songsenand
/
cgal
mirror of https://github.com/CGAL/cgal
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
3,124 Commits 3 Branches 156 Tags 974 MiB
Commit Graph

380 Commits

Author SHA1 Message Date
Sylvain Pion b7e176e429 - Bump for version 4.11. 1999-05-04 16:08:27 +00:00
Sylvain Pion 60588365dc - Moved operator/ (double, Interval_nt), because could not be inlined. 1999-05-04 16:03:55 +00:00
Sylvain Pion 6c27739955 - indenting. 1999-05-04 15:52:14 +00:00
Sylvain Pion 308e05b1ec - Added (currently disabled) code to stop constant propagation inside the 
constructors.
 1999-05-04 15:51:33 +00:00
Sylvain Pion f45ba11b4e - Fix for MipsPro. 1999-05-04 12:14:06 +00:00
Sylvain Pion b08fa90fda - Removed operator+ (double) and operator- (double). 
- Added some rounding mode check (when CGAL_IA_DEBUG) where needed.
- Added some operator?? (double, Interval_nt) so that the rounding mode is ok.
 1999-05-04 12:08:33 +00:00
Sylvain Pion 0bb931e489 - Changed CGAL_assertion to CGAL_warning (when CGAL_IA_DEBUG is defined). 1999-05-04 09:53:34 +00:00
Sylvain Pion 947f394b1c - Fixed a bug in the division code !!!???!!! 
- Added operator/ (const double d, const Interval_nt & t).
- Removed useless () around boolean tests.
 1999-05-04 09:38:11 +00:00
Sylvain Pion 7edafed41a - Introduced the macro CGAL_IA_FORCE_TO_DOUBLE() to work around a Intel 
processor "feature" that breaks the interval computation when there is
  underflow or overflow (the exponent of the FP registers a larger than those
  of doubles).
 1999-05-03 16:07:14 +00:00
Sylvain Pion 134132c41a - Fix a typo breaking MipsPro. 
- First sketch for M$VC++.
 1999-04-30 09:53:25 +00:00
Sylvain Pion bf45af9c35 - Fixed a typo preventing use of assembly... 1999-04-22 14:05:07 +00:00
Sylvain Pion 28cedd7dd9 - Added 2 comments. 1999-04-20 16:46:55 +00:00
Sylvain Pion ebfc0541b7 - Replaced data members value/cache by _value/_cache, and added a public 
member function .value() instead.
 1999-04-20 16:41:57 +00:00
Sylvain Pion 82208c27f7 - Removed commented #includes. 
- The caching functionnality is now private inthe class rather than global
  scope, and is much cleaner now.
- Removed the copy operator, because it's useless.
 1999-04-20 16:31:09 +00:00
Sylvain Pion 688dfd4709 - Updated to remove a warning. 1999-04-20 13:10:24 +00:00
Sylvain Pion aa074c5ef0 - Cleanup of the caching functions. 1999-04-20 13:08:41 +00:00
Sylvain Pion 815ee80905 - unsafe_comparison is a type, not an object. 1999-04-20 13:07:51 +00:00
Sylvain Pion 11926a84a1 - Changed "__i386" to "__i386__" and similar. 1999-04-20 12:36:07 +00:00
Sylvain Pion 859c96feed - Simplify setting of CGAL_IA_USE_ASSEMBLY: all supported CPUs + GCC => ok. 1999-04-19 15:59:08 +00:00
Sylvain Pion 60f9cf8c08 - Implemented the new caching features. Still needs some testing and 
benchmarking.
 1999-04-16 15:34:04 +00:00
Sylvain Pion 46caccb005 - Workaround for lack of partial specialization, conditionned by 
CGAL_CFG_NO_EXPLICIT_TEMPLATE_FUNCTION_ARGUMENT_SPECIFICATION.
