in the PNG format) don't take a png_info pointer, but it's almost
139 always safe to assume that a (png_struct*) has to be passed to call an API
140 function.
142 You can have more than one png_info structure associated with an image,
143 as illustrated in pngtest.c, one for information valid prior to the
144 IDAT chunks and another (called "end_info" below) for things after them.
146 The png.h header file is an invaluable reference for programming with libpng.
147 And while I'm on the topic, make sure you include the libpng header file:
149 #include <png.h>
151 and also (as of libpng-1.5.0) the zlib header file, if you need it:
153 #include <zlib.h>
155 Types
157 The png.h header file defines a number of integral types used by the
158 APIs. Most of these are fairly obvious; for example types corresponding
159 to integers of particular sizes and types for passing color values.
161 One exception is how non-integral numbers are handled. For application
162 convenience most APIs that take such numbers have C (double) arguments;
163 however, internally PNG, and libpng, use 32 bit signed integers and encode
164 the value by multiplying by 100,000. As of libpng 1.5.0 a convenience
165 macro PNG_FP_1 is defined in png.h along with a type (png_fixed_point)
166 which is simply (png_int_32).
168 All APIs that take (double) arguments also have a matching API that
169 takes the corresponding fixed point integer arguments. The fixed point
170 API has the same name as the floating point one with "_fixed" appended.
171 The actual range of values permitted in the APIs is frequently less than
172 the full range of (png_fixed_point) (-21474 to +21474). When APIs require
173 a non-negative argument the type is recorded as png_uint_32 above. Consult
174 the header file and the text below for more information.
176 Special care must be take with sCAL chunk handling because the chunk itself
177 uses non-integral values encoded as strings containing decimal floating point
178 numbers. See the comments in the header file.
180 Configuration
182 The main header file function declarations are frequently protected by C
183 preprocessing directives of the form:
185 #ifdef PNG_feature_SUPPORTED
186 declare-function
187 #endif
188 ...
189 #ifdef PNG_feature_SUPPORTED
190 use-function
191 #endif
193 The library can be built without support for these APIs, although a
194 standard build will have all implemented APIs. Application programs
195 should check the feature macros before using an API for maximum
196 portability. From libpng 1.5.0 the feature macros set during the build
197 of libpng are recorded in the header file "pnglibconf.h" and this file
198 is always included by png.h.
200 If you don't need to change the library configuration from the default, skip to
201 the next section ("Reading").
203 Notice that some of the makefiles in the 'scripts' directory and (in 1.5.0) all
204 of the build project files in the 'projects' directory simply copy
205 scripts/pnglibconf.h.prebuilt to pnglibconf.h. This means that these build
206 systems do not permit easy auto-configuration of the library - they only
207 support the default configuration.
209 The easiest way to make minor changes to the libpng configuration when
210 auto-configuration is supported is to add definitions to the command line
211 using (typically) CPPFLAGS. For example:
215 will change the internal libpng math implementation for gamma correction and
216 other arithmetic calculations to fixed point, avoiding the need for fast
217 floating point support. The result can be seen in the generated pnglibconf.h -
218 make sure it contains the changed feature macro setting.
220 If you need to make more extensive configuration changes - more than one or two
221 feature macro settings - you can either add -DPNG_USER_CONFIG to the build
222 command line and put a list of feature macro settings in pngusr.h or you can set
223 DFA_XTRA (a makefile variable) to a file containing the same information in the
224 form of 'option' settings.
226 A. Changing pnglibconf.h
228 A variety of methods exist to build libpng. Not all of these support
229 reconfiguration of pnglibconf.h. To reconfigure pnglibconf.h it must either be
230 rebuilt from scripts/pnglibconf.dfa using awk or it must be edited by hand.
232 Hand editing is achieved by copying scripts/pnglibconf.h.prebuilt to
233 pnglibconf.h and changing the lines defining the supported features, paying
234 very close attention to the 'option' information in scripts/pnglibconf.dfa
235 that describes those features and their requirements. This is easy to get
236 wrong.
238 B. Configuration using DFA_XTRA
240 Rebuilding from pnglibconf.dfa is easy if a functioning 'awk', or a later
241 variant such as 'nawk' or 'gawk', is available. The configure build will
242 automatically find an appropriate awk and build pnglibconf.h.
243 The scripts/pnglibconf.mak file contains a set of make rules for doing the
244 same thing if configure is not used, and many of the makefiles in the scripts
245 directory use this approach.
247 When rebuilding simply write a new file containing changed options and set
248 DFA_XTRA to the name of this file. This causes the build to append the new file
249 to the end of scripts/pnglibconf.dfa. The pngusr.dfa file should contain lines
250 of the following forms:
252 everything = off
254 This turns all optional features off. Include it at the start of pngusr.dfa to
255 make it easier to build a minimal configuration. You will need to turn at least
256 some features on afterward to enable either reading or writing code, or both.
258 option feature on
259 option feature off
261 Enable or disable a single feature. This will automatically enable other
262 features required by a feature that is turned on or disable other features that
263 require a feature which is turned off. Conflicting settings will cause an error
264 message to be emitted by awk.
266 setting feature default value
268 Changes the default value of setting 'feature' to 'value'. There are a small
269 number of settings listed at the top of pnglibconf.h, they are documented in the
270 source code. Most of these values have performance implications for the library
271 but most of them have no visible effect on the API. Some can also be overridden
272 from the API.
274 This method of building a customized pnglibconf.h is illustrated in
275 contrib/pngminim/*. See the "$(PNGCONF):" target in the makefile and
276 pngusr.dfa in these directories.
278 C. Configuration using PNG_USER_CONFIG
280 If -DPNG_USER_CONFIG is added to the CPPFLAGS when pnglibconf.h is built,
281 the file pngusr.h will automatically be included before the options in
282 scripts/pnglibconf.dfa are processed. Your pngusr.h file should contain only
283 macro definitions turning features on or off or setting settings.
285 Apart from the global setting "everything = off" all the options listed above
286 can be set using macros in pngusr.h:
288 #define PNG_feature_SUPPORTED
290 is equivalent to:
292 option feature on
294 #define PNG_NO_feature
296 is equivalent to:
298 option feature off
300 #define PNG_feature value
302 is equivalent to:
304 setting feature default value
306 Notice that in both cases, pngusr.dfa and pngusr.h, the contents of the
307 pngusr file you supply override the contents of scripts/pnglibconf.dfa
309 If confusing or incomprehensible behavior results it is possible to
310 examine the intermediate file pnglibconf.dfn to find the full set of
311 dependency information for each setting and option. Simply locate the
312 feature in the file and read the C comments that precede it.
314 This method is also illustrated in the contrib/pngminim/* makefiles and
315 pngusr.h.
317 III. Reading
319 We'll now walk you through the possible functions to call when reading
320 in a PNG file sequentially, briefly explaining the syntax and purpose
321 of each one. See example.c and png.h for more detail. While
322 progressive reading is covered in the next section, you will still
323 need some of the functions discussed in this section to read a PNG
324 file.
326 Setup
328 You will want to do the I/O initialization(*) before you get into libpng,
329 so if it doesn't work, you don't have much to undo. Of course, you
330 will also want to insure that you are, in fact, dealing with a PNG
331 file. Libpng provides a simple check to see if a file is a PNG file.
332 To use it, pass in the first 1 to 8 bytes of the file to the function
333 png_sig_cmp(), and it will return 0 (false) if the bytes match the
334 corresponding bytes of the PNG signature, or nonzero (true) otherwise.
335 Of course, the more bytes you pass in, the greater the accuracy of the
336 prediction.
338 If you are intending to keep the file pointer open for use in libpng,
339 you must ensure you don't read more than 8 bytes from the beginning
340 of the file, and you also have to make a call to png_set_sig_bytes()
341 with the number of bytes you read from the beginning. Libpng will
342 then only check the bytes (if any) that your program didn't read.
344 (*): If you are not using the standard I/O functions, you will need
345 to replace them with custom functions. See the discussion under
346 Customizing libpng.
348 FILE *fp = fopen(file_name, "rb");
349 if (!fp)
350 {
351 return (ERROR);
352 }
354 if (fread(header, 1, number, fp) != number)
355 {
356 return (ERROR);
357 }
359 is_png = !png_sig_cmp(header, 0, number);
360 if (!is_png)
361 {
362 return (NOT_PNG);
363 }
365 Next, png_struct and png_info need to be allocated and initialized. In
366 order to ensure that the size of these structures is correct even with a
367 dynamically linked libpng, there are functions to initialize and
368 allocate the structures. We also pass the library version, optional
369 pointers to error handling functions, and a pointer to a data struct for
370 use by the error functions, if necessary (the pointer and functions can
371 be NULL if the default error handlers are to be used). See the section
372 on Changes to Libpng below regarding the old initialization functions.
373 The structure allocation functions quietly return NULL if they fail to
374 create the structure, so your application should check for that.
376 png_structp png_ptr = png_create_read_struct
377 (PNG_LIBPNG_VER_STRING, (png_voidp)user_error_ptr,
378 user_error_fn, user_warning_fn);
380 if (!png_ptr)
381 return (ERROR);
383 png_infop info_ptr = png_create_info_struct(png_ptr);
385 if (!info_ptr)
386 {
387 png_destroy_read_struct(&png_ptr,
388 (png_infopp)NULL, (png_infopp)NULL);
389 return (ERROR);
390 }
392 If you want to use your own memory allocation routines,
393 use a libpng that was built with PNG_USER_MEM_SUPPORTED defined, and use
394 png_create_read_struct_2() instead of png_create_read_struct():
396 png_structp png_ptr = png_create_read_struct_2
397 (PNG_LIBPNG_VER_STRING, (png_voidp)user_error_ptr,
398 user_error_fn, user_warning_fn, (png_voidp)
399 user_mem_ptr, user_malloc_fn, user_free_fn);
401 The error handling routines passed to png_create_read_struct()
402 and the memory alloc/free routines passed to png_create_struct_2()
403 are only necessary if you are not using the libpng supplied error
404 handling and memory alloc/free functions.
406 When libpng encounters an error, it expects to longjmp back
407 to your routine. Therefore, you will need to call setjmp and pass
408 your png_jmpbuf(png_ptr). If you read the file from different
409 routines, you will need to update the longjmp buffer every time you enter
410 a new routine that will call a png_*() function.
412 See your documentation of setjmp/longjmp for your compiler for more
413 information on setjmp/longjmp. See the discussion on libpng error
414 handling in the Customizing Libpng section below for more information
415 on the libpng error handling. If an error occurs, and libpng longjmp's
416 back to your setjmp, you will want to call png_destroy_read_struct() to
417 free any memory.
419 if (setjmp(png_jmpbuf(png_ptr)))
420 {
421 png_destroy_read_struct(&png_ptr, &info_ptr,
422 &end_info);
423 fclose(fp);
424 return (ERROR);
425 }
427 Pass (png_infopp)NULL instead of &end_info if you didn't create
428 an end_info structure.
430 If you would rather avoid the complexity of setjmp/longjmp issues,
431 you can compile libpng with PNG_NO_SETJMP, in which case
432 errors will result in a call to PNG_ABORT() which defaults to abort().
434 You can #define PNG_ABORT() to a function that does something
435 more useful than abort(), as long as your function does not
436 return.
438 Now you need to set up the input code. The default for libpng is to
439 use the C function fread(). If you use this, you will need to pass a
440 valid FILE * in the function png_init_io(). Be sure that the file is
441 opened in binary mode. If you wish to handle reading data in another
442 way, you need not call the png_init_io() function, but you must then
443 implement the libpng I/O methods discussed in the Customizing Libpng
444 section below.
446 png_init_io(png_ptr, fp);
448 If you had previously opened the file and read any of the signature from
449 the beginning in order to see if this was a PNG file, you need to let
450 libpng know that there are some bytes missing from the start of the file.
452 png_set_sig_bytes(png_ptr, number);
454 You can change the zlib compression buffer size to be used while
455 reading compressed data with
457 png_set_compression_buffer_size(png_ptr, buffer_size);
459 where the default size is 8192 bytes. Note that the buffer size
460 is changed immediately and the buffer is reallocated immediately,
461 instead of setting a flag to be acted upon later.
463 If you want CRC errors to be handled in a different manner than
464 the default, use
466 png_set_crc_action(png_ptr, crit_action, ancil_action);
468 The values for png_set_crc_action() say how libpng is to handle CRC errors in
469 ancillary and critical chunks, and whether to use the data contained
470 therein. Starting with libpng-1.6.26, this also governs how an ADLER32 error
471 is handled while reading the IDAT chunk. Note that it is impossible to
472 "discard" data in a critical chunk.
474 Choices for (int) crit_action are
475 PNG_CRC_DEFAULT 0 error/quit
476 PNG_CRC_ERROR_QUIT 1 error/quit
477 PNG_CRC_WARN_USE 3 warn/use data
478 PNG_CRC_QUIET_USE 4 quiet/use data
479 PNG_CRC_NO_CHANGE 5 use the current value
481 Choices for (int) ancil_action are
482 PNG_CRC_DEFAULT 0 error/quit
483 PNG_CRC_ERROR_QUIT 1 error/quit
484 PNG_CRC_WARN_DISCARD 2 warn/discard data
485 PNG_CRC_WARN_USE 3 warn/use data
486 PNG_CRC_QUIET_USE 4 quiet/use data
487 PNG_CRC_NO_CHANGE 5 use the current value
489 When the setting for crit_action is PNG_CRC_QUIET_USE, the CRC and ADLER32
490 checksums are not only ignored, but they are not evaluated.
492 Setting up callback code
494 You can set up a callback function to handle any unknown chunks in the
495 input stream. You must supply the function
497 read_chunk_callback(png_structp png_ptr,
498 png_unknown_chunkp chunk);
499 {
500 /* The unknown chunk structure contains your
501 chunk data, along with similar data for any other
502 unknown chunks: */
504 png_byte name[5];
505 png_byte *data;
506 png_size_t size;
508 /* Note that libpng has already taken care of
509 the CRC handling */
511 /* put your code here. Search for your chunk in the
512 unknown chunk structure, process it, and return one
513 of the following: */
515 return (-n); /* chunk had an error */
516 return (0); /* did not recognize */
517 return (n); /* success */
518 }
520 (You can give your function another name that you like instead of
521 "read_chunk_callback")
523 To inform libpng about your function, use
525 png_set_read_user_chunk_fn(png_ptr, user_chunk_ptr,
526 read_chunk_callback);
528 This names not only the callback function, but also a user pointer that
529 you can retrieve with
531 png_get_user_chunk_ptr(png_ptr);
533 If you call the png_set_read_user_chunk_fn() function, then all unknown
534 chunks which the callback does not handle will be saved when read. You can
535 cause them to be discarded by returning '1' ("handled") instead of '0'. This
536 behavior will change in libpng 1.7 and the default handling set by the
537 png_set_keep_unknown_chunks() function, described below, will be used when the
538 callback returns 0. If you want the existing behavior you should set the global
539 default to PNG_HANDLE_CHUNK_IF_SAFE now; this is compatible with all current
540 versions of libpng and with 1.7. Libpng 1.6 issues a warning if you keep the
541 default, or PNG_HANDLE_CHUNK_NEVER, and the callback returns 0.
543 At this point, you can set up a callback function that will be
544 called after each row has been read, which you can use to control
545 a progress meter or the like. It's demonstrated in pngtest.c.
546 You must supply a function
548 void read_row_callback(png_structp png_ptr,
549 png_uint_32 row, int pass);
550 {
551 /* put your code here */
552 }
554 (You can give it another name that you like instead of "read_row_callback")
556 To inform libpng about your function, use
558 png_set_read_status_fn(png_ptr, read_row_callback);
560 When this function is called the row has already been completely processed and
561 the 'row' and 'pass' refer to the next row to be handled. For the
562 non-interlaced case the row that was just handled is simply one less than the
563 passed in row number, and pass will always be 0. For the interlaced case the
564 same applies unless the row value is 0, in which case the row just handled was
565 the last one from one of the preceding passes. Because interlacing may skip a
566 pass you cannot be sure that the preceding pass is just 'pass-1'; if you really
567 need to know what the last pass is record (row,pass) from the callback and use
568 the last recorded value each time.
570 As with the user transform you can find the output row using the
573 Unknown-chunk handling
575 Now you get to set the way the library processes unknown chunks in the
576 input PNG stream. Both known and unknown chunks will be read. Normal
577 behavior is that known chunks will be parsed into information in
578 various info_ptr members while unknown chunks will be discarded. This
579 behavior can be wasteful if your application will never use some known
580 chunk types. To change this, you can call:
582 png_set_keep_unknown_chunks(png_ptr, keep,
583 chunk_list, num_chunks);
585 keep - 0: default unknown chunk handling
586 1: ignore; do not keep
587 2: keep only if safe-to-copy
588 3: keep even if unsafe-to-copy
590 You can use these definitions:
596 chunk_list - list of chunks affected (a byte string,
597 five bytes per chunk, NULL or '\0' if
598 num_chunks is positive; ignored if
599 numchunks <= 0).
601 num_chunks - number of chunks affected; if 0, all
602 unknown chunks are affected. If positive,
603 only the chunks in the list are affected,
604 and if negative all unknown chunks and
605 all known chunks except for the IHDR,
606 PLTE, tRNS, IDAT, and IEND chunks are
607 affected.
