Increased libtool TH*_LIB_CURRENT and thus minor version of SONAME as
API of a few methods changes by adding 'const' to some of the char *
arguments in codec.h.

A 1.2.0 release is planned in two weeks.

Related to #2319.

CI currently does not thoroughly test encoding and decoding (except for
encoding zeroed dummy frames).
Testing --disable-asm is still useful because it ensures that the code compiles
on platforms where there are no ASM optimizations and the generic C code paths
are used.

Adjust clang CI run to report warning about reserved identifiers.

A crash was discovered using input fuzzying, in th_decode_ceaderin()
where the len value in the oc_fuff_tree_unpack() can end up as -1.
Added a check to ensure this do not happen.

Based on feedback from Timothy B. Terriberry.

The issue was discovered using gcc sanitazion, which reported the following:

huffdec.c:228:27: runtime error: shift exponent -1 is negative
    #0 0x5d471012bfd0 in oc_huff_tree_unpack /home/uos/libtheora-18570/theora/lib/huffdec.c:228
    #1 0x5d471012c134 in oc_huff_trees_unpack /home/uos/libtheora-18570/theora/lib/huffdec.c:392
    #2 0x5d471010a98c in oc_setup_unpack /home/uos/libtheora-18570/theora/lib/decinfo.c:169
    #3 0x5d471010a98c in oc_dec_headerin /home/uos/libtheora-18570/theora/lib/decinfo.c:238
    #4 0x5d471010a98c in th_decode_headerin /home/uos/libtheora-18570/theora/lib/decinfo.c:266
    #5 0x5d47100fd638 in TheoraDecoder::initialize() /home/uos/libtheora-18570/libtheora-18570/fuzzer.cpp:66
    #6 0x5d47100ffa76 in TheoraDecoder::Run() /home/uos/libtheora-18570/libtheora-18570/fuzzer.cpp:180
    #7 0x5d47100ffe48 in main /home/uos/libtheora-18570/libtheora-18570/fuzzer.cpp:240
    #8 0x7cc9a5e29d8f in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
    #9 0x7cc9a5e29e3f in __libc_start_main_impl ../csu/libc-start.c:392
    #10 0x5d47100f9964 in _start (/home/uos/libtheora-18570/libtheora-18570/poc1+0x83964)

Fixes github pull request #19.

Do not build examples, as it require libvorbis, and do not build
documentation as it require a host of programs not available
during cross building.

During video playout using player_example with gcc sanitaztions
enabled, new bit shifting of negative values were discovered.
Replaced this with multiplication too.

Leftover changes from a6766c94.

Use multiplication instead, allowing the compiler to optimize to
bitshifts if it believe it to be safe.

Partly solves github issue #18.

These have not been updated since before the code was migrated to
git, and both no longer make sense and are not correct.

The 'int x=x' style statement do not make sense and cause problems with
some compilers.  Replace it with more sensible initializers to ensure
the value have a well known starting point.  Using initializers
proposed by Timothy B. Terriberry in !25.

Added -Werror=uninitialized -Winit-self to default GCC build to ensure
this style of statement do not sneak into the code again.

Fixes #2317

The security issue CVE-2024-56431 was discovered using GCC address
sanitaztion.  Make it easier to run this check, and run it in
gitlab CI.

Based on a comment in !28.

Ran 'cp Doxygen.in Doxygen; doxygen -u; cp Doxygen Doxygen.in' and
removed DOT_FONTNAME reference to non-free font.

Check if argument is null before using this approach, and just return
zero when the value to do bit count on is zero.

Fixes a issue discovered by --enable-gcc-sanitizers in !31:

  ../../lib/encode.c:1384:49: runtime error: passing zero to clz(),
    which is not a valid argument
    #0 0x7f0af5dae97d in th_encode_ctl ../../lib/encode.c:1384
    #1 0x7f0af5e2eb97 in theora_encode_init ../../lib/encapiwrapper.c:72
    #2 0x55fd4beb18b7 in granulepos_test_encode
     ../../tests/granulepos_theora.c:71
    #3 0x55fd4beb126d in main ../../tests/granulepos_theora.c:137
    #4 0x7f0af51d3249 in __libc_start_call_main
     ../sysdeps/nptl/libc_start_call_main.h:58
    #5 0x7f0af51d3304 in __libc_start_main_impl ../csu/libc-start.c:360
    #6 0x55fd4beb1300 in _start
     (/home/user/libtheora/build/tests/.libs/granulepos_theoraenc+0x2300)

Fixes #2323

The last iteration of the loop execute 1<<63, which would push the
result into the signed bit of a signed 64 bit type, and this
move into currently undefined behaviour with C99.  Avoid the
issue by making the operation work on unsigned 64 bit type instead.