 1999-04-15 13:25:37 +00:00
Sylvain Pion 0e790c0677 - Changed "sign" to "CGAL::sign". 1999-04-15 11:21:41 +00:00
Sylvain Pion c5b8853cd1 - Changed compare to CGAL::compare (thanks LEDA). 1999-04-13 15:52:50 +00:00
Sylvain Pion bf7781caa0 - Updated to 2.0-I-8. 1999-04-13 15:50:10 +00:00
Sylvain Pion 09829b6da4 - Tried workaround for MipsPro based on convert_from_to() and 
CGAL_CFG_NO_EXPLICIT_TEMPLATE_FUNCTION_ARGUMENT_SPECIFICATION.
 1999-04-13 15:49:17 +00:00
Sylvain Pion 5cf818b3c6 - Tried a workaround for MipsPro based on: 
- convert_from_to(target, source)
  - CGAL_CFG_NO_EXPLICIT_TEMPLATE_FUNCTION_ARGUMENT_SPECIFICATION
 1999-04-13 15:48:17 +00:00
Sylvain Pion 6bd5bd6a71 - Removed cruft for the IRIX platform. Just kept the good workaround. 1999-04-13 14:03:22 +00:00
Sylvain Pion f5ca461671 - Several test possibilities for IRIX (all C versions fail on IRIX 6 !!!). 
Only kept for the future, will be cleaned up in the next commit.
 1999-04-13 13:53:43 +00:00
Sylvain Pion aef5b2afb5 - Bump for 3.5. 1999-04-01 15:50:47 +00:00
Sylvain Pion a7cd3afe15 - Removed commented include headers. 
- Made overlap_action return void, and modify the calling functions
  accordingly.
- Removed all friend functions, and replaced them using .inf() and .sup(), so
  that they don't need to be friend anymore (this should even make SunPro 5.0
  happy).
 1999-04-01 14:10:22 +00:00
Sylvain Pion 21b93bb834 - to_double, is_valid, is_finite, sign, compare do not need to be friend 
functions anymore, and they are not defined twice for the 2 classes.
- they use .inf() and .sup() to access the data members.
- overlap_action is now public.
 1999-04-01 10:15:12 +00:00
Sylvain Pion fec6af0b1c - Changed .inf to ._inf, same for sup. 
- Added .inf() and .sup().
 1999-04-01 08:58:51 +00:00
Sylvain Pion e036bf3979 - Include files changes. 1999-04-01 08:43:16 +00:00
Sylvain Pion 5298fc3390 - Use CGAL::to_double() for the LEDA NTs. 1999-03-30 14:07:43 +00:00
Sylvain Pion 2e51f92d39 - Added std:: for the iostream functions. 1999-03-26 17:33:13 +00:00
Sylvain Pion f7eb5d551c - Added std:: for the iostrem functions. 1999-03-26 17:32:36 +00:00
Sylvain Pion effc18ebf0 - Fix for IRIX. 1999-03-26 17:31:35 +00:00
Sylvain Pion ce960bc1cb - Updated for 2.0-I-5. 1999-03-19 20:07:31 +00:00
Sylvain Pion 496f9e4b57 - Replaced call of max<double>() to std::max. 1999-03-19 20:00:31 +00:00
Sylvain Pion 2aa2e5c7ae - Added a static data member "number_of_failures". 1999-03-12 16:52:58 +00:00
Sylvain Pion 62d24cdb20 - Fixed a Sign() vs sign() bug. 
- Re-formatting.
 1999-03-12 16:33:11 +00:00
Sylvain Pion eba314e01f - Use the new functions FPU_set_cw()/FPU_get_cw(). 1999-03-12 14:56:21 +00:00
Sylvain Pion 47f208026c - Shorten the names: 
FPU_get_control_word  =>  FPU_get_cw
   FPU_set_control_word  =>  FPU_set_cw
 1999-03-12 14:55:47 +00:00
Sylvain Pion 0bd1826da5 - Moved a few #define in _FPU.h. 
- Introduced the new FPU_get_cw() and FPU_set_cw().