609 Unknown chunks declared in this way will be saved as raw data onto a
610 list of png_unknown_chunk structures. If a chunk that is normally
611 known to libpng is named in the list, it will be handled as unknown,
612 according to the "keep" directive. If a chunk is named in successive
613 instances of png_set_keep_unknown_chunks(), the final instance will
614 take precedence. The IHDR and IEND chunks should not be named in
615 chunk_list; if they are, libpng will process them normally anyway.
616 If you know that your application will never make use of some particular
617 chunks, use PNG_HANDLE_CHUNK_NEVER (or 1) as demonstrated below.
619 Here is an example of the usage of png_set_keep_unknown_chunks(),
620 where the private "vpAg" chunk will later be processed by a user chunk
621 callback function:
623 png_byte vpAg[5]={118, 112, 65, 103, (png_byte) '\0'};
626 png_byte unused_chunks[]=
627 {
628 104, 73, 83, 84, (png_byte) '\0', /* hIST */
629 105, 84, 88, 116, (png_byte) '\0', /* iTXt */
630 112, 67, 65, 76, (png_byte) '\0', /* pCAL */
631 115, 67, 65, 76, (png_byte) '\0', /* sCAL */
632 115, 80, 76, 84, (png_byte) '\0', /* sPLT */
633 116, 73, 77, 69, (png_byte) '\0', /* tIME */
634 };
635 #endif
637 ...
640 /* ignore all unknown chunks
641 * (use global setting "2" for libpng16 and earlier):
642 */
643 png_set_keep_unknown_chunks(read_ptr, 2, NULL, 0);
645 /* except for vpAg: */
646 png_set_keep_unknown_chunks(read_ptr, 2, vpAg, 1);
648 /* also ignore unused known chunks: */
649 png_set_keep_unknown_chunks(read_ptr, 1, unused_chunks,
650 (int)(sizeof unused_chunks)/5);
651 #endif
653 User limits
655 The PNG specification allows the width and height of an image to be as
656 large as 2^31-1 (0x7fffffff), or about 2.147 billion rows and columns.
657 For safety, libpng imposes a default limit of 1 million rows and columns.
658 Larger images will be rejected immediately with a png_error() call. If
659 you wish to change these limits, you can use
661 png_set_user_limits(png_ptr, width_max, height_max);
663 to set your own limits (libpng may reject some very wide images
664 anyway because of potential buffer overflow conditions).
666 You should put this statement after you create the PNG structure and
667 before calling png_read_info(), png_read_png(), or png_process_data().
669 When writing a PNG datastream, put this statement before calling
670 png_write_info() or png_write_png().
672 If you need to retrieve the limits that are being applied, use
674 width_max = png_get_user_width_max(png_ptr);
675 height_max = png_get_user_height_max(png_ptr);
677 The PNG specification sets no limit on the number of ancillary chunks
678 allowed in a PNG datastream. By default, libpng imposes a limit of
679 a total of 1000 sPLT, tEXt, iTXt, zTXt, and unknown chunks to be stored.
680 If you have set up both info_ptr and end_info_ptr, the limit applies
681 separately to each. You can change the limit on the total number of such
682 chunks that will be stored, with
684 png_set_chunk_cache_max(png_ptr, user_chunk_cache_max);
686 where 0x7fffffffL means unlimited. You can retrieve this limit with
688 chunk_cache_max = png_get_chunk_cache_max(png_ptr);
690 Libpng imposes a limit of 8 Megabytes (8,000,000 bytes) on the amount of
691 memory that a compressed chunk other than IDAT can occupy, when decompressed.
692 You can change this limit with
694 png_set_chunk_malloc_max(png_ptr, user_chunk_malloc_max);
696 and you can retrieve the limit with
698 chunk_malloc_max = png_get_chunk_malloc_max(png_ptr);
700 Any chunks that would cause either of these limits to be exceeded will
701 be ignored.
703 Information about your system
705 If you intend to display the PNG or to incorporate it in other image data you
706 need to tell libpng information about your display or drawing surface so that
707 libpng can convert the values in the image to match the display.
709 From libpng-1.5.4 this information can be set before reading the PNG file
710 header. In earlier versions png_set_gamma() existed but behaved incorrectly if
711 called before the PNG file header had been read and png_set_alpha_mode() did not
712 exist.
714 If you need to support versions prior to libpng-1.5.4 test the version number
715 as illustrated below using "PNG_LIBPNG_VER >= 10504" and follow the procedures
716 described in the appropriate manual page.
718 You give libpng the encoding expected by your system expressed as a 'gamma'
719 value. You can also specify a default encoding for the PNG file in
720 case the required information is missing from the file. By default libpng
721 assumes that the PNG data matches your system, to keep this default call:
723 png_set_gamma(png_ptr, screen_gamma, output_gamma);
725 or you can use the fixed point equivalent:
727 png_set_gamma_fixed(png_ptr, PNG_FP_1*screen_gamma,
728 PNG_FP_1*output_gamma);
730 If you don't know the gamma for your system it is probably 2.2 - a good
731 approximation to the IEC standard for display systems (sRGB). If images are
732 too contrasty or washed out you got the value wrong - check your system
733 documentation!
735 Many systems permit the system gamma to be changed via a lookup table in the
736 display driver, a few systems, including older Macs, change the response by
737 default. As of 1.5.4 three special values are available to handle common
738 situations:
740 PNG_DEFAULT_sRGB: Indicates that the system conforms to the
741 IEC 61966-2-1 standard. This matches almost
742 all systems.
743 PNG_GAMMA_MAC_18: Indicates that the system is an older
744 (pre Mac OS 10.6) Apple Macintosh system with
745 the default settings.
746 PNG_GAMMA_LINEAR: Just the fixed point value for 1.0 - indicates
747 that the system expects data with no gamma
748 encoding.
750 You would use the linear (unencoded) value if you need to process the pixel
751 values further because this avoids the need to decode and re-encode each
752 component value whenever arithmetic is performed. A lot of graphics software
753 uses linear values for this reason, often with higher precision component values
754 to preserve overall accuracy.
757 The output_gamma value expresses how to decode the output values, not how
758 they are encoded. The values used correspond to the normal numbers used to
759 describe the overall gamma of a computer display system; for example 2.2 for
760 an sRGB conformant system. The values are scaled by 100000 in the _fixed
761 version of the API (so 220000 for sRGB.)
763 The inverse of the value is always used to provide a default for the PNG file
764 encoding if it has no gAMA chunk and if png_set_gamma() has not been called
765 to override the PNG gamma information.
767 When the ALPHA_OPTIMIZED mode is selected the output gamma is used to encode
768 opaque pixels however pixels with lower alpha values are not encoded,
769 regardless of the output gamma setting.
771 When the standard Porter Duff handling is requested with mode 1 the output
772 encoding is set to be linear and the output_gamma value is only relevant
773 as a default for input data that has no gamma information. The linear output
774 encoding will be overridden if png_set_gamma() is called - the results may be
775 highly unexpected!
777 The following numbers are derived from the sRGB standard and the research
778 behind it. sRGB is defined to be approximated by a PNG gAMA chunk value of
779 0.45455 (1/2.2) for PNG. The value implicitly includes any viewing
780 correction required to take account of any differences in the color
781 environment of the original scene and the intended display environment; the
782 value expresses how to *decode* the image for display, not how the original
783 data was *encoded*.
785 sRGB provides a peg for the PNG standard by defining a viewing environment.
786 sRGB itself, and earlier TV standards, actually use a more complex transform
787 (a linear portion then a gamma 2.4 power law) than PNG can express. (PNG is
788 limited to simple power laws.) By saying that an image for direct display on
789 an sRGB conformant system should be stored with a gAMA chunk value of 45455
790 ( and of the ISO PNG specification) the PNG specification
791 makes it possible to derive values for other display systems and
792 environments.
794 The Mac value is deduced from the sRGB based on an assumption that the actual
795 extra viewing correction used in early Mac display systems was implemented as
796 a power 1.45 lookup table.
798 Any system where a programmable lookup table is used or where the behavior of
799 the final display device characteristics can be changed requires system
800 specific code to obtain the current characteristic. However this can be
801 difficult and most PNG gamma correction only requires an approximate value.
803 By default, if png_set_alpha_mode() is not called, libpng assumes that all
804 values are unencoded, linear, values and that the output device also has a
805 linear characteristic. This is only very rarely correct - it is invariably
806 better to call png_set_alpha_mode() with PNG_DEFAULT_sRGB than rely on the
807 default if you don't know what the right answer is!
809 The special value PNG_GAMMA_MAC_18 indicates an older Mac system (pre Mac OS
810 10.6) which used a correction table to implement a somewhat lower gamma on an
811 otherwise sRGB system.
813 Both these values are reserved (not simple gamma values) in order to allow
814 more precise correction internally in the future.
816 NOTE: the values can be passed to either the fixed or floating
817 point APIs, but the floating point API will also accept floating point
818 values.
820 The second thing you may need to tell libpng about is how your system handles
821 alpha channel information. Some, but not all, PNG files contain an alpha
822 channel. To display these files correctly you need to compose the data onto a
823 suitable background, as described in the PNG specification.
825 Libpng only supports composing onto a single color (using png_set_background;
826 see below). Otherwise you must do the composition yourself and, in this case,
827 you may need to call png_set_alpha_mode:
829 #if PNG_LIBPNG_VER >= 10504
830 png_set_alpha_mode(png_ptr, mode, screen_gamma);
831 #else
832 png_set_gamma(png_ptr, screen_gamma, 1.0/screen_gamma);
833 #endif
835 The screen_gamma value is the same as the argument to png_set_gamma; however,
836 how it affects the output depends on the mode. png_set_alpha_mode() sets the
837 file gamma default to 1/screen_gamma, so normally you don't need to call
838 png_set_gamma. If you need different defaults call png_set_gamma() before
839 png_set_alpha_mode() - if you call it after it will override the settings made
840 by png_set_alpha_mode().
842 The mode is as follows:
844 PNG_ALPHA_PNG: The data is encoded according to the PNG
845 specification. Red, green and blue, or gray, components are
846 gamma encoded color values and are not premultiplied by the
847 alpha value. The alpha value is a linear measure of the
848 contribution of the pixel to the corresponding final output pixel.
850 You should normally use this format if you intend to perform
851 color correction on the color values; most, maybe all, color
852 correction software has no handling for the alpha channel and,
853 anyway, the math to handle pre-multiplied component values is
854 unnecessarily complex.
856 Before you do any arithmetic on the component values you need
857 to remove the gamma encoding and multiply out the alpha
858 channel. See the PNG specification for more detail. It is
859 important to note that when an image with an alpha channel is
860 scaled, linear encoded, pre-multiplied component values must
861 be used!
863 The remaining modes assume you don't need to do any further color correction or
864 that if you do, your color correction software knows all about alpha (it
865 probably doesn't!). They 'associate' the alpha with the color information by
866 storing color channel values that have been scaled by the alpha. The
867 advantage is that the color channels can be resampled (the image can be
868 scaled) in this form. The disadvantage is that normal practice is to store
869 linear, not (gamma) encoded, values and this requires 16-bit channels for
870 still images rather than the 8-bit channels that are just about sufficient if
871 gamma encoding is used. In addition all non-transparent pixel values,
872 including completely opaque ones, must be gamma encoded to produce the final
873 image. These are the 'STANDARD', 'ASSOCIATED' or 'PREMULTIPLIED' modes
874 described below (the latter being the two common names for associated alpha
875 color channels). Note that PNG files always contain non-associated color
876 channels; png_set_alpha_mode() with one of the modes causes the decoder to
877 convert the pixels to an associated form before returning them to your
878 application.
880 Since it is not necessary to perform arithmetic on opaque color values so
881 long as they are not to be resampled and are in the final color space it is
882 possible to optimize the handling of alpha by storing the opaque pixels in
883 the PNG format (adjusted for the output color space) while storing partially
884 opaque pixels in the standard, linear, format. The accuracy required for
885 standard alpha composition is relatively low, because the pixels are
886 isolated, therefore typically the accuracy loss in storing 8-bit linear
887 values is acceptable. (This is not true if the alpha channel is used to
888 simulate transparency over large areas - use 16 bits or the PNG mode in
889 this case!) This is the 'OPTIMIZED' mode. For this mode a pixel is
890 treated as opaque only if the alpha value is equal to the maximum value.
892 PNG_ALPHA_STANDARD: The data libpng produces is encoded in the
893 standard way assumed by most correctly written graphics software.
894 The gamma encoding will be removed by libpng and the
895 linear component values will be pre-multiplied by the
896 alpha channel.
898 With this format the final image must be re-encoded to
899 match the display gamma before the image is displayed.
900 If your system doesn't do that, yet still seems to
901 perform arithmetic on the pixels without decoding them,
902 it is broken - check out the modes below.
904 With PNG_ALPHA_STANDARD libpng always produces linear
905 component values, whatever screen_gamma you supply. The
906 screen_gamma value is, however, used as a default for
907 the file gamma if the PNG file has no gamma information.
909 If you call png_set_gamma() after png_set_alpha_mode() you
910 will override the linear encoding. Instead the
911 pre-multiplied pixel values will be gamma encoded but
912 the alpha channel will still be linear. This may
913 actually match the requirements of some broken software,
914 but it is unlikely.
916 While linear 8-bit data is often used it has
917 insufficient precision for any image with a reasonable
918 dynamic range. To avoid problems, and if your software
919 supports it, use png_set_expand_16() to force all
920 components to 16 bits.
922 PNG_ALPHA_OPTIMIZED: This mode is the same as PNG_ALPHA_STANDARD
923 except that completely opaque pixels are gamma encoded according to
924 the screen_gamma value. Pixels with alpha less than 1.0
925 will still have linear components.
927 Use this format if you have control over your
928 compositing software and so don't do other arithmetic
929 (such as scaling) on the data you get from libpng. Your
930 compositing software can simply copy opaque pixels to
931 the output but still has linear values for the
932 non-opaque pixels.
934 In normal compositing, where the alpha channel encodes
935 partial pixel coverage (as opposed to broad area
936 translucency), the inaccuracies of the 8-bit
937 representation of non-opaque pixels are irrelevant.
939 You can also try this format if your software is broken;
940 it might look better.
942 PNG_ALPHA_BROKEN: This is PNG_ALPHA_STANDARD; however, all component
943 values, including the alpha channel are gamma encoded. This is
944 broken because, in practice, no implementation that uses this choice
945 correctly undoes the encoding before handling alpha composition. Use this
946 choice only if other serious errors in the software or hardware you use
947 mandate it. In most cases of broken software or hardware the bug in the
948 final display manifests as a subtle halo around composited parts of the
949 image. You may not even perceive this as a halo; the composited part of
950 the image may simply appear separate from the background, as though it had
951 been cut out of paper and pasted on afterward.
953 If you don't have to deal with bugs in software or hardware, or if you can fix
954 them, there are three recommended ways of using png_set_alpha_mode():
956 png_set_alpha_mode(png_ptr, PNG_ALPHA_PNG,
957 screen_gamma);
959 You can do color correction on the result (libpng does not currently
960 support color correction internally). When you handle the alpha channel
961 you need to undo the gamma encoding and multiply out the alpha.
963 png_set_alpha_mode(png_ptr, PNG_ALPHA_STANDARD,
964 screen_gamma);
965 png_set_expand_16(png_ptr);
967 If you are using the high level interface, don't call png_set_expand_16();
968 instead pass PNG_TRANSFORM_EXPAND_16 to the interface.
970 With this mode you can't do color correction, but you can do arithmetic,
971 including composition and scaling, on the data without further processing.
973 png_set_alpha_mode(png_ptr, PNG_ALPHA_OPTIMIZED,
974 screen_gamma);
976 You can avoid the expansion to 16-bit components with this mode, but you
977 lose the ability to scale the image or perform other linear arithmetic.
978 All you can do is compose the result onto a matching output. Since this
979 mode is libpng-specific you also need to write your own composition
980 software.
982 The following are examples of calls to png_set_alpha_mode to achieve the
983 required overall gamma correction and, where necessary, alpha
984 premultiplication.
986 png_set_alpha_mode(pp, PNG_ALPHA_PNG, PNG_DEFAULT_sRGB);
988 This is the default libpng handling of the alpha channel - it is not
989 pre-multiplied into the color components. In addition the call states
990 that the output is for a sRGB system and causes all PNG files without gAMA
991 chunks to be assumed to be encoded using sRGB.
993 png_set_alpha_mode(pp, PNG_ALPHA_PNG, PNG_GAMMA_MAC);
995 In this case the output is assumed to be something like an sRGB conformant
996 display preceeded by a power-law lookup table of power 1.45. This is how
997 early Mac systems behaved.
999 png_set_alpha_mode(pp, PNG_ALPHA_STANDARD, PNG_GAMMA_LINEAR);
1001 This is the classic Jim Blinn approach and will work in academic
1002 environments where everything is done by the book. It has the shortcoming
1003 of assuming that input PNG data with no gamma information is linear - this
1004 is unlikely to be correct unless the PNG files where generated locally.
1005 Most of the time the output precision will be so low as to show
1006 significant banding in dark areas of the image.
1008 png_set_expand_16(pp);
1009 png_set_alpha_mode(pp, PNG_ALPHA_STANDARD, PNG_DEFAULT_sRGB);
1011 This is a somewhat more realistic Jim Blinn inspired approach. PNG files
1012 are assumed to have the sRGB encoding if not marked with a gamma value and
1013 the output is always 16 bits per component. This permits accurate scaling
1014 and processing of the data. If you know that your input PNG files were
1015 generated locally you might need to replace PNG_DEFAULT_sRGB with the
1016 correct value for your system.