This require libogg version to 1.3.4, raise autotools dependency check
to look for this.

Partly solves github issue #18.

Use multiplication instead, allowing the compiler to optimize to
bitshifts if it believe it to be safe.

Fixes a few issues discovered by --enable-gcc-sanitize-address in !31.

Use multiplication instead, allowing the compiler to optimize to
bitshifts if it believe it to be safe.

Partly solves github issue #18.

It is undefined what left-shifting a negative number should do, and the
compiler do not like trying to shift the -64 value 57 bits to the left.

Changed the definition of OC_Q57() to left shift one bit and
use multiplication, to change the operatoin to a well defined one.
Patch from Timothy B_ Terriberry.

This get rid of this clang message:

>  ../../lib/rate.c:746:15: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
>      log_scale=OC_Q57(-64);
>                ^~~~~~~~~~~
>  ../../lib/mathops.h:134:38: note: expanded from macro 'OC_Q57'
>  #define OC_Q57(_v) ((ogg_int64_t)(_v)<<57)
>                      ~~~~~~~~~~~~~~~~~^

MR !23 fixes #2321

This will make it easier for prospective users and developers to
locate the mailman lists.

Instead of shifting the negative number, shift one bit and multiply
it with the negative number.

Fixes the following clang message:
../../lib/decode.c:195:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0,10, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:198:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0,  0,1),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:208:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 1, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:210:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 2, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:212:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 3, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:214:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 4, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:216:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 5, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:220:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 1, -2,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:221:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 1, -3,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:227:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 6, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:228:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 7, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:229:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 8, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:230:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 9, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:236:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 2, -2,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:237:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 3, -2,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:238:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 2, -3,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:239:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 3, -3,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:242:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0,  0,1),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:253:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -1,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:257:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -2,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:260:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -3,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:262:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -4,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:264:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -5,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:266:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -6,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:270:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -7,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:271:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -8,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:277:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0, -9,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:278:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0,-10,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:279:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0,-11,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^
../../lib/decode.c:280:3: warning: shifting a negative signed value is undefined [-Wshift-negative-value]
  OC_DCT_CW_PACK( 0, 0,-12,0),
  ^~~~~~~~~~~~~~~~~~~~~~~~~~~
../../lib/decode.c:150:16: note: expanded from macro 'OC_DCT_CW_PACK'
 (_mag)-(_flip)<<OC_DCT_CW_MAG_SHIFT)
 ~~~~~~~~~~~~~~^

Fixes #2322

Switching type of cd from unsigned to int make sure the difference
passed to abs() wil be a signed value.

As dc is calculated to at most ((8 * 8 * (2^8-1)) + 32) / 64, ie 255
and the highest value it can hold before it is shifted 6 bit positions
to the right is (8 * 8 * (2^8-1)) + 32) ie. 16352, it should fit nicely
in an integer on any platform where int is at least 16 bit.

Fixes this clang message:

../../../lib/encfrag.c:104:26: warning: taking the absolute value of unsigned type 'unsigned int' has no effect [-Wabsolute-value]
    for(j=0;j<8;j++)sad+=abs(_src[j]-dc);
                         ^
../../../lib/encfrag.c:104:26: note: remove the call to 'abs' since unsigned values cannot be negative