 1999-03-12 14:54:37 +00:00
Sylvain Pion d34c5c0cd4 - Use the new get/set rounding mode functions. 1999-03-12 14:48:39 +00:00
Sylvain Pion 9731f06a82 - Replaced the static consts by enums. 1999-03-12 14:21:11 +00:00
Sylvain Pion 814d128d15 - Nearly complete rewrite, the get/set functions are working. 1999-03-12 14:00:58 +00:00
Sylvain Pion 5e9d7ec2e4 - Comments. 1999-03-11 15:19:05 +00:00
Sylvain Pion 04c28cf0db - Fixed template param. 1999-03-11 15:17:09 +00:00
Sylvain Pion c9c1ae46dc - Fixed template declaration problems. 
- Replaced Interval_nt::smallest by CGAL_IA_SMALLEST.
 1999-03-11 13:04:43 +00:00
Sylvain Pion 778d14af16 - Removed the line #include <CGAL/Interval_arithmetic.h>. 1999-03-08 09:40:03 +00:00
Sylvain Pion 29ee98b567 - STORAGE CHANGED: "inf" now stores the lower bound, not the opposite anymore. 
- => lots of "-" removals around the code.  Things are a LOT cleaner now,
  though probably slightly (<5%) slower.  Optimizing the compiler might be a
  good idea now.
- Added #include <CGAL/config.h> for CGAL_BEGIN_NAMESPACE.
- Replaced the static member constants by #defines, because the link fails
  otherwise (multiply defined symbols), and I _want_ constant propagation.
  CGAL_IA_MIN_DOUBLE, CGAL_IA_MAX_DOUBLE, CGAL_IA_SMALLEST, CGAL_IA_LARGEST
 1999-03-04 17:42:06 +00:00
Sylvain Pion 48e0e74899 - Replaced sqrt (double) by std::sqrt. 1999-03-04 17:32:59 +00:00
Sylvain Pion 69dff681ca - Updated for 2.0-I-2: namespace issues. 1999-03-04 17:20:49 +00:00
Sylvain Pion 374430045a - Reformating. 1999-03-04 14:50:21 +00:00
Sylvain Pion d24e34bf46 - New member function .is_point() {return (inf == sup);}. 1999-03-03 09:46:43 +00:00
Sylvain Pion f69b19e8ea - Some reformating 
- Added operators || (union) and && (intersection).
 1999-03-03 09:43:34 +00:00
Sylvain Pion 29ceffad58 - Changed stuff to use "struct converter". 1999-02-21 15:14:52 +00:00
Sylvain Pion 6601da2073 - Introduced namespace CGAL. 1999-02-19 17:17:07 +00:00
Sylvain Pion c1c83f6993 - Temporary version (waiting for "struct converter"). 
They will all need some re-writing.
 1999-02-19 15:13:41 +00:00
Sylvain Pion 200626696b - New flag CGAL_IA_DONT_USE_ASSEMBLY that disables assembly for all platforms 
(usefull for Irix 5.3 only at the moment).
- Fixed the asm constraint "g" to "m" in the x86 assembly version (egcs-1.2
  discovered it).
- Introduced a "typedef CGAL_FPU_CW_t" and use it (type storing the FPcw).
- Added CGAL_IA_GETFPCW() for Sparc and Alpha.
- Simplified Mips's assembly versions.
- Global static variable CGAL_FPU_CW to store the FPCW.
- Added CGAL_FPU_save_control_word() and CGAL_FPU_restore_control_word()
  (only assembly versions are working now).
  They are not used nor tested for the moment.
 1999-02-19 15:11:51 +00:00
Sylvain Pion 5942656090 - Standard header <iostream>. 
- Removed a few useless #include, since now I #include <IA.h>.
- Added a (forwarding) ctor with 2 parameters for Quotient<>(num,den).
- Added preliminary support for caching the converted value.
 1999-02-19 15:06:45 +00:00
Sylvain Pion 20f1a15273 - Standard header <iostream>. 
- Added operators (double, IA) for +,-,*,/,<,>...
- Changed PRECONDITION to ASSERTION.
- Default constructor is the buggy interval [1;-1] to detect use before
  definition in the user's code.
- Fine-tuned the multiplication (once again :).
- Same for division.
- Various code looking clean up.
- Removed a useless test in CGAL_sign()'s equality case.
- About the same for CGAL_compare().