1018 png_set_alpha_mode(pp, PNG_ALPHA_OPTIMIZED, PNG_DEFAULT_sRGB);
1020 If you just need to composite the PNG image onto an existing background
1021 and if you control the code that does this you can use the optimization
1022 setting. In this case you just copy completely opaque pixels to the
1023 output. For pixels that are not completely transparent (you just skip
1024 those) you do the composition math using png_composite or png_composite_16
1025 below then encode the resultant 8-bit or 16-bit values to match the output
1026 encoding.
1028 Other cases
1030 If neither the PNG nor the standard linear encoding work for you because
1031 of the software or hardware you use then you have a big problem. The PNG
1032 case will probably result in halos around the image. The linear encoding
1033 will probably result in a washed out, too bright, image (it's actually too
1034 contrasty.) Try the ALPHA_OPTIMIZED mode above - this will probably
1035 substantially reduce the halos. Alternatively try:
1037 png_set_alpha_mode(pp, PNG_ALPHA_BROKEN, PNG_DEFAULT_sRGB);
1039 This option will also reduce the halos, but there will be slight dark
1040 halos round the opaque parts of the image where the background is light.
1041 In the OPTIMIZED mode the halos will be light halos where the background
1042 is dark. Take your pick - the halos are unavoidable unless you can get
1043 your hardware/software fixed! (The OPTIMIZED approach is slightly
1044 faster.)
1046 When the default gamma of PNG files doesn't match the output gamma.
1047 If you have PNG files with no gamma information png_set_alpha_mode allows
1048 you to provide a default gamma, but it also sets the ouput gamma to the
1049 matching value. If you know your PNG files have a gamma that doesn't
1050 match the output you can take advantage of the fact that
1051 png_set_alpha_mode always sets the output gamma but only sets the PNG
1052 default if it is not already set:
1054 png_set_alpha_mode(pp, PNG_ALPHA_PNG, PNG_DEFAULT_sRGB);
1055 png_set_alpha_mode(pp, PNG_ALPHA_PNG, PNG_GAMMA_MAC);
1057 The first call sets both the default and the output gamma values, the
1058 second call overrides the output gamma without changing the default. This
1059 is easier than achieving the same effect with png_set_gamma. You must use
1060 PNG_ALPHA_PNG for the first call - internal checking in png_set_alpha will
1061 fire if more than one call to png_set_alpha_mode and png_set_background is
1062 made in the same read operation, however multiple calls with PNG_ALPHA_PNG
1063 are ignored.
1065 If you don't need, or can't handle, the alpha channel you can call
1066 png_set_background() to remove it by compositing against a fixed color. Don't
1067 call png_set_strip_alpha() to do this - it will leave spurious pixel values in
1068 transparent parts of this image.
1070 png_set_background(png_ptr, &background_color,
1073 The background_color is an RGB or grayscale value according to the data format
1074 libpng will produce for you. Because you don't yet know the format of the PNG
1075 file, if you call png_set_background at this point you must arrange for the
1076 format produced by libpng to always have 8-bit or 16-bit components and then
1077 store the color as an 8-bit or 16-bit color as appropriate. The color contains
1078 separate gray and RGB component values, so you can let libpng produce gray or
1079 RGB output according to the input format, but low bit depth grayscale images
1080 must always be converted to at least 8-bit format. (Even though low bit depth
1081 grayscale images can't have an alpha channel they can have a transparent
1082 color!)
1084 You set the transforms you need later, either as flags to the high level
1085 interface or libpng API calls for the low level interface. For reference the
1086 settings and API calls required are:
1088 8-bit values:
1090 png_set_expand(png_ptr); png_set_scale_16(png_ptr);
1092 If you must get exactly the same inaccurate results
1093 produced by default in versions prior to libpng-1.5.4,
1094 use PNG_TRANSFORM_STRIP_16 and png_set_strip_16(png_ptr)
1095 instead.
1097 16-bit values:
1099 png_set_expand_16(png_ptr);
1101 In either case palette image data will be expanded to RGB. If you just want
1102 color data you can add PNG_TRANSFORM_GRAY_TO_RGB or png_set_gray_to_rgb(png_ptr)
1103 to the list.
1105 Calling png_set_background before the PNG file header is read will not work
1106 prior to libpng-1.5.4. Because the failure may result in unexpected warnings or
1107 errors it is therefore much safer to call png_set_background after the head has
1108 been read. Unfortunately this means that prior to libpng-1.5.4 it cannot be
1109 used with the high level interface.
1111 The high-level read interface
1113 At this point there are two ways to proceed; through the high-level
1114 read interface, or through a sequence of low-level read operations.
1115 You can use the high-level interface if (a) you are willing to read
1116 the entire image into memory, and (b) the input transformations
1117 you want to do are limited to the following set:
1119 PNG_TRANSFORM_IDENTITY No transformation
1120 PNG_TRANSFORM_SCALE_16 Strip 16-bit samples to
1121 8-bit accurately
1122 PNG_TRANSFORM_STRIP_16 Chop 16-bit samples to
1123 8-bit less accurately
1124 PNG_TRANSFORM_STRIP_ALPHA Discard the alpha channel
1125 PNG_TRANSFORM_PACKING Expand 1, 2 and 4-bit
1126 samples to bytes
1127 PNG_TRANSFORM_PACKSWAP Change order of packed
1128 pixels to LSB first
1129 PNG_TRANSFORM_EXPAND Perform set_expand()
1130 PNG_TRANSFORM_INVERT_MONO Invert monochrome images
1131 PNG_TRANSFORM_SHIFT Normalize pixels to the
1132 sBIT depth
1134 to BGRA
1136 to AG
1137 PNG_TRANSFORM_INVERT_ALPHA Change alpha from opacity
1138 to transparency
1139 PNG_TRANSFORM_SWAP_ENDIAN Byte-swap 16-bit samples
1140 PNG_TRANSFORM_GRAY_TO_RGB Expand grayscale samples
1141 to RGB (or GA to RGBA)
1142 PNG_TRANSFORM_EXPAND_16 Expand samples to 16 bits
1144 (This excludes setting a background color, doing gamma transformation,
1145 quantizing, and setting filler.) If this is the case, simply do this:
1147 png_read_png(png_ptr, info_ptr, png_transforms, NULL)
1149 where png_transforms is an integer containing the bitwise OR of some
1150 set of transformation flags. This call is equivalent to png_read_info(),
1151 followed the set of transformations indicated by the transform mask,
1152 then png_read_image(), and finally png_read_end().
1154 (The final parameter of this call is not yet used. Someday it might point
1155 to transformation parameters required by some future input transform.)
1157 You must use png_transforms and not call any png_set_transform() functions
1158 when you use png_read_png().
1160 After you have called png_read_png(), you can retrieve the image data
1161 with
1163 row_pointers = png_get_rows(png_ptr, info_ptr);
1165 where row_pointers is an array of pointers to the pixel data for each row:
1167 png_bytep row_pointers[height];
1169 If you know your image size and pixel size ahead of time, you can allocate
1170 row_pointers prior to calling png_read_png() with
1172 if (height > PNG_UINT_32_MAX/(sizeof (png_byte)))
1173 png_error (png_ptr,
1174 "Image is too tall to process in memory");
1176 if (width > PNG_UINT_32_MAX/pixel_size)
1177 png_error (png_ptr,
1178 "Image is too wide to process in memory");
1180 row_pointers = png_malloc(png_ptr,
1181 height*(sizeof (png_bytep)));
1183 for (int i=0; i<height, i++)
1184 row_pointers[i]=NULL; /* security precaution */
1186 for (int i=0; i<height, i++)
1187 row_pointers[i]=png_malloc(png_ptr,
1188 width*pixel_size);
1190 png_set_rows(png_ptr, info_ptr, &row_pointers);
1192 Alternatively you could allocate your image in one big block and define
1193 row_pointers[i] to point into the proper places in your block.
1195 If you use png_set_rows(), the application is responsible for freeing
1196 row_pointers (and row_pointers[i], if they were separately allocated).
1198 If you don't allocate row_pointers ahead of time, png_read_png() will
1199 do it, and it'll be free'ed by libpng when you call png_destroy_*().
1201 The low-level read interface
1203 If you are going the low-level route, you are now ready to read all
1204 the file information up to the actual image data. You do this with a
1205 call to png_read_info().
1207 png_read_info(png_ptr, info_ptr);
1209 This will process all chunks up to but not including the image data.
1211 This also copies some of the data from the PNG file into the decode structure
1212 for use in later transformations. Important information copied in is:
1214 1) The PNG file gamma from the gAMA chunk. This overwrites the default value
1215 provided by an earlier call to png_set_gamma or png_set_alpha_mode.
1217 2) Prior to libpng-1.5.4 the background color from a bKGd chunk. This
1218 damages the information provided by an earlier call to png_set_background
1219 resulting in unexpected behavior. Libpng-1.5.4 no longer does this.
1221 3) The number of significant bits in each component value. Libpng uses this to
1222 optimize gamma handling by reducing the internal lookup table sizes.
1224 4) The transparent color information from a tRNS chunk. This can be modified by
1225 a later call to png_set_tRNS.
1227 Querying the info structure
1229 Functions are used to get the information from the info_ptr once it
1230 has been read. Note that these fields may not be completely filled
1231 in until png_read_end() has read the chunk data following the image.
1233 png_get_IHDR(png_ptr, info_ptr, &width, &height,
1234 &bit_depth, &color_type, &interlace_type,
1235 &compression_type, &filter_method);
1237 width - holds the width of the image
1238 in pixels (up to 2^31).
1240 height - holds the height of the image
1241 in pixels (up to 2^31).
1243 bit_depth - holds the bit depth of one of the
1244 image channels. (valid values are
1245 1, 2, 4, 8, 16 and depend also on
1246 the color_type. See also
1247 significant bits (sBIT) below).
1249 color_type - describes which color/alpha channels
1250 are present.
1252 (bit depths 1, 2, 4, 8, 16)
1254 (bit depths 8, 16)
1256 (bit depths 1, 2, 4, 8)
1258 (bit_depths 8, 16)
1260 (bit_depths 8, 16)
1266 interlace_type - (PNG_INTERLACE_NONE or
1269 compression_type - (must be PNG_COMPRESSION_TYPE_BASE
1270 for PNG 1.0)
1272 filter_method - (must be PNG_FILTER_TYPE_BASE
1273 for PNG 1.0, and can also be
1275 the PNG datastream is embedded in
1276 a MNG-1.0 datastream)
1278 Any of width, height, color_type, bit_depth,
1279 interlace_type, compression_type, or filter_method can
1280 be NULL if you are not interested in their values.
1282 Note that png_get_IHDR() returns 32-bit data into
1283 the application's width and height variables.
1284 This is an unsafe situation if these are not png_uint_32
1285 variables. In such situations, the
1286 png_get_image_width() and png_get_image_height()
1287 functions described below are safer.
1289 width = png_get_image_width(png_ptr,
1290 info_ptr);
1292 height = png_get_image_height(png_ptr,
1293 info_ptr);
1295 bit_depth = png_get_bit_depth(png_ptr,
1296 info_ptr);
1298 color_type = png_get_color_type(png_ptr,
1299 info_ptr);
1301 interlace_type = png_get_interlace_type(png_ptr,
1302 info_ptr);
1304 compression_type = png_get_compression_type(png_ptr,
1305 info_ptr);
1307 filter_method = png_get_filter_type(png_ptr,
1308 info_ptr);
1310 channels = png_get_channels(png_ptr, info_ptr);
1312 channels - number of channels of info for the
1313 color type (valid values are 1 (GRAY,
1314 PALETTE), 2 (GRAY_ALPHA), 3 (RGB),
1315 4 (RGB_ALPHA or RGB + filler byte))
1317 rowbytes = png_get_rowbytes(png_ptr, info_ptr);
1319 rowbytes - number of bytes needed to hold a row
1321 signature = png_get_signature(png_ptr, info_ptr);
1323 signature - holds the signature read from the
1324 file (if any). The data is kept in
1325 the same offset it would be if the
1326 whole signature were read (i.e. if an
1327 application had already read in 4
1328 bytes of signature before starting
1329 libpng, the remaining 4 bytes would
1330 be in signature[4] through signature[7]
1331 (see png_set_sig_bytes())).
1333 These are also important, but their validity depends on whether the chunk
1334 has been read. The png_get_valid(png_ptr, info_ptr, PNG_INFO_<chunk>) and
1335 png_get_<chunk>(png_ptr, info_ptr, ...) functions return non-zero if the
1336 data has been read, or zero if it is missing. The parameters to the
1337 png_get_<chunk> are set directly if they are simple data types, or a
1338 pointer into the info_ptr is returned for any complex types.
1340 The colorspace data from gAMA, cHRM, sRGB, iCCP, and sBIT chunks
1341 is simply returned to give the application information about how the
1342 image was encoded. Libpng itself only does transformations using the file
1343 gamma when combining semitransparent pixels with the background color, and,
1344 since libpng-1.6.0, when converting between 8-bit sRGB and 16-bit linear pixels
1345 within the simplified API. Libpng also uses the file gamma when converting
1346 RGB to gray, beginning with libpng-1.0.5, if the application calls
1347 png_set_rgb_to_gray()).
1349 png_get_PLTE(png_ptr, info_ptr, &palette,
1350 &num_palette);
1352 palette - the palette for the file
1353 (array of png_color)
1355 num_palette - number of entries in the palette
1357 png_get_gAMA(png_ptr, info_ptr, &file_gamma);
1358 png_get_gAMA_fixed(png_ptr, info_ptr, &int_file_gamma);
1360 file_gamma - the gamma at which the file is
1361 written (PNG_INFO_gAMA)
1363 int_file_gamma - 100,000 times the gamma at which the
1364 file is written
1366 png_get_cHRM(png_ptr, info_ptr, &white_x, &white_y, &red_x,
1367 &red_y, &green_x, &green_y, &blue_x, &blue_y)
1368 png_get_cHRM_XYZ(png_ptr, info_ptr, &red_X, &red_Y, &red_Z,
1369 &green_X, &green_Y, &green_Z, &blue_X, &blue_Y,
1370 &blue_Z)
1371 png_get_cHRM_fixed(png_ptr, info_ptr, &int_white_x,
1372 &int_white_y, &int_red_x, &int_red_y,
1373 &int_green_x, &int_green_y, &int_blue_x,
1374 &int_blue_y)
1375 png_get_cHRM_XYZ_fixed(png_ptr, info_ptr, &int_red_X, &int_red_Y,
1376 &int_red_Z, &int_green_X, &int_green_Y,
1377 &int_green_Z, &int_blue_X, &int_blue_Y,
1378 &int_blue_Z)
1380 {white,red,green,blue}_{x,y}
1381 A color space encoding specified using the
1382 chromaticities of the end points and the
1383 white point. (PNG_INFO_cHRM)
1385 {red,green,blue}_{X,Y,Z}
1386 A color space encoding specified using the
1387 encoding end points - the CIE tristimulus
1388 specification of the intended color of the red,
1389 green and blue channels in the PNG RGB data.
1390 The white point is simply the sum of the three
1391 end points. (PNG_INFO_cHRM)
1393 png_get_sRGB(png_ptr, info_ptr, &srgb_intent);
1395 srgb_intent - the rendering intent (PNG_INFO_sRGB)
1396 The presence of the sRGB chunk
1397 means that the pixel data is in the
1398 sRGB color space. This chunk also
1399 implies specific values of gAMA and
1400 cHRM.
1402 png_get_iCCP(png_ptr, info_ptr, &name,
1403 &compression_type, &profile, &proflen);
1405 name - The profile name.
1407 compression_type - The compression type; always
1409 You may give NULL to this argument to
1410 ignore it.
1412 profile - International Color Consortium color
1413 profile data. May contain NULs.
1415 proflen - length of profile data in bytes.
1417 png_get_sBIT(png_ptr, info_ptr, &sig_bit);
1419 sig_bit - the number of significant bits for
1420 (PNG_INFO_sBIT) each of the gray,
1421 red, green, and blue channels,
1422 whichever are appropriate for the
1423 given color type (png_color_16)
1425 png_get_tRNS(png_ptr, info_ptr, &trans_alpha,
1426 &num_trans, &trans_color);
1428 trans_alpha - array of alpha (transparency)
1429 entries for palette (PNG_INFO_tRNS)
1431 num_trans - number of transparent entries
1432 (PNG_INFO_tRNS)
1434 trans_color - graylevel or color sample values of
1435 the single transparent color for
1436 non-paletted images (PNG_INFO_tRNS)
1438 png_get_hIST(png_ptr, info_ptr, &hist);
1439 (PNG_INFO_hIST)
1441 hist - histogram of palette (array of
1442 png_uint_16)
1444 png_get_tIME(png_ptr, info_ptr, &mod_time);
1446 mod_time - time image was last modified
1449 png_get_bKGD(png_ptr, info_ptr, &background);
1451 background - background color (of type
1452 png_color_16p) (PNG_VALID_bKGD)
1453 valid 16-bit red, green and blue
1454 values, regardless of color_type
1456 num_comments = png_get_text(png_ptr, info_ptr,
1457 &text_ptr, &num_text);
1459 num_comments - number of comments
1461 text_ptr - array of png_text holding image
1464 text_ptr[i].compression - type of compression used
1470 text_ptr[i].key - keyword for comment. Must contain
1471 1-79 characters.
1473 text_ptr[i].text - text comments for current
1474 keyword. Can be empty.