    for(j=0;j<8;j++)sad+=abs(_src[j]-dc);
                         ^~~

This get rid of the following clang warning:

huffdec.c:439:7: warning: variable 'total' set but not used [-Wunused-but-set-variable]
  int total;
      ^

The OC_IDCT_CONSTS array in x86/sse2idct.c is 64 elements long, so
not use trying to pretend it is 128 elements long.

This avoid the following copmiler message:

In function 'oc_idct8x8_slow_mmx',
    inlined from 'oc_idct8x8_mmx' at ../../lib/x86/mmxidct.c:555:8:
../../lib/x86/mmxidct.c:290:3: warning: array subscript 'const struct <anonymous>[0]' is partly outside array bounds of 'const short unsigned int[64]' [-Warray-bounds]
  290 |   __asm__ __volatile__(
      |   ^~~~~~~
In file included from ../../lib/x86/mmxidct.c:20:
../../lib/x86/x86int.h: In function 'oc_idct8x8_mmx':
../../lib/x86/x86int.h:95:58: note: object 'OC_IDCT_CONSTS' of size 128
   95 | extern const unsigned short __attribute__((aligned(16))) OC_IDCT_CONSTS[64];
      |                                                          ^~~~~~~~~~~~~~
In function 'oc_idct8x8_10_mmx',
    inlined from 'oc_idct8x8_mmx' at ../../lib/x86/mmxidct.c:554:20:
../../lib/x86/mmxidct.c:492:3: warning: array subscript 'const struct <anonymous>[0]' is partly outside array bounds of 'const short unsigned int[64]' [-Warray-bounds]
  492 |   __asm__ __volatile__(
      |   ^~~~~~~
../../lib/x86/x86int.h: In function 'oc_idct8x8_mmx':
../../lib/x86/x86int.h:95:58: note: object 'OC_IDCT_CONSTS' of size 128
   95 | extern const unsigned short __attribute__((aligned(16))) OC_IDCT_CONSTS[64];
      |                                                          ^~~~~~~~~~~~~~
  CC       x86/sse2idct.lo
In function 'oc_idct8x8_slow_sse2',
    inlined from 'oc_idct8x8_sse2' at ../../lib/x86/sse2idct.c:453:8:
../../lib/x86/sse2idct.c:213:3: warning: array subscript 'const struct <anonymous>[0]' is partly outside array bounds of 'const short unsigned int[64]' [-Warray-bounds]
  213 |   __asm__ __volatile__(
      |   ^~~~~~~
../../lib/x86/sse2idct.c: In function 'oc_idct8x8_sse2':
../../lib/x86/sse2idct.c:26:56: note: object 'OC_IDCT_CONSTS' of size 128
   26 | const unsigned short __attribute__((aligned(16),used)) OC_IDCT_CONSTS[64]={
      |                                                        ^~~~~~~~~~~~~~
In function 'oc_idct8x8_10_sse2',
    inlined from 'oc_idct8x8_sse2' at ../../lib/x86/sse2idct.c:452:20:
../../lib/x86/sse2idct.c:398:3: warning: array subscript 'const struct <anonymous>[0]' is partly outside array bounds of 'const short unsigned int[64]' [-Warray-bounds]
  398 |   __asm__ __volatile__(
      |   ^~~~~~~
../../lib/x86/sse2idct.c: In function 'oc_idct8x8_sse2':
../../lib/x86/sse2idct.c:26:56: note: object 'OC_IDCT_CONSTS' of size 128
   26 | const unsigned short __attribute__((aligned(16),used)) OC_IDCT_CONSTS[64]={
      |                                                        ^~~~~~~~~~~~~~

The arrays are 5 elements long, but some intermittent methods claim it
was just four elements.

Made sure the implementation matches the prototype.

Rewrites 2fb974cf.

This will avoid running valgrind on bash and instead validate the
test programs themselves.

This avoid a crash on bogus input, just exit with a message instead.

Fixes #2305

Using ogg_uint32_t and uint8_t instead.

Get rid of the following compiler message:

../../examples/tiff2theora.c:283:1: warning: ‘uint32’ is deprecated [-Wdeprecated-declarations]
  283 | rgb_to_yuv(uint32 *raster,
      | ^~~~~~~~~~
../../examples/tiff2theora.c: In function ‘rgb_to_yuv’:
../../examples/tiff2theora.c:314:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  314 |         uint8 r0 = TIFFGetR(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:315:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  315 |         uint8 g0 = TIFFGetG(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:316:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  316 |         uint8 b0 = TIFFGetB(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:317:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  317 |         uint8 r1 = TIFFGetR(raster[(h-y)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:318:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  318 |         uint8 g1 = TIFFGetG(raster[(h-y)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:319:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  319 |         uint8 b1 = TIFFGetB(raster[(h-y)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:320:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  320 |         uint8 r2 = TIFFGetR(raster[(h-y1)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:321:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  321 |         uint8 g2 = TIFFGetG(raster[(h-y1)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:322:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  322 |         uint8 b2 = TIFFGetB(raster[(h-y1)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:323:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  323 |         uint8 r3 = TIFFGetR(raster[(h-y1)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:324:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  324 |         uint8 g3 = TIFFGetG(raster[(h-y1)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:325:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  325 |         uint8 b3 = TIFFGetB(raster[(h-y1)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:347:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  347 |         uint8 r = TIFFGetR(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:348:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  348 |         uint8 g = TIFFGetG(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:349:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  349 |         uint8 b = TIFFGetB(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:360:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  360 |         uint8 r0 = TIFFGetR(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:361:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  361 |         uint8 g0 = TIFFGetG(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:362:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  362 |         uint8 b0 = TIFFGetB(raster[(h-y)*w + x]);