 1999-02-19 15:04:17 +00:00
Sylvain Pion eb271e36dd - += and -= rely on + and -, no redefinition (it was for speed, but...). 
- Added conditional include for Quotient<>.
 1999-02-04 13:25:13 +00:00
Sylvain Pion 8fb164ffcc - First version for CGAL_convert_to(Quotient<RT>). 1999-02-02 14:00:37 +00:00
Sylvain Pion 9b6d7b2acd - get_rounding_mode simplified (2 cases are enough). 1999-02-02 13:47:40 +00:00
Sylvain Pion b8721294f2 - Added a 2 parameter constructor, because Quotient has one. 
- Added interval() and exact() member functions.
 1999-02-01 16:29:35 +00:00
Sylvain Pion 91714be61f - Adapted to the .interval() and so changes to the script. 1999-01-29 17:03:20 +00:00
Sylvain Pion f4213551e7 - moved conditionnal include lines at the beginning. 1999-01-19 16:17:36 +00:00
Sylvain Pion 8ac8cdf045 - Use CGAL_square<double>(), it's here now. 1999-01-05 19:14:32 +00:00
Sylvain Pion 5464c1b10d - Fixed bug in CGAL_square(): -a*a -> a*-a. 1998-12-23 12:18:50 +00:00
Sylvain Pion 99771e0853 - Typo. 1998-12-22 14:31:05 +00:00
Sylvain Pion c87231fc7a - Move template ctor before the specialised one. 1998-12-22 13:58:33 +00:00
Sylvain Pion 6c2d539e37 - New predicates for rtH2. It doesn't seem complete for 1.2 yet, though. 1998-12-22 13:57:20 +00:00
Sylvain Pion f5719011e0 - Write specialisations as they should be: 
template <> ... convert_to <IA> ().
 1998-12-22 13:48:42 +00:00
Sylvain Pion 6f0cf1d2f7 - CGAL_square(double) doesn't exist yet. 1998-12-22 12:51:17 +00:00
Sylvain Pion 263f481f20 - Replace the expensive by a condition on CGAL_IA_DEBUG, so that they are not 
enabled by default.
 1998-12-22 12:43:02 +00:00
Sylvain Pion 34f5af32c1 - Replace the expensive checks by the flag CGAL_IA_DEBUG. 1998-12-22 12:40:01 +00:00
Sylvain Pion 674de45cf9 - Added "const" for tags operators. 
- Splitted function headers.
 1998-12-21 14:17:02 +00:00
Sylvain Pion 6bc091291e - Added [expensive] post conditions to check the interval is really containing 
the initial value.  It might be removed one day because it's slow.
 1998-12-17 17:46:11 +00:00
Sylvain Pion ed1367a87e - Added CGAL_square forwarding. 1998-12-17 17:39:26 +00:00
Sylvain Pion 8924234ce8 - Added CGAL_square() for both types. 
- Rewrote the header of each function in a new style.
 1998-12-17 17:37:54 +00:00
Sylvain Pion 5b413a0fc6 - Assembly for Mips disabled again. 1998-12-16 16:49:15 +00:00
Sylvain Pion bd450f0bbe - removed "const" from istream(). 1998-12-15 17:42:02 +00:00
Sylvain Pion 6a0e233bf3 - It apppears that most CGAL_to_double() functions tend to depend on the 
rounding mode to be set to the nearest.  So I have to do it.  This is done
  for GMP and LEDA.
- For leda_rational, I use .to_double(), which is less error prone to
  overflows (but even slower), but I add 3 ulps, because it's the result of a
  FP division.
 1998-12-15 11:38:11 +00:00
Sylvain Pion 9b5f96f84f - New function CGAL_FPU_set_rounding_mode(). Currently based on a switch on 
the old functions.
 1998-12-13 18:33:32 +00:00
Sylvain Pion 3b7473d48a - Made only "epsilon" volatile, it should be faster. 1998-12-13 18:19:51 +00:00
Sylvain Pion 3443c802fa - Added lots of preconditions on the rounding mode. 1998-12-13 17:38:37 +00:00
Sylvain Pion 7a6195de19 - Added preconditions on the rounding mode, where necessary. 1998-12-13 17:38:05 +00:00
Sylvain Pion e8e91971c2 - Added enum CGAL_FPU_rounding_mode, and function 
CGAL_FPU_get_rounding_mode(), that tests empirically the current rounding
  mode.  To be used in the test-suite, and in some pre-conditions.