1476 text_ptr[i].text_length - length of text string,
1477 after decompression, 0 for iTXt
1479 text_ptr[i].itxt_length - length of itxt string,
1480 after decompression, 0 for tEXt/zTXt
1482 text_ptr[i].lang - language of comment (empty
1483 string for unknown).
1485 text_ptr[i].lang_key - keyword in UTF-8
1486 (empty string for unknown).
1488 Note that the itxt_length, lang, and lang_key
1489 members of the text_ptr structure only exist when the
1490 library is built with iTXt chunk support. Prior to
1491 libpng-1.4.0 the library was built by default without
1492 iTXt support. Also note that when iTXt is supported,
1493 they contain NULL pointers when the "compression"
1494 field contains PNG_TEXT_COMPRESSION_NONE or
1497 num_text - number of comments (same as
1498 num_comments; you can put NULL here
1499 to avoid the duplication)
1501 Note while png_set_text() will accept text, language,
1502 and translated keywords that can be NULL pointers, the
1503 structure returned by png_get_text will always contain
1504 regular zero-terminated C strings. They might be
1505 empty strings but they will never be NULL pointers.
1507 num_spalettes = png_get_sPLT(png_ptr, info_ptr,
1508 &palette_ptr);
1510 num_spalettes - number of sPLT chunks read.
1512 palette_ptr - array of palette structures holding
1513 contents of one or more sPLT chunks
1514 read.
1516 png_get_oFFs(png_ptr, info_ptr, &offset_x, &offset_y,
1517 &unit_type);
1519 offset_x - positive offset from the left edge
1520 of the screen (can be negative)
1522 offset_y - positive offset from the top edge
1523 of the screen (can be negative)
1527 png_get_pHYs(png_ptr, info_ptr, &res_x, &res_y,
1528 &unit_type);
1530 res_x - pixels/unit physical resolution in
1531 x direction
1533 res_y - pixels/unit physical resolution in
1534 x direction
1536 unit_type - PNG_RESOLUTION_UNKNOWN,
1539 png_get_sCAL(png_ptr, info_ptr, &unit, &width,
1540 &height)
1542 unit - physical scale units (an integer)
1544 width - width of a pixel in physical scale units
1546 height - height of a pixel in physical scale units
1547 (width and height are doubles)
1549 png_get_sCAL_s(png_ptr, info_ptr, &unit, &width,
1550 &height)
1552 unit - physical scale units (an integer)
1554 width - width of a pixel in physical scale units
1555 (expressed as a string)
1557 height - height of a pixel in physical scale units
1558 (width and height are strings like "2.54")
1560 num_unknown_chunks = png_get_unknown_chunks(png_ptr,
1561 info_ptr, &unknowns)
1563 unknowns - array of png_unknown_chunk
1564 structures holding unknown chunks
1566 unknowns[i].name - name of unknown chunk
1568 unknowns[i].data - data of unknown chunk
1570 unknowns[i].size - size of unknown chunk's data
1572 unknowns[i].location - position of chunk in file
1574 The value of "i" corresponds to the order in which the
1575 chunks were read from the PNG file or inserted with the
1576 png_set_unknown_chunks() function.
1578 The value of "location" is a bitwise "or" of
1580 PNG_HAVE_IHDR (0x01)
1581 PNG_HAVE_PLTE (0x02)
1582 PNG_AFTER_IDAT (0x08)
1584 The data from the pHYs chunk can be retrieved in several convenient
1585 forms:
1587 res_x = png_get_x_pixels_per_meter(png_ptr,
1588 info_ptr)
1590 res_y = png_get_y_pixels_per_meter(png_ptr,
1591 info_ptr)
1593 res_x_and_y = png_get_pixels_per_meter(png_ptr,
1594 info_ptr)
1596 res_x = png_get_x_pixels_per_inch(png_ptr,
1597 info_ptr)
1599 res_y = png_get_y_pixels_per_inch(png_ptr,
1600 info_ptr)
1602 res_x_and_y = png_get_pixels_per_inch(png_ptr,
1603 info_ptr)
1605 aspect_ratio = png_get_pixel_aspect_ratio(png_ptr,
1606 info_ptr)
1608 Each of these returns 0 [signifying "unknown"] if
1609 the data is not present or if res_x is 0;
1610 res_x_and_y is 0 if res_x != res_y
1612 Note that because of the way the resolutions are
1613 stored internally, the inch conversions won't
1614 come out to exactly even number. For example,
1615 72 dpi is stored as 0.28346 pixels/meter, and
1616 when this is retrieved it is 71.9988 dpi, so
1617 be sure to round the returned value appropriately
1618 if you want to display a reasonable-looking result.
1620 The data from the oFFs chunk can be retrieved in several convenient
1621 forms:
1623 x_offset = png_get_x_offset_microns(png_ptr, info_ptr);
1625 y_offset = png_get_y_offset_microns(png_ptr, info_ptr);
1627 x_offset = png_get_x_offset_inches(png_ptr, info_ptr);
1629 y_offset = png_get_y_offset_inches(png_ptr, info_ptr);
1631 Each of these returns 0 [signifying "unknown" if both
1632 x and y are 0] if the data is not present or if the
1633 chunk is present but the unit is the pixel. The
1634 remark about inexact inch conversions applies here
1635 as well, because a value in inches can't always be
1636 converted to microns and back without some loss
1637 of precision.
1639 For more information, see the
1640 PNG specification for chunk contents. Be careful with trusting
1641 rowbytes, as some of the transformations could increase the space
1642 needed to hold a row (expand, filler, gray_to_rgb, etc.).
1643 See png_read_update_info(), below.
1645 A quick word about text_ptr and num_text. PNG stores comments in
1646 keyword/text pairs, one pair per chunk, with no limit on the number
1647 of text chunks, and a 2^31 byte limit on their size. While there are
1648 suggested keywords, there is no requirement to restrict the use to these
1649 strings. It is strongly suggested that keywords and text be sensible
1650 to humans (that's the point), so don't use abbreviations. Non-printing
1651 symbols are not allowed. See the PNG specification for more details.
1652 There is also no requirement to have text after the keyword.
1654 Keywords should be limited to 79 Latin-1 characters without leading or
1655 trailing spaces, but non-consecutive spaces are allowed within the
1656 keyword. It is possible to have the same keyword any number of times.
1657 The text_ptr is an array of png_text structures, each holding a
1658 pointer to a language string, a pointer to a keyword and a pointer to
1659 a text string. The text string, language code, and translated
1660 keyword may be empty or NULL pointers. The keyword/text
1661 pairs are put into the array in the order that they are received.
1662 However, some or all of the text chunks may be after the image, so, to
1663 make sure you have read all the text chunks, don't mess with these
1664 until after you read the stuff after the image. This will be
1665 mentioned again below in the discussion that goes with png_read_end().
1667 Input transformations
1669 After you've read the header information, you can set up the library
1670 to handle any special transformations of the image data. The various
1671 ways to transform the data will be described in the order that they
1672 should occur. This is important, as some of these change the color
1673 type and/or bit depth of the data, and some others only work on
1674 certain color types and bit depths.
1676 Transformations you request are ignored if they don't have any meaning for a
1677 particular input data format. However some transformations can have an effect
1678 as a result of a previous transformation. If you specify a contradictory set of
1679 transformations, for example both adding and removing the alpha channel, you
1680 cannot predict the final result.
1682 The color used for the transparency values should be supplied in the same
1683 format/depth as the current image data. It is stored in the same format/depth
1684 as the image data in a tRNS chunk, so this is what libpng expects for this data.
1686 The color used for the background value depends on the need_expand argument as
1687 described below.
1689 Data will be decoded into the supplied row buffers packed into bytes
1690 unless the library has been told to transform it into another format.
1691 For example, 4 bit/pixel paletted or grayscale data will be returned
1692 2 pixels/byte with the leftmost pixel in the high-order bits of the byte,
1693 unless png_set_packing() is called. 8-bit RGB data will be stored
1694 in RGB RGB RGB format unless png_set_filler() or png_set_add_alpha()
1695 is called to insert filler bytes, either before or after each RGB triplet.
1697 16-bit RGB data will be returned RRGGBB RRGGBB, with the most significant
1698 byte of the color value first, unless png_set_scale_16() is called to
1699 transform it to regular RGB RGB triplets, or png_set_filler() or
1700 png_set_add alpha() is called to insert two filler bytes, either before
1701 or after each RRGGBB triplet. Similarly, 8-bit or 16-bit grayscale data can
1702 be modified with png_set_filler(), png_set_add_alpha(), png_set_strip_16(),
1703 or png_set_scale_16().
1705 The following code transforms grayscale images of less than 8 to 8 bits,
1706 changes paletted images to RGB, and adds a full alpha channel if there is
1707 transparency information in a tRNS chunk. This is most useful on
1708 grayscale images with bit depths of 2 or 4 or if there is a multiple-image
1709 viewing application that wishes to treat all images in the same way.
1711 if (color_type == PNG_COLOR_TYPE_PALETTE)
1712 png_set_palette_to_rgb(png_ptr);
1714 if (png_get_valid(png_ptr, info_ptr,
1715 PNG_INFO_tRNS)) png_set_tRNS_to_alpha(png_ptr);
1717 if (color_type == PNG_COLOR_TYPE_GRAY &&
1718 bit_depth < 8) png_set_expand_gray_1_2_4_to_8(png_ptr);
1720 The first two functions are actually aliases for png_set_expand(), added
1721 in libpng version 1.0.4, with the function names expanded to improve code
1722 readability. In some future version they may actually do different
1723 things.
1725 As of libpng version 1.2.9, png_set_expand_gray_1_2_4_to_8() was
1726 added. It expands the sample depth without changing tRNS to alpha.
1728 As of libpng version 1.5.2, png_set_expand_16() was added. It behaves as
1729 png_set_expand(); however, the resultant channels have 16 bits rather than 8.
1730 Use this when the output color or gray channels are made linear to avoid fairly
1731 severe accuracy loss.
1733 if (bit_depth < 16)
1734 png_set_expand_16(png_ptr);
1736 PNG can have files with 16 bits per channel. If you only can handle
1737 8 bits per channel, this will strip the pixels down to 8-bit.
1739 if (bit_depth == 16)
1740 #if PNG_LIBPNG_VER >= 10504
1741 png_set_scale_16(png_ptr);
1742 #else
1743 png_set_strip_16(png_ptr);
1744 #endif
1746 (The more accurate "png_set_scale_16()" API became available in libpng version
1747 1.5.4).
1749 If you need to process the alpha channel on the image separately from the image
1750 data (for example if you convert it to a bitmap mask) it is possible to have
1751 libpng strip the channel leaving just RGB or gray data:
1753 if (color_type & PNG_COLOR_MASK_ALPHA)
1754 png_set_strip_alpha(png_ptr);
1756 If you strip the alpha channel you need to find some other way of dealing with
1757 the information. If, instead, you want to convert the image to an opaque
1758 version with no alpha channel use png_set_background; see below.
1760 As of libpng version 1.5.2, almost all useful expansions are supported, the
1761 major ommissions are conversion of grayscale to indexed images (which can be
1762 done trivially in the application) and conversion of indexed to grayscale (which
1763 can be done by a trivial manipulation of the palette.)
1765 In the following table, the 01 means grayscale with depth<8, 31 means
1766 indexed with depth<8, other numerals represent the color type, "T" means
1767 the tRNS chunk is present, A means an alpha channel is present, and O
1768 means tRNS or alpha is present but all pixels in the image are opaque.
1770 FROM 01 31 0 0T 0O 2 2T 2O 3 3T 3O 4A 4O 6A 6O
1771 TO
1772 01 - [G] - - - - - - - - - - - - -
1773 31 [Q] Q [Q] [Q] [Q] Q Q Q Q Q Q [Q] [Q] Q Q
1774 0 1 G + . . G G G G G G B B GB GB
1775 0T lt Gt t + . Gt G G Gt G G Bt Bt GBt GBt
1776 0O lt Gt t . + Gt Gt G Gt Gt G Bt Bt GBt GBt
1777 2 C P C C C + . . C - - CB CB B B
1778 2T Ct - Ct C C t + t - - - CBt CBt Bt Bt
1779 2O Ct - Ct C C t t + - - - CBt CBt Bt Bt
1780 3 [Q] p [Q] [Q] [Q] Q Q Q + . . [Q] [Q] Q Q
1781 3T [Qt] p [Qt][Q] [Q] Qt Qt Qt t + t [Qt][Qt] Qt Qt
1782 3O [Qt] p [Qt][Q] [Q] Qt Qt Qt t t + [Qt][Qt] Qt Qt
1785 6A CA PA CA C C A T tT PA P P C CBA + BA
1786 6O CA PBA CA C C A tT T PA P P CBA C BA +
1788 Within the matrix,
1789 "+" identifies entries where 'from' and 'to' are the same.
1790 "-" means the transformation is not supported.
1791 "." means nothing is necessary (a tRNS chunk can just be ignored).
1792 "t" means the transformation is obtained by png_set_tRNS.
1793 "A" means the transformation is obtained by png_set_add_alpha().
1794 "X" means the transformation is obtained by png_set_expand().
1795 "1" means the transformation is obtained by
1796 png_set_expand_gray_1_2_4_to_8() (and by png_set_expand()
1797 if there is no transparency in the original or the final
1798 format).
1799 "C" means the transformation is obtained by png_set_gray_to_rgb().
1800 "G" means the transformation is obtained by png_set_rgb_to_gray().
1801 "P" means the transformation is obtained by
1802 png_set_expand_palette_to_rgb().
1803 "p" means the transformation is obtained by png_set_packing().
1804 "Q" means the transformation is obtained by png_set_quantize().
1805 "T" means the transformation is obtained by
1806 png_set_tRNS_to_alpha().
1807 "B" means the transformation is obtained by
1808 png_set_background(), or png_strip_alpha().
1810 When an entry has multiple transforms listed all are required to cause the
1811 right overall transformation. When two transforms are separated by a comma
1812 either will do the job. When transforms are enclosed in [] the transform should
1813 do the job but this is currently unimplemented - a different format will result
1814 if the suggested transformations are used.
1816 In PNG files, the alpha channel in an image
1817 is the level of opacity. If you need the alpha channel in an image to
1818 be the level of transparency instead of opacity, you can invert the
1819 alpha channel (or the tRNS chunk data) after it's read, so that 0 is
1820 fully opaque and 255 (in 8-bit or paletted images) or 65535 (in 16-bit
1821 images) is fully transparent, with
1823 png_set_invert_alpha(png_ptr);
1825 PNG files pack pixels of bit depths 1, 2, and 4 into bytes as small as
1826 they can, resulting in, for example, 8 pixels per byte for 1 bit
1827 files. This code expands to 1 pixel per byte without changing the
1828 values of the pixels:
1830 if (bit_depth < 8)
1831 png_set_packing(png_ptr);
1833 PNG files have possible bit depths of 1, 2, 4, 8, and 16. All pixels
1834 stored in a PNG image have been "scaled" or "shifted" up to the next
1835 higher possible bit depth (e.g. from 5 bits/sample in the range [0,31]
1836 to 8 bits/sample in the range [0, 255]). However, it is also possible
1837 to convert the PNG pixel data back to the original bit depth of the
1838 image. This call reduces the pixels back down to the original bit depth:
1840 png_color_8p sig_bit;
1842 if (png_get_sBIT(png_ptr, info_ptr, &sig_bit))
1843 png_set_shift(png_ptr, sig_bit);
1845 PNG files store 3-color pixels in red, green, blue order. This code
1846 changes the storage of the pixels to blue, green, red:
1848 if (color_type == PNG_COLOR_TYPE_RGB ||
1849 color_type == PNG_COLOR_TYPE_RGB_ALPHA)
1850 png_set_bgr(png_ptr);
1852 PNG files store RGB pixels packed into 3 or 6 bytes. This code expands them
1853 into 4 or 8 bytes for windowing systems that need them in this format:
1855 if (color_type == PNG_COLOR_TYPE_RGB)
1856 png_set_filler(png_ptr, filler, PNG_FILLER_BEFORE);
1858 where "filler" is the 8-bit or 16-bit number to fill with, and the location
1859 is either PNG_FILLER_BEFORE or PNG_FILLER_AFTER, depending upon whether
1860 you want the filler before the RGB or after. When filling an 8-bit pixel,
1861 the least significant 8 bits of the number are used, if a 16-bit number is
1862 supplied. This transformation does not affect images that already have full
1863 alpha channels. To add an opaque alpha channel, use filler=0xffff and
1864 PNG_FILLER_AFTER which will generate RGBA pixels.
1866 Note that png_set_filler() does not change the color type. If you want
1867 to do that, you can add a true alpha channel with
1869 if (color_type == PNG_COLOR_TYPE_RGB ||
1870 color_type == PNG_COLOR_TYPE_GRAY)
1871 png_set_add_alpha(png_ptr, filler, PNG_FILLER_AFTER);
1873 where "filler" contains the alpha value to assign to each pixel.
1874 The png_set_add_alpha() function was added in libpng-1.2.7.
1876 If you are reading an image with an alpha channel, and you need the
1877 data as ARGB instead of the normal PNG format RGBA:
1879 if (color_type == PNG_COLOR_TYPE_RGB_ALPHA)
1880 png_set_swap_alpha(png_ptr);
1882 For some uses, you may want a grayscale image to be represented as
1883 RGB. This code will do that conversion:
1885 if (color_type == PNG_COLOR_TYPE_GRAY ||
1886 color_type == PNG_COLOR_TYPE_GRAY_ALPHA)
1887 png_set_gray_to_rgb(png_ptr);
1889 Conversely, you can convert an RGB or RGBA image to grayscale or grayscale
1890 with alpha.
