ginac.git
19 months agoPrepare for release 1.7.3 (yeah, again).
Richard Kreckel [Wed, 31 Jan 2018 22:07:35 +0000 (23:07 +0100)]
Prepare for release 1.7.3 (yeah, again).

19 months agoFix two comments in ginac/normal.cpp.
Richard Kreckel [Wed, 31 Jan 2018 22:00:08 +0000 (23:00 +0100)]
Fix two comments in ginac/normal.cpp.

And, while at it, make some disabled debugging output code compile when
it is enabled.

19 months agoFix out-of-bounds error in multivar_diophant().
Richard Kreckel [Wed, 31 Jan 2018 21:55:02 +0000 (22:55 +0100)]
Fix out-of-bounds error in multivar_diophant().

This could trigger an assertion in the factorization exams introduced
in afb0ccaa49 if the library was compiled with -DDO_GINAC_ASSERT.

19 months agoShortcut conversion ex->numeric->ex in expairseq::make_flat().
Richard Kreckel [Wed, 31 Jan 2018 11:18:37 +0000 (12:18 +0100)]
Shortcut conversion ex->numeric->ex in expairseq::make_flat().

Of course, passing an ex by reference is always faster than converting
it to a numeric first (even if it is a numeric).

19 months agoImprove gcd(a, b) where one argument is a power of a symbol.
Richard Kreckel [Wed, 31 Jan 2018 09:04:25 +0000 (10:04 +0100)]
Improve gcd(a, b) where one argument is a power of a symbol.

The already implemented recursion
  gcd(x^n, x*p(x)) -> x*gcd(x^(n-1), p(x))
is not ambitious enough: If p(x) has a factor of x, it just goes through
the same step again, and if p(x) has a factor which is a power of x, this
is reapeted many times.

Example:
  gcd(x^n, expand(x^n*(1+x)))
used to go recurse through the gcd routine n times, which could
easily lead to a stack overflow for n about 10^5.

To improve the situation, introduce a special case for gcd where one
argument is a power of a symbol and just cancel the common factor.

This turned out to be the root cause of segfaults in matrix elimination
algorithms, as reported by Patrick Schulz and Vitaly Magerya:
Cf. <https://www.ginac.de/pipermail/ginac-list/2018-January/thread.html>

19 months agoMake matrix::solve() work with non-normal zeros.
Richard Kreckel [Sun, 28 Jan 2018 23:07:17 +0000 (00:07 +0100)]
Make matrix::solve() work with non-normal zeros.

Normalize elements of augmented matrix before checking if they are zero.
This ensures that we don't divide by a non-normal zero. Also added a test.

This bug was reported by Vitaly Magerya <vmagerya@gmail.com>.

19 months ago[CHECK] Add some more factorization exams.
Richard Kreckel [Sun, 28 Jan 2018 18:59:54 +0000 (19:59 +0100)]
[CHECK] Add some more factorization exams.

Add the 15 multivariate polynomials from P. S. Wang's paper "An
Improved Multivariate Polynomial Factoring Algorithm" as exams.

Don't add them as benchmarks since timings vary considerably
depending on internal choice of variables. In any case, this should
never take hours. (But before 630db9a0b0 it did, occasionally.)

19 months agoAdd optional algorithm selection to matrix::inverse().
Richard Kreckel [Sun, 28 Jan 2018 18:23:50 +0000 (19:23 +0100)]
Add optional algorithm selection to matrix::inverse().

Following a proposal by Vitaly Magerya, a signature with a solve_algo
is added to class matrix. (Not using a default argument in order to
not break the ABI.)

19 months agoFix previous commit 'Speed up special cases of square-free...'.
Richard Kreckel [Thu, 25 Jan 2018 18:59:00 +0000 (19:59 +0100)]
Fix previous commit 'Speed up special cases of square-free...'.

Huh, I swear 4ffb3cbb3 worked for me! Sorry for that.