 1998-12-13 17:23:43 +00:00
Sylvain Pion 632645e683 - line too long fixed. 1998-12-11 11:47:01 +00:00
Sylvain Pion c03eda30e3 - Mips assembly re-enabled. 1998-12-10 16:51:32 +00:00
Sylvain Pion f82270de10 - #include <CGAL/misc.h> for CGAL_convert_to<>(). 1998-12-10 16:06:12 +00:00
Sylvain Pion 55b70d5e94 - Checks that CGAL_IA_NO_EXCEPTION is not defined ! 
- Conditionnal inclusions not made if already made.
 1998-12-08 17:48:25 +00:00
Sylvain Pion e6980cac88 - Big renaming... 
- Use typedef for IA.
 1998-12-08 17:23:07 +00:00
Sylvain Pion 1539586454 - Removed <IA> in declaration of CGAL_convert_to<IA>(). 1998-12-08 17:20:15 +00:00
Sylvain Pion d1bb64cb4a - Moved the CGAL_convert_to<>() general template here. 
(will be in kernel_basic.h normally).
 1998-12-08 17:16:34 +00:00
Sylvain Pion 49efafed43 - Commented out the CGAL_convert_to<>() definitions, that really should not be 
here.  Currently moved to Interval_arithmetic.h, but should be in
  number_utils.h, or in separate conversion files.
 1998-12-08 13:50:30 +00:00
Sylvain Pion 3ede82ecd2 - Renaming: CGAL_to_Interval_nt_advanced() to 
CGAL_convert_to<CGAL_Interval_nt_advanced>.
 1998-12-08 13:16:52 +00:00
Sylvain Pion 6e8540613c - Renamed CGAL_to_Interval_nt*() to CGAL_convert_to<CGAL_Interval_nt*>(). 1998-12-08 13:13:42 +00:00
Sylvain Pion 4b2bce06b4 - Big renaming: 
- /Filter/ 		-> /Arithmetic_filter/
  - Filter.h		-> Arithmetic_filter.h
  - CGAL_to_exact_type 	-> CGAL_convert_to
  - CGAL_FILTER_H 	-> CGAL_ARITHMETIC_FILTER_H
  - CGAL_Filtering 	-> CGAL_Filtered_exact
 1998-12-08 13:03:30 +00:00
Sylvain Pion 62fd519627 Filter.h renamed Arithmetic_filter.h. 1998-12-08 12:55:09 +00:00
Sylvain Pion 67e14022cd Moved to Arithmetic_filter/. 1998-12-08 12:53:48 +00:00
Sylvain Pion 4f3b42af42 Moved from Filter/. 1998-12-08 12:53:19 +00:00
Sylvain Pion 6f6f508da2 - Added conditional include for ftC3 and rtH3. 1998-12-08 12:32:42 +00:00
Sylvain Pion b9d343c117 New files. 1998-12-07 17:27:41 +00:00
Sylvain Pion 1c18283a42 - Added CGAL_compare_deltax_deltayC2() 
- Fixed the "inline" bug.
 1998-12-07 17:27:24 +00:00
Sylvain Pion e391699415 - Changed the script location. 
- Added ftC3 and rtH2.
- Fixed the generic make rule.
 1998-12-07 15:02:22 +00:00
Sylvain Pion 2ba0dea489 - Added member function bool overlap(const IA &d). 
- Added flag CGAL_IA_NO_EXCEPTION, which prevents throwing.
 1998-12-05 16:28:17 +00:00
Sylvain Pion dadd0d2470 - Default CGAL_to_exact_type<ET>(CT) based on ET(ct). 
- Added operator=()
- Added constructor(CGAL_Filtering<>), to avoid recursive construction.
 1998-12-05 14:55:43 +00:00
Sylvain Pion 65df7a8b81 - Added forgottent definition of CGAL_to_double(IA_nt). 