1892 if (color_type == PNG_COLOR_TYPE_RGB ||
1893 color_type == PNG_COLOR_TYPE_RGB_ALPHA)
1894 png_set_rgb_to_gray(png_ptr, error_action,
1895 double red_weight, double green_weight);
1897 error_action = 1: silently do the conversion
1899 error_action = 2: issue a warning if the original
1900 image has any pixel where
1901 red != green or red != blue
1903 error_action = 3: issue an error and abort the
1904 conversion if the original
1905 image has any pixel where
1906 red != green or red != blue
1908 red_weight: weight of red component
1910 green_weight: weight of green component
1911 If either weight is negative, default
1912 weights are used.
1914 In the corresponding fixed point API the red_weight and green_weight values are
1915 simply scaled by 100,000:
1917 png_set_rgb_to_gray(png_ptr, error_action,
1918 png_fixed_point red_weight,
1919 png_fixed_point green_weight);
1921 If you have set error_action = 1 or 2, you can
1922 later check whether the image really was gray, after processing
1923 the image rows, with the png_get_rgb_to_gray_status(png_ptr) function.
1924 It will return a png_byte that is zero if the image was gray or
1925 1 if there were any non-gray pixels. Background and sBIT data
1926 will be silently converted to grayscale, using the green channel
1927 data for sBIT, regardless of the error_action setting.
1929 The default values come from the PNG file cHRM chunk if present; otherwise, the
1930 defaults correspond to the ITU-R recommendation 709, and also the sRGB color
1931 space, as recommended in the Charles Poynton's Colour FAQ,
1932 Copyright (c) 2006-11-28 Charles Poynton, in section 9:
1934 <http://www.poynton.com/notes/colour_and_gamma/ColorFAQ.html#RTFToC9>
1936 Y = 0.2126 * R + 0.7152 * G + 0.0722 * B
1938 Previous versions of this document, 1998 through 2002, recommended a slightly
1939 different formula:
1941 Y = 0.212671 * R + 0.715160 * G + 0.072169 * B
1943 Libpng uses an integer approximation:
1945 Y = (6968 * R + 23434 * G + 2366 * B)/32768
1947 The calculation is done in a linear colorspace, if the image gamma
1948 can be determined.
1950 The png_set_background() function has been described already; it tells libpng to
1951 composite images with alpha or simple transparency against the supplied
1952 background color. For compatibility with versions of libpng earlier than
1953 libpng-1.5.4 it is recommended that you call the function after reading the file
1954 header, even if you don't want to use the color in a bKGD chunk, if one exists.
1956 If the PNG file contains a bKGD chunk (PNG_INFO_bKGD valid),
1957 you may use this color, or supply another color more suitable for
1958 the current display (e.g., the background color from a web page). You
1959 need to tell libpng how the color is represented, both the format of the
1960 component values in the color (the number of bits) and the gamma encoding of the
1961 color. The function takes two arguments, background_gamma_mode and need_expand
1962 to convey this information; however, only two combinations are likely to be
1963 useful:
1965 png_color_16 my_background;
1966 png_color_16p image_background;
1968 if (png_get_bKGD(png_ptr, info_ptr, &image_background))
1969 png_set_background(png_ptr, image_background,
1970 PNG_BACKGROUND_GAMMA_FILE, 1/*needs to be expanded*/, 1);
1971 else
1972 png_set_background(png_ptr, &my_background,
1973 PNG_BACKGROUND_GAMMA_SCREEN, 0/*do not expand*/, 1);
1975 The second call was described above - my_background is in the format of the
1976 final, display, output produced by libpng. Because you now know the format of
1977 the PNG it is possible to avoid the need to choose either 8-bit or 16-bit
1978 output and to retain palette images (the palette colors will be modified
1979 appropriately and the tRNS chunk removed.) However, if you are doing this,
1980 take great care not to ask for transformations without checking first that
1981 they apply!
1983 In the first call the background color has the original bit depth and color type
1984 of the PNG file. So, for palette images the color is supplied as a palette
1985 index and for low bit greyscale images the color is a reduced bit value in
1986 image_background->gray.
1988 If you didn't call png_set_gamma() before reading the file header, for example
1989 if you need your code to remain compatible with older versions of libpng prior
1990 to libpng-1.5.4, this is the place to call it.
1992 Do not call it if you called png_set_alpha_mode(); doing so will damage the
1993 settings put in place by png_set_alpha_mode(). (If png_set_alpha_mode() is
1994 supported then you can certainly do png_set_gamma() before reading the PNG
1995 header.)
1997 This API unconditionally sets the screen and file gamma values, so it will
1998 override the value in the PNG file unless it is called before the PNG file
1999 reading starts. For this reason you must always call it with the PNG file
2000 value when you call it in this position:
2002 if (png_get_gAMA(png_ptr, info_ptr, &file_gamma))
2003 png_set_gamma(png_ptr, screen_gamma, file_gamma);
2005 else
2006 png_set_gamma(png_ptr, screen_gamma, 0.45455);
2008 If you need to reduce an RGB file to a paletted file, or if a paletted
2009 file has more entries than will fit on your screen, png_set_quantize()
2010 will do that. Note that this is a simple match quantization that merely
2011 finds the closest color available. This should work fairly well with
2012 optimized palettes, but fairly badly with linear color cubes. If you
2013 pass a palette that is larger than maximum_colors, the file will
2014 reduce the number of colors in the palette so it will fit into
2015 maximum_colors. If there is a histogram, libpng will use it to make
2016 more intelligent choices when reducing the palette. If there is no
2017 histogram, it may not do as good a job.
2019 if (color_type & PNG_COLOR_MASK_COLOR)
2020 {
2021 if (png_get_valid(png_ptr, info_ptr,
2023 {
2024 png_uint_16p histogram = NULL;
2026 png_get_hIST(png_ptr, info_ptr,
2027 &histogram);
2028 png_set_quantize(png_ptr, palette, num_palette,
2029 max_screen_colors, histogram, 1);
2030 }
2032 else
2033 {
2034 png_color std_color_cube[MAX_SCREEN_COLORS] =
2035 { ... colors ... };
2037 png_set_quantize(png_ptr, std_color_cube,
2039 NULL,0);
2040 }
2041 }
2043 PNG files describe monochrome as black being zero and white being one.
2044 The following code will reverse this (make black be one and white be
2045 zero):
2047 if (bit_depth == 1 && color_type == PNG_COLOR_TYPE_GRAY)
2048 png_set_invert_mono(png_ptr);
2050 This function can also be used to invert grayscale and gray-alpha images:
2052 if (color_type == PNG_COLOR_TYPE_GRAY ||
2053 color_type == PNG_COLOR_TYPE_GRAY_ALPHA)
2054 png_set_invert_mono(png_ptr);
2056 PNG files store 16-bit pixels in network byte order (big-endian,
2057 ie. most significant bits first). This code changes the storage to the
2058 other way (little-endian, i.e. least significant bits first, the
2059 way PCs store them):
2061 if (bit_depth == 16)
2062 png_set_swap(png_ptr);
2064 If you are using packed-pixel images (1, 2, or 4 bits/pixel), and you
2065 need to change the order the pixels are packed into bytes, you can use:
2067 if (bit_depth < 8)
2068 png_set_packswap(png_ptr);
2070 Finally, you can write your own transformation function if none of
2071 the existing ones meets your needs. This is done by setting a callback
2072 with
2074 png_set_read_user_transform_fn(png_ptr,
2075 read_transform_fn);
2077 You must supply the function
2079 void read_transform_fn(png_structp png_ptr, png_row_infop
2080 row_info, png_bytep data)
2082 See pngtest.c for a working example. Your function will be called
2083 after all of the other transformations have been processed. Take care with
2084 interlaced images if you do the interlace yourself - the width of the row is the
2085 width in 'row_info', not the overall image width.
2087 If supported, libpng provides two information routines that you can use to find
2088 where you are in processing the image:
2090 png_get_current_pass_number(png_structp png_ptr);
2091 png_get_current_row_number(png_structp png_ptr);
2093 Don't try using these outside a transform callback - firstly they are only
2094 supported if user transforms are supported, secondly they may well return
2095 unexpected results unless the row is actually being processed at the moment they
2096 are called.
2098 With interlaced
2099 images the value returned is the row in the input sub-image image. Use
2100 PNG_ROW_FROM_PASS_ROW(row, pass) and PNG_COL_FROM_PASS_COL(col, pass) to
2101 find the output pixel (x,y) given an interlaced sub-image pixel (row,col,pass).
2103 The discussion of interlace handling above contains more information on how to
2104 use these values.
2106 You can also set up a pointer to a user structure for use by your
2107 callback function, and you can inform libpng that your transform
2108 function will change the number of channels or bit depth with the
2109 function
2111 png_set_user_transform_info(png_ptr, user_ptr,
2112 user_depth, user_channels);
2114 The user's application, not libpng, is responsible for allocating and
2115 freeing any memory required for the user structure.
2117 You can retrieve the pointer via the function
2118 png_get_user_transform_ptr(). For example:
2120 voidp read_user_transform_ptr =
2121 png_get_user_transform_ptr(png_ptr);
2123 The last thing to handle is interlacing; this is covered in detail below,
2124 but you must call the function here if you want libpng to handle expansion
2125 of the interlaced image.
2127 number_of_passes = png_set_interlace_handling(png_ptr);
2129 After setting the transformations, libpng can update your png_info
2130 structure to reflect any transformations you've requested with this
2131 call.
2133 png_read_update_info(png_ptr, info_ptr);
2135 This is most useful to update the info structure's rowbytes
2136 field so you can use it to allocate your image memory. This function
2137 will also update your palette with the correct screen_gamma and
2138 background if these have been given with the calls above. You may
2139 only call png_read_update_info() once with a particular info_ptr.
2141 After you call png_read_update_info(), you can allocate any
2142 memory you need to hold the image. The row data is simply
2143 raw byte data for all forms of images. As the actual allocation
2144 varies among applications, no example will be given. If you
2145 are allocating one large chunk, you will need to build an
2146 array of pointers to each row, as it will be needed for some
2147 of the functions below.
2149 Remember: Before you call png_read_update_info(), the png_get_*()
2150 functions return the values corresponding to the original PNG image.
2151 After you call png_read_update_info the values refer to the image
2152 that libpng will output. Consequently you must call all the png_set_
2153 functions before you call png_read_update_info(). This is particularly
2154 important for png_set_interlace_handling() - if you are going to call
2155 png_read_update_info() you must call png_set_interlace_handling() before
2156 it unless you want to receive interlaced output.
2158 Reading image data
2160 After you've allocated memory, you can read the image data.
2161 The simplest way to do this is in one function call. If you are
2162 allocating enough memory to hold the whole image, you can just
2163 call png_read_image() and libpng will read in all the image data
2164 and put it in the memory area supplied. You will need to pass in
2165 an array of pointers to each row.
2167 This function automatically handles interlacing, so you don't
2168 need to call png_set_interlace_handling() (unless you call
2169 png_read_update_info()) or call this function multiple times, or any
2170 of that other stuff necessary with png_read_rows().
2172 png_read_image(png_ptr, row_pointers);
2174 where row_pointers is:
2176 png_bytep row_pointers[height];
2178 You can point to void or char or whatever you use for pixels.
2180 If you don't want to read in the whole image at once, you can
2181 use png_read_rows() instead. If there is no interlacing (check
2182 interlace_type == PNG_INTERLACE_NONE), this is simple:
2184 png_read_rows(png_ptr, row_pointers, NULL,
2185 number_of_rows);
2187 where row_pointers is the same as in the png_read_image() call.
2189 If you are doing this just one row at a time, you can do this with
2190 a single row_pointer instead of an array of row_pointers:
2192 png_bytep row_pointer = row;
2193 png_read_row(png_ptr, row_pointer, NULL);
2195 If the file is interlaced (interlace_type != 0 in the IHDR chunk), things
2196 get somewhat harder. The only current (PNG Specification version 1.2)
2197 interlacing type for PNG is (interlace_type == PNG_INTERLACE_ADAM7);
2198 a somewhat complicated 2D interlace scheme, known as Adam7, that
2199 breaks down an image into seven smaller images of varying size, based
2200 on an 8x8 grid. This number is defined (from libpng 1.5) as
2203 libpng can fill out those images or it can give them to you "as is".
2204 It is almost always better to have libpng handle the interlacing for you.
2205 If you want the images filled out, there are two ways to do that. The one
2206 mentioned in the PNG specification is to expand each pixel to cover
2207 those pixels that have not been read yet (the "rectangle" method).
2208 This results in a blocky image for the first pass, which gradually
2209 smooths out as more pixels are read. The other method is the "sparkle"
2210 method, where pixels are drawn only in their final locations, with the
2211 rest of the image remaining whatever colors they were initialized to
2212 before the start of the read. The first method usually looks better,
2213 but tends to be slower, as there are more pixels to put in the rows.
2215 If, as is likely, you want libpng to expand the images, call this before
2216 calling png_start_read_image() or png_read_update_info():
2218 if (interlace_type == PNG_INTERLACE_ADAM7)
2219 number_of_passes
2220 = png_set_interlace_handling(png_ptr);
2222 This will return the number of passes needed. Currently, this is seven,
2223 but may change if another interlace type is added. This function can be
2224 called even if the file is not interlaced, where it will return one pass.
2225 You then need to read the whole image 'number_of_passes' times. Each time
2226 will distribute the pixels from the current pass to the correct place in
2227 the output image, so you need to supply the same rows to png_read_rows in
2228 each pass.
2230 If you are not going to display the image after each pass, but are
2231 going to wait until the entire image is read in, use the sparkle
2232 effect. This effect is faster and the end result of either method
2233 is exactly the same. If you are planning on displaying the image
2234 after each pass, the "rectangle" effect is generally considered the
2235 better looking one.
2237 If you only want the "sparkle" effect, just call png_read_row() or
2238 png_read_rows() as
2239 normal, with the third parameter NULL. Make sure you make pass over
2240 the image number_of_passes times, and you don't change the data in the
2241 rows between calls. You can change the locations of the data, just
2242 not the data. Each pass only writes the pixels appropriate for that
2243 pass, and assumes the data from previous passes is still valid.
2245 png_read_rows(png_ptr, row_pointers, NULL,
2246 number_of_rows);
2247 or
2248 png_read_row(png_ptr, row_pointers, NULL);
2250 If you only want the first effect (the rectangles), do the same as
2251 before except pass the row buffer in the third parameter, and leave
2252 the second parameter NULL.
2254 png_read_rows(png_ptr, NULL, row_pointers,
2255 number_of_rows);
2256 or
2257 png_read_row(png_ptr, NULL, row_pointers);
2259 If you don't want libpng to handle the interlacing details, just call
2260 png_read_rows() PNG_INTERLACE_ADAM7_PASSES times to read in all the images.
2261 Each of the images is a valid image by itself; however, you will almost
2262 certainly need to distribute the pixels from each sub-image to the
2263 correct place. This is where everything gets very tricky.
2265 If you want to retrieve the separate images you must pass the correct
2266 number of rows to each successive call of png_read_rows(). The calculation
2267 gets pretty complicated for small images, where some sub-images may
2268 not even exist because either their width or height ends up zero.
2269 libpng provides two macros to help you in 1.5 and later versions:
2271 png_uint_32 width = PNG_PASS_COLS(image_width, pass_number);
2272 png_uint_32 height = PNG_PASS_ROWS(image_height, pass_number);
2274 Respectively these tell you the width and height of the sub-image
2275 corresponding to the numbered pass. 'pass' is in in the range 0 to 6 -
2276 this can be confusing because the specification refers to the same passes
2277 as 1 to 7! Be careful, you must check both the width and height before
2278 calling png_read_rows() and not call it for that pass if either is zero.
2280 You can, of course, read each sub-image row by row. If you want to
2281 produce optimal code to make a pixel-by-pixel transformation of an
2282 interlaced image this is the best approach; read each row of each pass,
2283 transform it, and write it out to a new interlaced image.
2285 If you want to de-interlace the image yourself libpng provides further
2286 macros to help that tell you where to place the pixels in the output image.
2287 Because the interlacing scheme is rectangular - sub-image pixels are always
2288 arranged on a rectangular grid - all you need to know for each pass is the
2289 starting column and row in the output image of the first pixel plus the
2290 spacing between each pixel. As of libpng 1.5 there are four macros to
2291 retrieve this information:
2293 png_uint_32 x = PNG_PASS_START_COL(pass);
2294 png_uint_32 y = PNG_PASS_START_ROW(pass);
2295 png_uint_32 xStep = 1U << PNG_PASS_COL_SHIFT(pass);
2296 png_uint_32 yStep = 1U << PNG_PASS_ROW_SHIFT(pass);
2298 These allow you to write the obvious loop:
2300 png_uint_32 input_y = 0;
2301 png_uint_32 output_y = PNG_PASS_START_ROW(pass);
2303 while (output_y < output_image_height)
2304 {
2305 png_uint_32 input_x = 0;
2306 png_uint_32 output_x = PNG_PASS_START_COL(pass);
2308 while (output_x < output_image_width)
2309 {
2310 image[output_y][output_x] =
2311 subimage[pass][input_y][input_x++];
2313 output_x += xStep;
2314 }
2316 ++input_y;
2317 output_y += yStep;
2318 }
2320 Notice that the steps between successive output rows and columns are
2321 returned as shifts. This is possible because the pixels in the subimages
2322 are always a power of 2 apart - 1, 2, 4 or 8 pixels - in the original
2323 image. In practice you may need to directly calculate the output coordinate
2324 given an input coordinate. libpng provides two further macros for this
2325 purpose:
2327 png_uint_32 output_x = PNG_COL_FROM_PASS_COL(input_x, pass);
2328 png_uint_32 output_y = PNG_ROW_FROM_PASS_ROW(input_y, pass);
2330 Finally a pair of macros are provided to tell you if a particular image
2331 row or column appears in a given pass:
2333 int col_in_pass = PNG_COL_IN_INTERLACE_PASS(output_x, pass);
2334 int row_in_pass = PNG_ROW_IN_INTERLACE_PASS(output_y, pass);
2336 Bear in mind that you will probably also need to check the width and height
2337 of the pass in addition to the above to be sure the pass even exists!