      |         ^~~~~
../../examples/tiff2theora.c:363:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  363 |         uint8 r1 = TIFFGetR(raster[(h-y)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:364:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  364 |         uint8 g1 = TIFFGetG(raster[(h-y)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c:365:9: warning: ‘uint8’ is deprecated [-Wdeprecated-declarations]
  365 |         uint8 b1 = TIFFGetB(raster[(h-y)*w + x1]);
      |         ^~~~~
../../examples/tiff2theora.c: In function ‘tiff_read’:
../../examples/tiff2theora.c:387:3: warning: ‘uint32’ is deprecated [-Wdeprecated-declarations]
  387 |   uint32 width;
      |   ^~~~~~
../../examples/tiff2theora.c:388:3: warning: ‘uint32’ is deprecated [-Wdeprecated-declarations]
  388 |   uint32 height;
      |   ^~~~~~
../../examples/tiff2theora.c:390:3: warning: ‘uint32’ is deprecated [-Wdeprecated-declarations]
  390 |   uint32 *raster;
      |   ^~~~~~
../../examples/tiff2theora.c:404:3: warning: ‘uint32’ is deprecated [-Wdeprecated-declarations]
  404 |   raster = malloc(pixels*sizeof(uint32));
      |   ^~~~~~

The ftime method has been removed from POSIX and is depricated on Linux.
clock_gettime() work better on Linux, but is not available on Windows,
so fall back to ftime() if autotools/scons fail to detect clock_gettime().

Based on proposal in
opusfile!9.

Fixes the following compiler messages:

../../examples/dump_video.c: In function ‘main’:
../../examples/dump_video.c:490:5: warning: ‘ftime’ is deprecated: Use gettimeofday or clock_gettime instead [-Werror=deprecated-declarations]
  490 |     ftime(&start);
      |     ^~~~~
In file included from ../../examples/dump_video.c:40:
/usr/include/x86_64-linux-gnu/sys/timeb.h:29:12: note: declared here
   29 | extern int ftime (struct timeb *__timebuf)
      |            ^~~~~
../../examples/dump_video.c:491:5: warning: ‘ftime’ is deprecated: Use gettimeofday or clock_gettime instead [-Werror=deprecated-declarations]
  491 |     ftime(&last);
      |     ^~~~~
/usr/include/x86_64-linux-gnu/sys/timeb.h:29:12: note: declared here
   29 | extern int ftime (struct timeb *__timebuf)
      |            ^~~~~
../../examples/dump_video.c:505:13: warning: ‘ftime’ is deprecated: Use gettimeofday or clock_gettime instead [-Werror=deprecated-declarations]
  505 |             ftime(&after);
      |             ^~~~~
/usr/include/x86_64-linux-gnu/sys/timeb.h:29:12: note: declared here
   29 | extern int ftime (struct timeb *__timebuf)
      |            ^~~~~

The header used signed char _bf[256], make sure the method uses the
same.  Adjust pointer arithmetic to use array notation to make operation
explicit.

Avoid the following compiler message:

../../lib/state.c:1056:15: warning: argument 2 of type 'signed char *' declared as a pointer [-Warray-parameter=]
 1056 |  signed char *_bv,int _refi,int _pli,int _fragy0,int _fragy_end){
      |  ~~~~~~~~~~~~~^~~
In file included from ../../lib/state.c:20:
../../lib/state.h:526:14: note: previously declared as an array 'signed char[256]'
  526 |  signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
      |  ~~~~~~~~~~~~^~~~~~~~

This avoid obsolete notation.

Increase autoconf dependency to version 2.71.

Added info.c to libtheoraenc source and dropped library dependency on libtheoradec.

Addresses issue reported in https://bugs.debian.org/923940 about some symbols being unresolved.

Both POSIX, the Linux documentation and UNIX Network Programming volume 1
by Stephens state that NULL is accepted.

This avoid the following compiler message:

../../examples/player_example.c: In function ‘main’:
../../examples/player_example.c:839:22: warning: passing argument 2 to ‘restrict’-qualified parameter aliases with argument 4 [-Werror=restrict]
  839 |           n=select(n,&empty,&writefs,&empty,&timeout);
      |                      ^~~~~~          ~~~~~~
../../examples/player_example.c:843:18: warning: passing argument 2 to ‘restrict’-qualified parameter aliases with argument 4 [-Werror=restrict]
  843 |         select(n,&empty,&writefs,&empty,NULL);
      |                  ^~~~~~          ~~~~~~

Apparently this still has users, so test it, rather than
removing it. Verify we can build and clean up without error.

Add a build description for gitlab's continuous integration build
and test runners. This gives us ongoing converage on commmits
and merge requests in the upstream repository.

Based on the gcc:9 image we currently use for most projects.
It already includes make and autotools, but we need to install
dependencies for the example programs and documentation.

  - debian buster has fig2dev in its own package.
  - latex-extras is needed for wrapfig, which has been superceeded
    by lwarp-wrapfig.

NB defining a separate job at the test stage clears all the
generated artefacts, which means we lose our configure script
and makefiles. Therefore isn't more efficient to just do the
tests as extra steps in the build stage.