- Added istream operator>>(IA) (via a double).
 1998-12-01 16:39:56 +00:00
Sylvain Pion c84a8911f3 Use a template member function (the constructor)... and pray that it's 
supported by the commercial compilers...
 1998-12-01 14:36:37 +00:00
Sylvain Pion 35b57efb0e - Put back CGAL_DENY_INEXACT_OPERATIONS_ON_FILTER, since when deriving, the 
specialized predicates are not used...
 1998-12-01 14:08:24 +00:00
Sylvain Pion d058b1adb6 - Changes "class" to "struct", and removed "public:". 
- Added missing constructors for CGAL_Filter_allow_inexact.
 1998-12-01 13:52:39 +00:00
Sylvain Pion 030e4fbfcd - Introduced the class CGAL_Filtering_allow_inexact to replace 
CGAL_DENY_INEXACT_OPERATIONS_ON_FILTER
- Added forwarding code (to the CT value) for the functions:
  CGAL_is_valid, CGAL_is_finite, CGAL_to_double, CGAL_sign, CGAL_compare,
  CGAL_abs, CGAL_min, CGAL_max, sqrt, CGAL_io_tag, CGAL_number_type_tag,
  operator>>, operator<<
- Added conditional include directive for predicates_on_rtH2.h.
 1998-12-01 13:37:23 +00:00
Sylvain Pion 861307904b Cast function for Olivier's CGAL_Fixed_precision_nt. 1998-12-01 13:03:02 +00:00
Sylvain Pion 10afe69c95 Added conditionnal #include directive for IA/IA_Fixed.h. 1998-12-01 12:54:55 +00:00
Sylvain Pion ffc2b7df51 Added CGAL_number_type_tag() 1998-12-01 12:51:38 +00:00
Sylvain Pion 4913337b50 CGAL_io_tag() added. 1998-11-30 16:03:59 +00:00
Sylvain Pion e22af6cf31 Removed "multi-line // comment" warning. 1998-11-29 11:05:16 +00:00
Sylvain Pion 358ee282a1 Changed SETFPCW to CGAL_IA_SETFPCW, same for GETFPCW. 1998-11-29 11:03:22 +00:00
Sylvain Pion 5fddddc9be Changed __USE_ASSEMBLY to CGAL_IA_USE_ASSEMBLY. 1998-11-29 11:00:42 +00:00
Sylvain Pion e3c364c0ed Updated to latest C2's predicates_on_ftC2.h. 1998-11-26 16:50:57 +00:00
Sylvain Pion 83b8634348 - Added the flag CGAL_DENY_INEXACT_OPERATIONS_ON_FILTER 
- Added sqrt() forwarding.
 1998-11-25 16:51:17 +00:00
Sylvain Pion a4c088ca00 - Changed again the semantic of sqrt() when negative. 
- Made static const members for largest and smallest.
- Declared the values of {min,max}_double outside => CC accepts it now.
 1998-11-25 16:38:40 +00:00
Sylvain Pion 5439295901 Replaced CGAL_Interval_smallest by CGAL_Interval_nt_advanced::smallest. 1998-11-25 16:28:19 +00:00
Sylvain Pion 756b56cc01 Updated after the script. 1998-11-23 15:23:37 +00:00
Sylvain Pion e4b1b35ce4 - Added double member constants "min_double" and "max_double". 
- Added global constants CGAL_Interval_smallest and CGAL_Interval_largest.
- Added a template for CGAL_to_Interval_nt() and a specialisation for doubles.
 1998-11-23 15:19:19 +00:00
Sylvain Pion d7e1b23f69 Replaced "CGAL_Interval_nt_advanced::min_double" with "CGAL_Interval_smallest". 1998-11-23 15:07:03 +00:00
Sylvain Pion b1e465e8da - Completely reworked, changed to CGAL_to_Interval_nt_advanced(), supposing 
the rounding mode set to +infinity.