2339 With any luck you are convinced by now that you don't want to do your own
2340 interlace handling. In reality normally the only good reason for doing this
2341 is if you are processing PNG files on a pixel-by-pixel basis and don't want
2342 to load the whole file into memory when it is interlaced.
2344 libpng includes a test program, pngvalid, that illustrates reading and
2345 writing of interlaced images. If you can't get interlacing to work in your
2346 code and don't want to leave it to libpng (the recommended approach), see
2347 how pngvalid.c does it.
2349 Finishing a sequential read
2351 After you are finished reading the image through the
2352 low-level interface, you can finish reading the file.
2354 If you want to use a different crc action for handling CRC errors in
2355 chunks after the image data, you can call png_set_crc_action()
2356 again at this point.
2358 If you are interested in comments or time, which may be stored either
2359 before or after the image data, you should pass the separate png_info
2360 struct if you want to keep the comments from before and after the image
2361 separate.
2363 png_infop end_info = png_create_info_struct(png_ptr);
2365 if (!end_info)
2366 {
2367 png_destroy_read_struct(&png_ptr, &info_ptr,
2368 (png_infopp)NULL);
2369 return (ERROR);
2370 }
2372 png_read_end(png_ptr, end_info);
2374 If you are not interested, you should still call png_read_end()
2375 but you can pass NULL, avoiding the need to create an end_info structure.
2376 If you do this, libpng will not process any chunks after IDAT other than
2377 skipping over them and perhaps (depending on whether you have called
2378 png_set_crc_action) checking their CRCs while looking for the IEND chunk.
2380 png_read_end(png_ptr, (png_infop)NULL);
2382 If you don't call png_read_end(), then your file pointer will be
2383 left pointing to the first chunk after the last IDAT, which is probably
2384 not what you want if you expect to read something beyond the end of
2385 the PNG datastream.
2387 When you are done, you can free all memory allocated by libpng like this:
2389 png_destroy_read_struct(&png_ptr, &info_ptr,
2390 &end_info);
2392 or, if you didn't create an end_info structure,
2394 png_destroy_read_struct(&png_ptr, &info_ptr,
2395 (png_infopp)NULL);
2397 It is also possible to individually free the info_ptr members that
2398 point to libpng-allocated storage with the following function:
2400 png_free_data(png_ptr, info_ptr, mask, seq)
2402 mask - identifies data to be freed, a mask
2403 containing the bitwise OR of one or
2404 more of
2410 or simply PNG_FREE_ALL
2412 seq - sequence number of item to be freed
2413 (-1 for all items)
2415 This function may be safely called when the relevant storage has
2416 already been freed, or has not yet been allocated, or was allocated
2417 by the user and not by libpng, and will in those cases do nothing.
2418 The "seq" parameter is ignored if only one item of the selected data
2419 type, such as PLTE, is allowed. If "seq" is not -1, and multiple items
2420 are allowed for the data type identified in the mask, such as text or
2421 sPLT, only the n'th item in the structure is freed, where n is "seq".
2423 The default behavior is only to free data that was allocated internally
2424 by libpng. This can be changed, so that libpng will not free the data,
2425 or so that it will free data that was allocated by the user with png_malloc()
2426 or png_calloc() and passed in via a png_set_*() function, with
2428 png_data_freer(png_ptr, info_ptr, freer, mask)
2430 freer - one of
2435 mask - which data elements are affected
2436 same choices as in png_free_data()
2438 This function only affects data that has already been allocated.
2439 You can call this function after reading the PNG data but before calling
2440 any png_set_*() functions, to control whether the user or the png_set_*()
2441 function is responsible for freeing any existing data that might be present,
2442 and again after the png_set_*() functions to control whether the user
2443 or png_destroy_*() is supposed to free the data. When the user assumes
2444 responsibility for libpng-allocated data, the application must use
2445 png_free() to free it, and when the user transfers responsibility to libpng
2446 for data that the user has allocated, the user must have used png_malloc()
2447 or png_calloc() to allocate it.
2449 If you allocated your row_pointers in a single block, as suggested above in
2450 the description of the high level read interface, you must not transfer
2451 responsibility for freeing it to the png_set_rows or png_read_destroy function,
2452 because they would also try to free the individual row_pointers[i].
2454 If you allocated text_ptr.text, text_ptr.lang, and text_ptr.translated_keyword
2455 separately, do not transfer responsibility for freeing text_ptr to libpng,
2456 because when libpng fills a png_text structure it combines these members with
2457 the key member, and png_free_data() will free only text_ptr.key. Similarly,
2458 if you transfer responsibility for free'ing text_ptr from libpng to your
2459 application, your application must not separately free those members.
2461 The png_free_data() function will turn off the "valid" flag for anything
2462 it frees. If you need to turn the flag off for a chunk that was freed by
2463 your application instead of by libpng, you can use
2465 png_set_invalid(png_ptr, info_ptr, mask);
2467 mask - identifies the chunks to be made invalid,
2468 containing the bitwise OR of one or
2469 more of
2479 For a more compact example of reading a PNG image, see the file example.c.
2481 Reading PNG files progressively
2483 The progressive reader is slightly different from the non-progressive
2484 reader. Instead of calling png_read_info(), png_read_rows(), and
2485 png_read_end(), you make one call to png_process_data(), which calls
2486 callbacks when it has the info, a row, or the end of the image. You
2487 set up these callbacks with png_set_progressive_read_fn(). You don't
2488 have to worry about the input/output functions of libpng, as you are
2489 giving the library the data directly in png_process_data(). I will
2490 assume that you have read the section on reading PNG files above,
2491 so I will only highlight the differences (although I will show
2492 all of the code).
2494 png_structp png_ptr;
2495 png_infop info_ptr;
2497 /* An example code fragment of how you would
2498 initialize the progressive reader in your
2499 application. */
2500 int
2501 initialize_png_reader()
2502 {
2503 png_ptr = png_create_read_struct
2504 (PNG_LIBPNG_VER_STRING, (png_voidp)user_error_ptr,
2505 user_error_fn, user_warning_fn);
2507 if (!png_ptr)
2508 return (ERROR);
2510 info_ptr = png_create_info_struct(png_ptr);
2512 if (!info_ptr)
2513 {
2514 png_destroy_read_struct(&png_ptr,
2515 (png_infopp)NULL, (png_infopp)NULL);
2516 return (ERROR);
2517 }
2519 if (setjmp(png_jmpbuf(png_ptr)))
2520 {
2521 png_destroy_read_struct(&png_ptr, &info_ptr,
2522 (png_infopp)NULL);
2523 return (ERROR);
2524 }
2526 /* This one's new. You can provide functions
2527 to be called when the header info is valid,
2528 when each row is completed, and when the image
2529 is finished. If you aren't using all functions,
2530 you can specify NULL parameters. Even when all
2531 three functions are NULL, you need to call
2532 png_set_progressive_read_fn(). You can use
2533 any struct as the user_ptr (cast to a void pointer
2534 for the function call), and retrieve the pointer
2535 from inside the callbacks using the function
2537 png_get_progressive_ptr(png_ptr);
2539 which will return a void pointer, which you have
2540 to cast appropriately.
2541 */
2542 png_set_progressive_read_fn(png_ptr, (void *)user_ptr,
2543 info_callback, row_callback, end_callback);
2545 return 0;
2546 }
2548 /* A code fragment that you call as you receive blocks
2549 of data */
2550 int
2551 process_data(png_bytep buffer, png_uint_32 length)
2552 {
2553 if (setjmp(png_jmpbuf(png_ptr)))
2554 {
2555 png_destroy_read_struct(&png_ptr, &info_ptr,
2556 (png_infopp)NULL);
2557 return (ERROR);
2558 }
2560 /* This one's new also. Simply give it a chunk
2561 of data from the file stream (in order, of
2562 course). On machines with segmented memory
2563 models machines, don't give it any more than
2564 64K. The library seems to run fine with sizes
2565 of 4K. Although you can give it much less if
2566 necessary (I assume you can give it chunks of
2567 1 byte, I haven't tried less than 256 bytes
2568 yet). When this function returns, you may
2569 want to display any rows that were generated
2570 in the row callback if you don't already do
2571 so there.
2572 */
2573 png_process_data(png_ptr, info_ptr, buffer, length);
2575 /* At this point you can call png_process_data_skip if
2576 you want to handle data the library will skip yourself;
2577 it simply returns the number of bytes to skip (and stops
2578 libpng skipping that number of bytes on the next
2579 png_process_data call).
2580 return 0;
2581 }
2583 /* This function is called (as set by
2584 png_set_progressive_read_fn() above) when enough data
2585 has been supplied so all of the header has been
2586 read.
2587 */
2588 void
2589 info_callback(png_structp png_ptr, png_infop info)
2590 {
2591 /* Do any setup here, including setting any of
2592 the transformations mentioned in the Reading
2593 PNG files section. For now, you _must_ call
2594 either png_start_read_image() or
2595 png_read_update_info() after all the
2596 transformations are set (even if you don't set
2597 any). You may start getting rows before
2598 png_process_data() returns, so this is your
2599 last chance to prepare for that.
2601 This is where you turn on interlace handling,
2602 assuming you don't want to do it yourself.
2604 If you need to you can stop the processing of
2605 your original input data at this point by calling
2606 png_process_data_pause. This returns the number
2607 of unprocessed bytes from the last png_process_data
2608 call - it is up to you to ensure that the next call
2609 sees these bytes again. If you don't want to bother
2610 with this you can get libpng to cache the unread
2611 bytes by setting the 'save' parameter (see png.h) but
2612 then libpng will have to copy the data internally.
2613 */
2614 }
2616 /* This function is called when each row of image
2617 data is complete */
2618 void
2619 row_callback(png_structp png_ptr, png_bytep new_row,
2620 png_uint_32 row_num, int pass)
2621 {
2622 /* If the image is interlaced, and you turned
2623 on the interlace handler, this function will
2624 be called for every row in every pass. Some
2625 of these rows will not be changed from the
2626 previous pass. When the row is not changed,
2627 the new_row variable will be NULL. The rows
2628 and passes are called in order, so you don't
2629 really need the row_num and pass, but I'm
2630 supplying them because it may make your life
2631 easier.
2633 If you did not turn on interlace handling then
2634 the callback is called for each row of each
2635 sub-image when the image is interlaced. In this
2636 case 'row_num' is the row in the sub-image, not
2637 the row in the output image as it is in all other
2638 cases.
2640 For the non-NULL rows of interlaced images when
2641 you have switched on libpng interlace handling,
2642 you must call png_progressive_combine_row()
2643 passing in the row and the old row. You can
2644 call this function for NULL rows (it will just
2645 return) and for non-interlaced images (it just
2646 does the memcpy for you) if it will make the
2647 code easier. Thus, you can just do this for
2648 all cases if you switch on interlace handling;
2649 */
2651 png_progressive_combine_row(png_ptr, old_row,
2652 new_row);
2654 /* where old_row is what was displayed
2655 previously for the row. Note that the first
2656 pass (pass == 0, really) will completely cover
2657 the old row, so the rows do not have to be
2658 initialized. After the first pass (and only
2659 for interlaced images), you will have to pass
2660 the current row, and the function will combine
2661 the old row and the new row.
2663 You can also call png_process_data_pause in this
2664 callback - see above.
2665 */
2666 }
2668 void
2669 end_callback(png_structp png_ptr, png_infop info)
2670 {
2671 /* This function is called after the whole image
2672 has been read, including any chunks after the
2673 image (up to and including the IEND). You
2674 will usually have the same info chunk as you
2675 had in the header, although some data may have
2676 been added to the comments and time fields.
2678 Most people won't do much here, perhaps setting
2679 a flag that marks the image as finished.
2680 */
2681 }
2685 IV. Writing
2687 Much of this is very similar to reading. However, everything of
2688 importance is repeated here, so you won't have to constantly look
2689 back up in the reading section to understand writing.
2691 Setup
2693 You will want to do the I/O initialization before you get into libpng,
2694 so if it doesn't work, you don't have anything to undo. If you are not
2695 using the standard I/O functions, you will need to replace them with
2696 custom writing functions. See the discussion under Customizing libpng.
2698 FILE *fp = fopen(file_name, "wb");
2700 if (!fp)
2701 return (ERROR);
2703 Next, png_struct and png_info need to be allocated and initialized.
2704 As these can be both relatively large, you may not want to store these
2705 on the stack, unless you have stack space to spare. Of course, you
2706 will want to check if they return NULL. If you are also reading,
2707 you won't want to name your read structure and your write structure
2708 both "png_ptr"; you can call them anything you like, such as
2709 "read_ptr" and "write_ptr". Look at pngtest.c, for example.
2711 png_structp png_ptr = png_create_write_struct
2712 (PNG_LIBPNG_VER_STRING, (png_voidp)user_error_ptr,
2713 user_error_fn, user_warning_fn);
2715 if (!png_ptr)
2716 return (ERROR);
2718 png_infop info_ptr = png_create_info_struct(png_ptr);
2719 if (!info_ptr)
2720 {
2721 png_destroy_write_struct(&png_ptr,
2722 (png_infopp)NULL);
2723 return (ERROR);
2724 }
2726 If you want to use your own memory allocation routines,
2727 define PNG_USER_MEM_SUPPORTED and use
2728 png_create_write_struct_2() instead of png_create_write_struct():
2730 png_structp png_ptr = png_create_write_struct_2
2731 (PNG_LIBPNG_VER_STRING, (png_voidp)user_error_ptr,
2732 user_error_fn, user_warning_fn, (png_voidp)
2733 user_mem_ptr, user_malloc_fn, user_free_fn);
2735 After you have these structures, you will need to set up the
2736 error handling. When libpng encounters an error, it expects to
2737 longjmp() back to your routine. Therefore, you will need to call
2738 setjmp() and pass the png_jmpbuf(png_ptr). If you
2739 write the file from different routines, you will need to update
2740 the png_jmpbuf(png_ptr) every time you enter a new routine that will
2741 call a png_*() function. See your documentation of setjmp/longjmp
2742 for your compiler for more information on setjmp/longjmp. See
2743 the discussion on libpng error handling in the Customizing Libpng
2744 section below for more information on the libpng error handling.
2746 if (setjmp(png_jmpbuf(png_ptr)))
2747 {
2748 png_destroy_write_struct(&png_ptr, &info_ptr);
2749 fclose(fp);
2750 return (ERROR);
2751 }
2752 ...
2753 return;
2755 If you would rather avoid the complexity of setjmp/longjmp issues,
2756 you can compile libpng with PNG_NO_SETJMP, in which case
2757 errors will result in a call to PNG_ABORT() which defaults to abort().
2759 You can #define PNG_ABORT() to a function that does something
2760 more useful than abort(), as long as your function does not
2761 return.
2763 Checking for invalid palette index on write was added at libpng
2764 1.5.10. If a pixel contains an invalid (out-of-range) index libpng issues
2765 a benign error. This is enabled by default because this condition is an
2766 error according to the PNG specification, Clause 11.3.2, but the error can
2767 be ignored in each png_ptr with
2769 png_set_check_for_invalid_index(png_ptr, 0);
2771 If the error is ignored, or if png_benign_error() treats it as a warning,
2772 any invalid pixels are written as-is by the encoder, resulting in an
2773 invalid PNG datastream as output. In this case the application is
2774 responsible for ensuring that the pixel indexes are in range when it writes
2775 a PLTE chunk with fewer entries than the bit depth would allow.
2777 Now you need to set up the output code. The default for libpng is to
2778 use the C function fwrite(). If you use this, you will need to pass a
2779 valid FILE * in the function png_init_io(). Be sure that the file is
2780 opened in binary mode. Again, if you wish to handle writing data in
2781 another way, see the discussion on libpng I/O handling in the Customizing
2782 Libpng section below.
2784 png_init_io(png_ptr, fp);
2786 If you are embedding your PNG into a datastream such as MNG, and don't
2787 want libpng to write the 8-byte signature, or if you have already
2788 written the signature in your application, use
2790 png_set_sig_bytes(png_ptr, 8);
2792 to inform libpng that it should not write a signature.
2794 Write callbacks
2796 At this point, you can set up a callback function that will be
2797 called after each row has been written, which you can use to control
2798 a progress meter or the like. It's demonstrated in pngtest.c.