 1998-11-19 12:52:57 +00:00
Sylvain Pion 1c9b971740 Use more "typedef CGAL_Interval_nt IA". 1998-11-18 16:39:59 +00:00
Sylvain Pion d5273396c5 sqrt() semantic has changed. 1998-11-18 16:30:30 +00:00
Sylvain Pion d7eaf1fd54 Comment modifs. 1998-11-18 16:28:53 +00:00
Sylvain Pion 42c4802f46 Makefile to generate the filtered predicates files from those in $CGAL. 1998-11-18 15:38:33 +00:00
Sylvain Pion 286607c0fc Fixed bug for the empty ctor, revealed by goo'ol' g++. 1998-11-16 10:56:10 +00:00
Sylvain Pion 113263bfb3 Fixed typo. 1998-11-16 10:42:00 +00:00
Sylvain Pion 37bd9349e0 Commented assembly version for Mips. 1998-11-16 08:43:41 +00:00
Sylvain Pion 2bae6bfc58 Fixed line > 80char problem. 1998-11-12 16:40:01 +00:00
Sylvain Pion c47e4572cb Added conditionnal inclusion of <CGAL/Filter/predicates_on_ftC2.h>. 1998-11-12 16:11:33 +00:00
Sylvain Pion 2d75f4a70f - Passing "double" instead of "double &" seems faster, so I changed all 
concerned function prototypes.
- Moved ctor def CGAL_Interval_nt(const CGAL_Interval_nt_advanced &d) at the
  beginning of the class body so that it can be inlined.
- Redefined (grrr) friends CGAL_{abs,sign, compare, min...} for
  CGAL_Interval_nt.
- overlap_action() should maybe be redefined for CGAL_Interval_nt, but this
  should use a virtual function, so the code is ready, but I need the first
  benchmarks to activate it.
- New (undocumented for the moment) member function is_same() that does a
  "deep compare" (test equality of both bounds of 2 intervals).
- commented out the friendlyness of operator* (const double, const IA &).
 1998-11-12 16:10:42 +00:00
Sylvain Pion bc05e04689 First release. Not entirely produced by the script, but mostly. 1998-11-12 16:05:45 +00:00
Sylvain Pion 0596e62707 Modified comment aout egcs not defining __i386 (it's fixed). 1998-11-12 16:04:43 +00:00
Sylvain Pion 04810e2dc4 - Added "typedef CGAL_Interval_nt IA;", and done some substitutions. 1998-11-05 15:22:10 +00:00
Sylvain Pion 89fe4aa29b - CGAL_to_interval_nt() now returns a CGAL_Interval_nt_advanced. 1998-11-05 14:32:22 +00:00
Sylvain Pion bf02d922bc - Added assembly support for Mips with GNU compilers. 
- Removed C compatibility (inline, // comments).
 1998-11-05 14:31:01 +00:00
Sylvain Pion b1b92487d7 - Added specializations for the functions CGAL_{sign,compare,abs,min,max}. 
- overlap_action() returns an int(0), to shut up compiler, and provide
  a default answer, when we don't want the exception.
- The test operators now "return overlap_action();".
- One test less in operator==().
 1998-11-05 14:28:18 +00:00
Sylvain Pion 9cdc072750 First release. The wrapper works at least on a few cases. 1998-10-30 16:51:45 +00:00
Sylvain Pion 872c4c7fe1 Changed member functions is_valid() and is_finite() to be non member functions 
CGAL_is_valid(IA) and CGAL_is_finite(IA).
 1998-10-30 12:22:37 +00:00
Sylvain Pion fd6727c9a8 CGAL_to_interval_nt(double) now returns an IA_advanced. 1998-10-28 17:20:41 +00:00
Sylvain Pion 6256a74fa0 New empty member class unsafe_comparison. 
New private function overlap_action(), that throws the above exception, when
comparing unsafely.  CGAL_IA_DONT_THROW_EXCEPTION defines the behaviour at
compile time.
 1998-10-26 13:41:00 +00:00
Sylvain Pion f979f9578a Bump for 2.3: small cleanup. 1998-10-23 08:53:48 +00:00
Sylvain Pion 5a2090bde6 Changed "class" to "struct". 
Put the cast "IA_advanved -> IA" private.
Put the 2 doubles at the top of the definition.
 1998-10-22 13:23:12 +00:00
Sylvain Pion 7fa30708d0 Added for the 2 classes, the 2 operators: *(IA, double) and *(double, IA). 1998-10-21 16:51:40 +00:00
Sylvain Pion 814447927d - operator<< doesn't need to be a friend function anymore. 
- using the GNU extension optimization "Named return value".
- new protected function verify_overlap() (to clean up the comparison
  functions codes).
- operator*() and operator/() are outside the class body, so they can be made
  not inlined.
 1998-10-20 13:42:32 +00:00
Sylvain Pion 4f8701d25b 2 small comments added. 1998-10-16 14:30:50 +00:00
Sylvain Pion 60476ef3cf CGAL_Interval_nt operator-() needed to be defined... 1998-10-15 15:59:55 +00:00
Sylvain Pion 4fd875f2c2 Fixed (duplicate code...) the CGAL_interval_nt operator?=(). 1998-10-15 15:36:31 +00:00
Sylvain Pion bfc9ee9ef3 Put the operators' code outside the class body. 
Made operator<<() a friend function, with const argument.
 1998-10-15 11:58:32 +00:00
Sylvain Pion 7bd25e7203 - Changed the default constructor to empty, instead of initializing to 0. 
- Corrected is_valid() when inf = sup.
 1998-10-02 11:14:06 +00:00
Sylvain Pion f798ba6488 Fixed sqrt() so that it passes the test-suite (it was correct though). 1998-10-02 11:06:25 +00:00
Sylvain Pion 61cd9b457b Debuggued and improved the asm statements. 1998-09-25 12:01:38 +00:00
Sylvain Pion 3618f393e7 Speeded up a bit CGAL_Interval_nt, by removing spurious constructors. 1998-09-25 11:52:28 +00:00
Sylvain Pion 566b72d63e Removed useless "inline"s in the class definition. 
Removed old commented code.
 1998-09-23 15:42:32 +00:00
Sylvain Pion c8b319f9aa Removed useless and time consuming copy constructors. 1998-09-11 17:55:48 +00:00
Sylvain Pion d3189c9f27 "lines over 80 characters" problem fixed. 1998-08-18 07:20:47 +00:00
Sylvain Pion e2058f1837 Headers updated for CVS keywords. 1998-08-17 15:33:57 +00:00
Sylvain Pion efc75b318b Updated to 2.0. 1998-08-17 15:33:36 +00:00
Sylvain Pion ef69b25cc3 Use keyword substitution in the headers. 1998-08-17 13:57:58 +00:00
Sylvain Pion 243f7f40fc Changed subdirectory name IA -> Interval_arithmetic 1998-08-14 13:54:04 +00:00
Sylvain Pion 911bf9f945 The e-mail of the coordinator changed from Herve to Mariette. 1998-08-14 13:29:37 +00:00
Sylvain Pion b482a785af Removed old files from CVS 1998-08-14 11:59:41 +00:00
Sylvain Pion 86236d5def Initial revision 1998-08-14 11:46:12 +00:00
Sylvain Pion abf8645805 This commit was generated by cvs2svn to compensate for changes in r425, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:46:12 +00:00
Sylvain Pion b77af23941 This commit was generated by cvs2svn to compensate for changes in r423, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:33:46 +00:00
Sylvain Pion f6e708c35f This commit was generated by cvs2svn to compensate for changes in r421, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:33:29 +00:00
Sylvain Pion ff2e669c3e This commit was generated by cvs2svn to compensate for changes in r419, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:33:02 +00:00
Sylvain Pion d76b1fdd02 This commit was generated by cvs2svn to compensate for changes in r417, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:32:48 +00:00
Sylvain Pion fc9c84c747 This commit was generated by cvs2svn to compensate for changes in r415, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:32:31 +00:00
Sylvain Pion 929ec3f994 This commit was generated by cvs2svn to compensate for changes in r413, 
which included commits to RCS files with non-trunk default branches.
 1998-08-14 11:31:56 +00:00
Sylvain Pion e761721edd Initial revision 1998-08-14 11:31:15 +00:00