2799 You must supply a function
2801 void write_row_callback(png_structp png_ptr, png_uint_32 row,
2802 int pass);
2803 {
2804 /* put your code here */
2805 }
2807 (You can give it another name that you like instead of "write_row_callback")
2809 To inform libpng about your function, use
2811 png_set_write_status_fn(png_ptr, write_row_callback);
2813 When this function is called the row has already been completely processed and
2814 it has also been written out. The 'row' and 'pass' refer to the next row to be
2815 handled. For the
2816 non-interlaced case the row that was just handled is simply one less than the
2817 passed in row number, and pass will always be 0. For the interlaced case the
2818 same applies unless the row value is 0, in which case the row just handled was
2819 the last one from one of the preceding passes. Because interlacing may skip a
2820 pass you cannot be sure that the preceding pass is just 'pass-1', if you really
2821 need to know what the last pass is record (row,pass) from the callback and use
2822 the last recorded value each time.
2824 As with the user transform you can find the output row using the
2827 You now have the option of modifying how the compression library will
2828 run. The following functions are mainly for testing, but may be useful
2829 in some cases, like if you need to write PNG files extremely fast and
2830 are willing to give up some compression, or if you want to get the
2831 maximum possible compression at the expense of slower writing. If you
2832 have no special needs in this area, let the library do what it wants by
2833 not calling this function at all, as it has been tuned to deliver a good
2834 speed/compression ratio. The second parameter to png_set_filter() is
2835 the filter method, for which the only valid values are 0 (as of the
2836 July 1999 PNG specification, version 1.2) or 64 (if you are writing
2837 a PNG datastream that is to be embedded in a MNG datastream). The third
2838 parameter is a flag that indicates which filter type(s) are to be tested
2839 for each scanline. See the PNG specification for details on the specific
2840 filter types.
2843 /* turn on or off filtering, and/or choose
2844 specific filters. You can use either a single
2845 PNG_FILTER_VALUE_NAME or the bitwise OR of one
2846 or more PNG_FILTER_NAME masks.
2847 */
2848 png_set_filter(png_ptr, 0,
2856 If an application wants to start and stop using particular filters during
2857 compression, it should start out with all of the filters (to ensure that
2858 the previous row of pixels will be stored in case it's needed later),
2859 and then add and remove them after the start of compression.
2861 If you are writing a PNG datastream that is to be embedded in a MNG
2862 datastream, the second parameter can be either 0 or 64.
2864 The png_set_compression_*() functions interface to the zlib compression
2865 library, and should mostly be ignored unless you really know what you are
2866 doing. The only generally useful call is png_set_compression_level()
2867 which changes how much time zlib spends on trying to compress the image
2868 data. See the Compression Library (zlib.h and algorithm.txt, distributed
2869 with zlib) for details on the compression levels.
2871 #include zlib.h
2873 /* Set the zlib compression level */
2874 png_set_compression_level(png_ptr,
2877 /* Set other zlib parameters for compressing IDAT */
2878 png_set_compression_mem_level(png_ptr, 8);
2879 png_set_compression_strategy(png_ptr,
2881 png_set_compression_window_bits(png_ptr, 15);
2882 png_set_compression_method(png_ptr, 8);
2883 png_set_compression_buffer_size(png_ptr, 8192)
2885 /* Set zlib parameters for text compression
2886 * If you don't call these, the parameters
2887 * fall back on those defined for IDAT chunks
2888 */
2889 png_set_text_compression_mem_level(png_ptr, 8);
2890 png_set_text_compression_strategy(png_ptr,
2892 png_set_text_compression_window_bits(png_ptr, 15);
2893 png_set_text_compression_method(png_ptr, 8);
2895 Setting the contents of info for output
2897 You now need to fill in the png_info structure with all the data you
2898 wish to write before the actual image. Note that the only thing you
2899 are allowed to write after the image is the text chunks and the time
2900 chunk (as of PNG Specification 1.2, anyway). See png_write_end() and
2901 the latest PNG specification for more information on that. If you
2902 wish to write them before the image, fill them in now, and flag that
2903 data as being valid. If you want to wait until after the data, don't
2904 fill them until png_write_end(). For all the fields in png_info and
2905 their data types, see png.h. For explanations of what the fields
2906 contain, see the PNG specification.
2908 Some of the more important parts of the png_info are:
2910 png_set_IHDR(png_ptr, info_ptr, width, height,
2911 bit_depth, color_type, interlace_type,
2912 compression_type, filter_method)
2914 width - holds the width of the image
2915 in pixels (up to 2^31).
2917 height - holds the height of the image
2918 in pixels (up to 2^31).
2920 bit_depth - holds the bit depth of one of the
2921 image channels.
2922 (valid values are 1, 2, 4, 8, 16
2923 and depend also on the
2924 color_type. See also significant
2925 bits (sBIT) below).
2927 color_type - describes which color/alpha
2928 channels are present.
2930 (bit depths 1, 2, 4, 8, 16)
2932 (bit depths 8, 16)
2934 (bit depths 1, 2, 4, 8)
2936 (bit_depths 8, 16)
2938 (bit_depths 8, 16)
2944 interlace_type - PNG_INTERLACE_NONE or
2947 compression_type - (must be
2950 filter_method - (must be PNG_FILTER_TYPE_DEFAULT
2951 or, if you are writing a PNG to
2952 be embedded in a MNG datastream,
2953 can also be
2956 If you call png_set_IHDR(), the call must appear before any of the
2957 other png_set_*() functions, because they might require access to some of
2958 the IHDR settings. The remaining png_set_*() functions can be called
2959 in any order.
2961 If you wish, you can reset the compression_type, interlace_type, or
2962 filter_method later by calling png_set_IHDR() again; if you do this, the
2963 width, height, bit_depth, and color_type must be the same in each call.
2965 png_set_PLTE(png_ptr, info_ptr, palette,
2966 num_palette);
2968 palette - the palette for the file
2969 (array of png_color)
2970 num_palette - number of entries in the palette
2973 png_set_gAMA(png_ptr, info_ptr, file_gamma);
2974 png_set_gAMA_fixed(png_ptr, info_ptr, int_file_gamma);
2976 file_gamma - the gamma at which the image was
2977 created (PNG_INFO_gAMA)
2979 int_file_gamma - 100,000 times the gamma at which
2980 the image was created
2982 png_set_cHRM(png_ptr, info_ptr, white_x, white_y, red_x, red_y,
2983 green_x, green_y, blue_x, blue_y)
2984 png_set_cHRM_XYZ(png_ptr, info_ptr, red_X, red_Y, red_Z, green_X,
2985 green_Y, green_Z, blue_X, blue_Y, blue_Z)
2986 png_set_cHRM_fixed(png_ptr, info_ptr, int_white_x, int_white_y,
2987 int_red_x, int_red_y, int_green_x, int_green_y,
2988 int_blue_x, int_blue_y)
2989 png_set_cHRM_XYZ_fixed(png_ptr, info_ptr, int_red_X, int_red_Y,
2990 int_red_Z, int_green_X, int_green_Y, int_green_Z,
2991 int_blue_X, int_blue_Y, int_blue_Z)
2993 {white,red,green,blue}_{x,y}
2994 A color space encoding specified using the chromaticities
2995 of the end points and the white point.
2997 {red,green,blue}_{X,Y,Z}
2998 A color space encoding specified using the encoding end
2999 points - the CIE tristimulus specification of the intended
3000 color of the red, green and blue channels in the PNG RGB
3001 data. The white point is simply the sum of the three end
3002 points.
3004 png_set_sRGB(png_ptr, info_ptr, srgb_intent);
3006 srgb_intent - the rendering intent
3007 (PNG_INFO_sRGB) The presence of
3008 the sRGB chunk means that the pixel
3009 data is in the sRGB color space.
3010 This chunk also implies specific
3011 values of gAMA and cHRM. Rendering
3012 intent is the CSS-1 property that
3013 has been defined by the International
3014 Color Consortium
3015 (http://www.color.org).
3016 It can be one of
3023 png_set_sRGB_gAMA_and_cHRM(png_ptr, info_ptr,
3024 srgb_intent);
3026 srgb_intent - the rendering intent
3027 (PNG_INFO_sRGB) The presence of the
3028 sRGB chunk means that the pixel
3029 data is in the sRGB color space.
3030 This function also causes gAMA and
3031 cHRM chunks with the specific values
3032 that are consistent with sRGB to be
3033 written.
3035 png_set_iCCP(png_ptr, info_ptr, name, compression_type,
3036 profile, proflen);
3038 name - The profile name.
3040 compression_type - The compression type; always
3042 You may give NULL to this argument to
3043 ignore it.
3045 profile - International Color Consortium color
3046 profile data. May contain NULs.
3048 proflen - length of profile data in bytes.
3050 png_set_sBIT(png_ptr, info_ptr, sig_bit);
3052 sig_bit - the number of significant bits for
3053 (PNG_INFO_sBIT) each of the gray, red,
3054 green, and blue channels, whichever are
3055 appropriate for the given color type
3056 (png_color_16)
3058 png_set_tRNS(png_ptr, info_ptr, trans_alpha,
3059 num_trans, trans_color);
3061 trans_alpha - array of alpha (transparency)
3062 entries for palette (PNG_INFO_tRNS)
3064 num_trans - number of transparent entries
3065 (PNG_INFO_tRNS)
3067 trans_color - graylevel or color sample values
3068 (in order red, green, blue) of the
3069 single transparent color for
3070 non-paletted images (PNG_INFO_tRNS)
3072 png_set_hIST(png_ptr, info_ptr, hist);
3074 hist - histogram of palette (array of
3075 png_uint_16) (PNG_INFO_hIST)
3077 png_set_tIME(png_ptr, info_ptr, mod_time);
3079 mod_time - time image was last modified
3082 png_set_bKGD(png_ptr, info_ptr, background);
3084 background - background color (of type
3085 png_color_16p) (PNG_VALID_bKGD)
3087 png_set_text(png_ptr, info_ptr, text_ptr, num_text);
3089 text_ptr - array of png_text holding image
3092 text_ptr[i].compression - type of compression used
3097 text_ptr[i].key - keyword for comment. Must contain
3098 1-79 characters.
3099 text_ptr[i].text - text comments for current
3100 keyword. Can be NULL or empty.
3101 text_ptr[i].text_length - length of text string,
3102 after decompression, 0 for iTXt
3103 text_ptr[i].itxt_length - length of itxt string,
3104 after decompression, 0 for tEXt/zTXt
3105 text_ptr[i].lang - language of comment (NULL or
3106 empty for unknown).
3107 text_ptr[i].translated_keyword - keyword in UTF-8 (NULL
3108 or empty for unknown).
3110 Note that the itxt_length, lang, and lang_key
3111 members of the text_ptr structure only exist when the
3112 library is built with iTXt chunk support. Prior to
3113 libpng-1.4.0 the library was built by default without
3114 iTXt support. Also note that when iTXt is supported,
3115 they contain NULL pointers when the "compression"
3116 field contains PNG_TEXT_COMPRESSION_NONE or
3119 num_text - number of comments
3121 png_set_sPLT(png_ptr, info_ptr, &palette_ptr,
3122 num_spalettes);
3124 palette_ptr - array of png_sPLT_struct structures
3125 to be added to the list of palettes
3126 in the info structure.
3127 num_spalettes - number of palette structures to be
3128 added.
3130 png_set_oFFs(png_ptr, info_ptr, offset_x, offset_y,
3131 unit_type);
3133 offset_x - positive offset from the left
3134 edge of the screen
3136 offset_y - positive offset from the top
3137 edge of the screen
3141 png_set_pHYs(png_ptr, info_ptr, res_x, res_y,
3142 unit_type);
3144 res_x - pixels/unit physical resolution
3145 in x direction
3147 res_y - pixels/unit physical resolution
3148 in y direction
3150 unit_type - PNG_RESOLUTION_UNKNOWN,
3153 png_set_sCAL(png_ptr, info_ptr, unit, width, height)
3155 unit - physical scale units (an integer)
3157 width - width of a pixel in physical scale units
3159 height - height of a pixel in physical scale units
3160 (width and height are doubles)
3162 png_set_sCAL_s(png_ptr, info_ptr, unit, width, height)
3164 unit - physical scale units (an integer)
3166 width - width of a pixel in physical scale units
3167 expressed as a string
3169 height - height of a pixel in physical scale units
3170 (width and height are strings like "2.54")
3172 png_set_unknown_chunks(png_ptr, info_ptr, &unknowns,
3173 num_unknowns)
3175 unknowns - array of png_unknown_chunk
3176 structures holding unknown chunks
3177 unknowns[i].name - name of unknown chunk
3178 unknowns[i].data - data of unknown chunk
3179 unknowns[i].size - size of unknown chunk's data
3180 unknowns[i].location - position to write chunk in file
3181 0: do not write chunk
3182 PNG_HAVE_IHDR: before PLTE
3183 PNG_HAVE_PLTE: before IDAT
3186 The "location" member is set automatically according to
3187 what part of the output file has already been written.
3188 You can change its value after calling png_set_unknown_chunks()
3189 as demonstrated in pngtest.c. Within each of the "locations",
3190 the chunks are sequenced according to their position in the
3191 structure (that is, the value of "i", which is the order in which
3192 the chunk was either read from the input file or defined with
3193 png_set_unknown_chunks).
3195 A quick word about text and num_text. text is an array of png_text
3196 structures. num_text is the number of valid structures in the array.
3197 Each png_text structure holds a language code, a keyword, a text value,
3198 and a compression type.
3200 The compression types have the same valid numbers as the compression
3201 types of the image data. Currently, the only valid number is zero.
3202 However, you can store text either compressed or uncompressed, unlike
3203 images, which always have to be compressed. So if you don't want the
3204 text compressed, set the compression type to PNG_TEXT_COMPRESSION_NONE.
3205 Because tEXt and zTXt chunks don't have a language field, if you
3207 any language code or translated keyword will not be written out.
3209 Until text gets around a few hundred bytes, it is not worth compressing it.
3210 After the text has been written out to the file, the compression type
3212 so that it isn't written out again at the end (in case you are calling
3213 png_write_end() with the same struct).
3215 The keywords that are given in the PNG Specification are:
3217 Title Short (one line) title or
3218 caption for image
3220 Author Name of image's creator
3222 Description Description of image (possibly long)
3224 Copyright Copyright notice
3226 Creation Time Time of original image creation
3227 (usually RFC 1123 format, see below)
3229 Software Software used to create the image
3231 Disclaimer Legal disclaimer
3233 Warning Warning of nature of content
3235 Source Device used to create the image
3237 Comment Miscellaneous comment; conversion
3238 from other image format
3240 The keyword-text pairs work like this. Keywords should be short
3241 simple descriptions of what the comment is about. Some typical
3242 keywords are found in the PNG specification, as is some recommendations
3243 on keywords. You can repeat keywords in a file. You can even write
3244 some text before the image and some after. For example, you may want
3245 to put a description of the image before the image, but leave the
3246 disclaimer until after, so viewers working over modem connections
3247 don't have to wait for the disclaimer to go over the modem before
3248 they start seeing the image. Finally, keywords should be full
3249 words, not abbreviations. Keywords and text are in the ISO 8859-1
3250 (Latin-1) character set (a superset of regular ASCII) and can not
3251 contain NUL characters, and should not contain control or other
3252 unprintable characters. To make the comments widely readable, stick
3253 with basic ASCII, and avoid machine specific character set extensions
3254 like the IBM-PC character set. The keyword must be present, but
3255 you can leave off the text string on non-compressed pairs.
3256 Compressed pairs must have a text string, as only the text string
3257 is compressed anyway, so the compression would be meaningless.
3259 PNG supports modification time via the png_time structure. Two
3260 conversion routines are provided, png_convert_from_time_t() for
3261 time_t and png_convert_from_struct_tm() for struct tm. The
3262 time_t routine uses gmtime(). You don't have to use either of
3263 these, but if you wish to fill in the png_time structure directly,
3264 you should provide the time in universal time (GMT) if possible
3265 instead of your local time. Note that the year number is the full
3266 year (e.g. 1998, rather than 98 - PNG is year 2000 compliant!), and
3267 that months start with 1.
3269 If you want to store the time of the original image creation, you should
3270 use a plain tEXt chunk with the "Creation Time" keyword. This is
3271 necessary because the "creation time" of a PNG image is somewhat vague,
3272 depending on whether you mean the PNG file, the time the image was
3273 created in a non-PNG format, a still photo from which the image was
3274 scanned, or possibly the subject matter itself. In order to facilitate
3275 machine-readable dates, it is recommended that the "Creation Time"
3276 tEXt chunk use RFC 1123 format dates (e.g. "22 May 1997 18:07:10 GMT"),
3277 although this isn't a requirement. Unlike the tIME chunk, the
3278 "Creation Time" tEXt chunk is not expected to be automatically changed
3279 by the software. To facilitate the use of RFC 1123 dates, a function
3280 png_convert_to_rfc1123_buffer(buffer, png_timep) is provided to
3281 convert from PNG time to an RFC 1123 format string. The caller must provide
3282 a writeable buffer of at least 29 bytes.
3284 Writing unknown chunks
3286 You can use the png_set_unknown_chunks function to queue up private chunks
3287 for writing. You give it a chunk name, location, raw data, and a size. You
3288 also must use png_set_keep_unknown_chunks() to ensure that libpng will
3289 handle them. That's all there is to it. The chunks will be written by the
3290 next following png_write_info_before_PLTE, png_write_info, or png_write_end
3291 function, depending upon the specified location. Any chunks previously
3292 read into the info structure's unknown-chunk list will also be written out
3293 in a sequence that satisfies the PNG specification's ordering rules.
3295 Here is an example of writing two private chunks, prVt and miNE:
3298 /* Set unknown chunk data */
3299 png_unknown_chunk unk_chunk[2];
3300 strcpy((char *) unk_chunk[0].name, "prVt";
3301 unk_chunk[0].data = (unsigned char *) "PRIVATE DATA";
3302 unk_chunk[0].size = strlen(unk_chunk[0].data)+1;
3303 unk_chunk[0].location = PNG_HAVE_IHDR;
3304 strcpy((char *) unk_chunk[1].name, "miNE";
3305 unk_chunk[1].data = (unsigned char *) "MY CHUNK DATA";
3306 unk_chunk[1].size = strlen(unk_chunk[0].data)+1;
3307 unk_chunk[1].location = PNG_AFTER_IDAT;
3308 png_set_unknown_chunks(write_ptr, write_info_ptr,
3309 unk_chunk, 2);
3310 /* Needed because miNE is not safe-to-copy */
3311 png_set_keep_unknown_chunks(png, PNG_HANDLE_CHUNK_ALWAYS,
3312 (png_bytep) "miNE", 1);
3313 # if PNG_LIBPNG_VER < 10600
3314 /* Deal with unknown chunk location bug in 1.5.x and earlier */
3315 png_set_unknown_chunk_location(png, info, 0, PNG_HAVE_IHDR);
3316 png_set_unknown_chunk_location(png, info, 1, PNG_AFTER_IDAT);
3317 # endif
3318 # if PNG_LIBPNG_VER < 10500
3319 /* PNG_AFTER_IDAT writes two copies of the chunk prior to libpng-1.5.0,
3320 * one before IDAT and another after IDAT, so don't use it; only use
3321 * PNG_HAVE_IHDR location. This call resets the location previously
3322 * set by assignment and png_set_unknown_chunk_location() for chunk 1.
3323 */
3324 png_set_unknown_chunk_location(png, info, 1, PNG_HAVE_IHDR);
3325 # endif
3326 #endif
3328 The high-level write interface
3330 At this point there are two ways to proceed; through the high-level
3331 write interface, or through a sequence of low-level write operations.
3332 You can use the high-level interface if your image data is present
3333 in the info structure. All defined output
3334 transformations are permitted, enabled by the following masks.
3336 PNG_TRANSFORM_IDENTITY No transformation
3337 PNG_TRANSFORM_PACKING Pack 1, 2 and 4-bit samples
3338 PNG_TRANSFORM_PACKSWAP Change order of packed
3339 pixels to LSB first
3340 PNG_TRANSFORM_INVERT_MONO Invert monochrome images
3341 PNG_TRANSFORM_SHIFT Normalize pixels to the
3342 sBIT depth
3344 to BGRA
3346 to AG
3347 PNG_TRANSFORM_INVERT_ALPHA Change alpha from opacity
3348 to transparency
3349 PNG_TRANSFORM_SWAP_ENDIAN Byte-swap 16-bit samples
3350 PNG_TRANSFORM_STRIP_FILLER Strip out filler
3351 bytes (deprecated).
3353 filler bytes
3355 filler bytes
3357 If you have valid image data in the info structure (you can use
3358 png_set_rows() to put image data in the info structure), simply do this:
3360 png_write_png(png_ptr, info_ptr, png_transforms, NULL)
3362 where png_transforms is an integer containing the bitwise OR of some set of
3363 transformation flags. This call is equivalent to png_write_info(),
3364 followed the set of transformations indicated by the transform mask,
3365 then png_write_image(), and finally png_write_end().
3367 (The final parameter of this call is not yet used. Someday it might point
3368 to transformation parameters required by some future output transform.)
3370 You must use png_transforms and not call any png_set_transform() functions
3371 when you use png_write_png().
3373 The low-level write interface
3375 If you are going the low-level route instead, you are now ready to
3376 write all the file information up to the actual image data. You do
3377 this with a call to png_write_info().
3379 png_write_info(png_ptr, info_ptr);
3381 Note that there is one transformation you may need to do before
3382 png_write_info(). In PNG files, the alpha channel in an image is the
3383 level of opacity. If your data is supplied as a level of transparency,
3384 you can invert the alpha channel before you write it, so that 0 is
3385 fully transparent and 255 (in 8-bit or paletted images) or 65535
3386 (in 16-bit images) is fully opaque, with
3388 png_set_invert_alpha(png_ptr);
3390 This must appear before png_write_info() instead of later with the
3391 other transformations because in the case of paletted images the tRNS
3392 chunk data has to be inverted before the tRNS chunk is written. If
3393 your image is not a paletted image, the tRNS data (which in such cases
3394 represents a single color to be rendered as transparent) won't need to
3395 be changed, and you can safely do this transformation after your
3396 png_write_info() call.
3398 If you need to write a private chunk that you want to appear before
3399 the PLTE chunk when PLTE is present, you can write the PNG info in
3400 two steps, and insert code to write your own chunk between them:
3402 png_write_info_before_PLTE(png_ptr, info_ptr);
3403 png_set_unknown_chunks(png_ptr, info_ptr, ...);
3404 png_write_info(png_ptr, info_ptr);
3406 After you've written the file information, you can set up the library
3407 to handle any special transformations of the image data. The various
3408 ways to transform the data will be described in the order that they
3409 should occur. This is important, as some of these change the color
3410 type and/or bit depth of the data, and some others only work on
3411 certain color types and bit depths. Even though each transformation
3412 checks to see if it has data that it can do something with, you should
3413 make sure to only enable a transformation if it will be valid for the
3414 data. For example, don't swap red and blue on grayscale data.
3416 PNG files store RGB pixels packed into 3 or 6 bytes. This code tells
3417 the library to strip input data that has 4 or 8 bytes per pixel down
3418 to 3 or 6 bytes (or strip 2 or 4-byte grayscale+filler data to 1 or 2
3419 bytes per pixel).
3421 png_set_filler(png_ptr, 0, PNG_FILLER_BEFORE);
3423 where the 0 is unused, and the location is either PNG_FILLER_BEFORE or
3424 PNG_FILLER_AFTER, depending upon whether the filler byte in the pixel
3425 is stored XRGB or RGBX.
3427 PNG files pack pixels of bit depths 1, 2, and 4 into bytes as small as
3428 they can, resulting in, for example, 8 pixels per byte for 1 bit files.
3429 If the data is supplied at 1 pixel per byte, use this code, which will
3430 correctly pack the pixels into a single byte:
3432 png_set_packing(png_ptr);
3434 PNG files reduce possible bit depths to 1, 2, 4, 8, and 16. If your
3435 data is of another bit depth, you can write an sBIT chunk into the
3436 file so that decoders can recover the original data if desired.
3438 /* Set the true bit depth of the image data */
3439 if (color_type & PNG_COLOR_MASK_COLOR)
3440 {
3441 sig_bit.red = true_bit_depth;
3442 sig_bit.green = true_bit_depth;
3443 sig_bit.blue = true_bit_depth;
3444 }
3446 else
3447 {
3448 sig_bit.gray = true_bit_depth;
3449 }
3451 if (color_type & PNG_COLOR_MASK_ALPHA)
3452 {
3453 sig_bit.alpha = true_bit_depth;
3454 }
3456 png_set_sBIT(png_ptr, info_ptr, &sig_bit);
3458 If the data is stored in the row buffer in a bit depth other than
3459 one supported by PNG (e.g. 3 bit data in the range 0-7 for a 4-bit PNG),
3460 this will scale the values to appear to be the correct bit depth as
3461 is required by PNG.
3463 png_set_shift(png_ptr, &sig_bit);
3465 PNG files store 16-bit pixels in network byte order (big-endian,
3466 ie. most significant bits first). This code would be used if they are
3467 supplied the other way (little-endian, i.e. least significant bits
3468 first, the way PCs store them):
3470 if (bit_depth > 8)
3471 png_set_swap(png_ptr);
3473 If you are using packed-pixel images (1, 2, or 4 bits/pixel), and you
3474 need to change the order the pixels are packed into bytes, you can use:
3476 if (bit_depth < 8)
3477 png_set_packswap(png_ptr);
3479 PNG files store 3 color pixels in red, green, blue order. This code
3480 would be used if they are supplied as blue, green, red:
3482 png_set_bgr(png_ptr);
3484 PNG files describe monochrome as black being zero and white being
3485 one. This code would be used if the pixels are supplied with this reversed
3486 (black being one and white being zero):
3488 png_set_invert_mono(png_ptr);
3490 Finally, you can write your own transformation function if none of
3491 the existing ones meets your needs. This is done by setting a callback
3492 with
3494 png_set_write_user_transform_fn(png_ptr,
3495 write_transform_fn);
3497 You must supply the function
3499 void write_transform_fn(png_structp png_ptr, png_row_infop
3500 row_info, png_bytep data)
3502 See pngtest.c for a working example. Your function will be called
3503 before any of the other transformations are processed. If supported
3504 libpng also supplies an information routine that may be called from
3505 your callback:
3507 png_get_current_row_number(png_ptr);
3508 png_get_current_pass_number(png_ptr);
3510 This returns the current row passed to the transform. With interlaced
3511 images the value returned is the row in the input sub-image image. Use
3512 PNG_ROW_FROM_PASS_ROW(row, pass) and PNG_COL_FROM_PASS_COL(col, pass) to
3513 find the output pixel (x,y) given an interlaced sub-image pixel (row,col,pass).
3515 The discussion of interlace handling above contains more information on how to
3516 use these values.
3518 You can also set up a pointer to a user structure for use by your
3519 callback function.
3521 png_set_user_transform_info(png_ptr, user_ptr, 0, 0);
3523 The user_channels and user_depth parameters of this function are ignored
3524 when writing; you can set them to zero as shown.
3526 You can retrieve the pointer via the function png_get_user_transform_ptr().
3527 For example:
3529 voidp write_user_transform_ptr =
3530 png_get_user_transform_ptr(png_ptr);
3532 It is possible to have libpng flush any pending output, either manually,
3533 or automatically after a certain number of lines have been written. To
3534 flush the output stream a single time call:
3536 png_write_flush(png_ptr);
3538 and to have libpng flush the output stream periodically after a certain
3539 number of scanlines have been written, call:
3541 png_set_flush(png_ptr, nrows);
3543 Note that the distance between rows is from the last time png_write_flush()
3544 was called, or the first row of the image if it has never been called.
3545 So if you write 50 lines, and then png_set_flush 25, it will flush the
3546 output on the next scanline, and every 25 lines thereafter, unless
3547 png_write_flush() is called before 25 more lines have been written.
3548 If nrows is too small (less than about 10 lines for a 640 pixel wide
3549 RGB image) the image compression may decrease noticeably (although this
3550 may be acceptable for real-time applications). Infrequent flushing will
3551 only degrade the compression performance by a few percent over images
3552 that do not use flushing.
3554 Writing the image data
3556 That's it for the transformations. Now you can write the image data.
3557 The simplest way to do this is in one function call. If you have the
3558 whole image in memory, you can just call png_write_image() and libpng
3559 will write the image. You will need to pass in an array of pointers to
3560 each row. This function automatically handles interlacing, so you don't
3561 need to call png_set_interlace_handling() or call this function multiple
3562 times, or any of that other stuff necessary with png_write_rows().
3564 png_write_image(png_ptr, row_pointers);
3566 where row_pointers is:
3568 png_byte *row_pointers[height];
3570 You can point to void or char or whatever you use for pixels.
3572 If you don't want to write the whole image at once, you can
3573 use png_write_rows() instead. If the file is not interlaced,
3574 this is simple:
3576 png_write_rows(png_ptr, row_pointers,
3577 number_of_rows);
3579 row_pointers is the same as in the png_write_image() call.
3581 If you are just writing one row at a time, you can do this with
3582 a single row_pointer instead of an array of row_pointers:
3584 png_bytep row_pointer = row;
3586 png_write_row(png_ptr, row_pointer);
3588 When the file is interlaced, things can get a good deal more complicated.
3589 The only currently (as of the PNG Specification version 1.2, dated July
3590 1999) defined interlacing scheme for PNG files is the "Adam7" interlace
3591 scheme, that breaks down an image into seven smaller images of varying
3592 size. libpng will build these images for you, or you can do them
3593 yourself. If you want to build them yourself, see the PNG specification
3594 for details of which pixels to write when.
3596 If you don't want libpng to handle the interlacing details, just
3597 use png_set_interlace_handling() and call png_write_rows() the
3598 correct number of times to write all the sub-images
3599 (png_set_interlace_handling() returns the number of sub-images.)
3601 If you want libpng to build the sub-images, call this before you start
3602 writing any rows:
3604 number_of_passes = png_set_interlace_handling(png_ptr);
3606 This will return the number of passes needed. Currently, this is seven,
3607 but may change if another interlace type is added.
3609 Then write the complete image number_of_passes times.
3611 png_write_rows(png_ptr, row_pointers, number_of_rows);
3613 Think carefully before you write an interlaced image. Typically code that
3614 reads such images reads all the image data into memory, uncompressed, before
3615 doing any processing. Only code that can display an image on the fly can
3616 take advantage of the interlacing and even then the image has to be exactly
3617 the correct size for the output device, because scaling an image requires
3618 adjacent pixels and these are not available until all the passes have been
3619 read.
3621 If you do write an interlaced image you will hardly ever need to handle
3622 the interlacing yourself. Call png_set_interlace_handling() and use the
3623 approach described above.
3625 The only time it is conceivable that you will really need to write an
3626 interlaced image pass-by-pass is when you have read one pass by pass and
3627 made some pixel-by-pixel transformation to it, as described in the read
3628 code above. In this case use the PNG_PASS_ROWS and PNG_PASS_COLS macros
3629 to determine the size of each sub-image in turn and simply write the rows
3630 you obtained from the read code.
3632 Finishing a sequential write
3634 After you are finished writing the image, you should finish writing
3635 the file. If you are interested in writing comments or time, you should
3636 pass an appropriately filled png_info pointer. If you are not interested,
3637 you can pass NULL.
3639 png_write_end(png_ptr, info_ptr);
3641 When you are done, you can free all memory used by libpng like this:
3643 png_destroy_write_struct(&png_ptr, &info_ptr);
3645 It is also possible to individually free the info_ptr members that
3646 point to libpng-allocated storage with the following function:
3648 png_free_data(png_ptr, info_ptr, mask, seq)
3650 mask - identifies data to be freed, a mask
3651 containing the bitwise OR of one or
3652 more of
3658 or simply PNG_FREE_ALL
3660 seq - sequence number of item to be freed
3661 (-1 for all items)
3663 This function may be safely called when the relevant storage has
3664 already been freed, or has not yet been allocated, or was allocated
3665 by the user and not by libpng, and will in those cases do nothing.
3666 The "seq" parameter is ignored if only one item of the selected data
3667 type, such as PLTE, is allowed. If "seq" is not -1, and multiple items
3668 are allowed for the data type identified in the mask, such as text or
3669 sPLT, only the n'th item in the structure is freed, where n is "seq".
3671 If you allocated data such as a palette that you passed in to libpng
3672 with png_set_*, you must not free it until just before the call to
3673 png_destroy_write_struct().
3675 The default behavior is only to free data that was allocated internally
3676 by libpng. This can be changed, so that libpng will not free the data,
3677 or so that it will free data that was allocated by the user with png_malloc()
3678 or png_calloc() and passed in via a png_set_*() function, with
3680 png_data_freer(png_ptr, info_ptr, freer, mask)
3682 freer - one of
3687 mask - which data elements are affected
3688 same choices as in png_free_data()
3690 For example, to transfer responsibility for some data from a read structure
3691 to a write structure, you could use
3693 png_data_freer(read_ptr, read_info_ptr,
3697 png_data_freer(write_ptr, write_info_ptr,
3701 thereby briefly reassigning responsibility for freeing to the user but
3702 immediately afterwards reassigning it once more to the write_destroy
3703 function. Having done this, it would then be safe to destroy the read
3704 structure and continue to use the PLTE, tRNS, and hIST data in the write
3705 structure.
3707 This function only affects data that has already been allocated.
3708 You can call this function before calling after the png_set_*() functions
3709 to control whether the user or png_destroy_*() is supposed to free the data.
3710 When the user assumes responsibility for libpng-allocated data, the
3711 application must use
3712 png_free() to free it, and when the user transfers responsibility to libpng
3713 for data that the user has allocated, the user must have used png_malloc()
3714 or png_calloc() to allocate it.
3716 If you allocated text_ptr.text, text_ptr.lang, and text_ptr.translated_keyword
3717 separately, do not transfer responsibility for freeing text_ptr to libpng,
3718 because when libpng fills a png_text structure it combines these members with
3719 the key member, and png_free_data() will free only text_ptr.key. Similarly,
3720 if you transfer responsibility for free'ing text_ptr from libpng to your
3721 application, your application must not separately free those members.
3722 For a more compact example of writing a PNG image, see the file example.c.
3724 V. Simplified API
3726 The simplified API, which became available in libpng-1.6.0, hides the details
3727 of both libpng and the PNG file format itself.
3728 It allows PNG files to be read into a very limited number of
3729 in-memory bitmap formats or to be written from the same formats. If these
3730 formats do not accommodate your needs then you can, and should, use the more
3731 sophisticated APIs above - these support a wide variety of in-memory formats
3732 and a wide variety of sophisticated transformations to those formats as well
3733 as a wide variety of APIs to manipulate ancilliary information.
3735 To read a PNG file using the simplified API:
3737 1) Declare a 'png_image' structure (see below) on the stack, set the
3738 version field to PNG_IMAGE_VERSION and the 'opaque' pointer to NULL
3739 (this is REQUIRED, your program may crash if you don't do it.)
3741 2) Call the appropriate png_image_begin_read... function.