19 months agoSpeed up special cases of square-free factorization.
Richard Kreckel [Wed, 24 Jan 2018 21:37:24 +0000 (22:37 +0100)]
Speed up special cases of square-free factorization.

Square-free factorization of polynomials containing a factor which is
a high power P of a symbol x did scale like O(P) in space and time.
This patch introduces a shortcut in Yun's algorithm, such that the
computation is only O(1) in space and time.

This makes it possible to compute, say sqrfree(x^P + x^(P+1)) =>
(1+x)*x^P with P=123456789.

Found this to be a bottleneck while debugging one of Vitaly Magerya's
examples.

20 months agoPrepare for release 1.7.3.
Richard Kreckel [Mon, 1 Jan 2018 20:21:41 +0000 (21:21 +0100)]
Prepare for release 1.7.3.

20 months agoHappy New Year!
Richard Kreckel [Mon, 1 Jan 2018 19:47:26 +0000 (20:47 +0100)]
Happy New Year!

20 months ago[BUGFIX] Speed up some cases of polynomial factorization.
Richard Kreckel [Fri, 29 Dec 2017 21:16:54 +0000 (22:16 +0100)]
[BUGFIX] Speed up some cases of polynomial factorization.

Speed up a naive O(N) modulus operation in multivar_diophant().
This loop caused some polynomials to seemingly hang forever.

20 months agoDo not expand pre-factored polynomials.
Vitaly Magerya [Mon, 18 Dec 2017 18:28:01 +0000 (19:28 +0100)]
Do not expand pre-factored polynomials.

See <https://www.ginac.de/pipermail/ginac-list/2017-December/002155.html>.

20 months agoMinor sreamlining of some internal code using class lst...
Richard Kreckel [Mon, 18 Dec 2017 18:19:34 +0000 (19:19 +0100)]
Minor sreamlining of some internal code using class lst...

...without any external effect at all.

2 years agoAdd LaTeX pretty-print for function derivatives.
Vladimir V. Kisil [Tue, 12 Sep 2017 21:07:59 +0000 (23:07 +0200)]
Add LaTeX pretty-print for function derivatives.

2 years agoUpdate m4/ax_cxx_compile_stdcxx.m4 to current upstream version.
Richard Kreckel [Tue, 12 Sep 2017 20:32:20 +0000 (22:32 +0200)]
Update m4/ax_cxx_compile_stdcxx.m4 to current upstream version.

2 years ago[PATCH] Fix bug in evaluation of real and imaginary parts of powers.
Vladimir V. Kisil [Mon, 11 Sep 2017 21:12:36 +0000 (23:12 +0200)]
[PATCH] Fix bug in evaluation of real and imaginary parts of powers.

Be more careful when simplifying Re(a^c) -> a^c and Im(a^c) -> 0.

Thanks to Jan Rheinländer for reporting this.

2 years agoIntroducing method symbol::get_TeX_name() to match existing symbol::get_name().
Vladimir V. Kisil [Mon, 17 Jul 2017 07:37:21 +0000 (09:37 +0200)]
Introducing method symbol::get_TeX_name() to match existing symbol::get_name().

Signed-off-by: Vladimir V. Kisil <kisilv@maths.leeds.ac.uk>
2 years agoFix excessive backslashes in the LaTeX function names.
Vladimir V. Kisil [Mon, 17 Jul 2017 07:35:40 +0000 (09:35 +0200)]
Fix excessive backslashes in the LaTeX function names.

Signed-off-by: Vladimir V. Kisil <kisilv@maths.leeds.ac.uk>
2 years agochecks: prefer initializer lists over push_back on vectors.
Richard Kreckel [Fri, 17 Feb 2017 22:32:19 +0000 (23:32 +0100)]
checks: prefer initializer lists over push_back on vectors.

2 years agoClean up combinatorial helpers.
Richard Kreckel [Wed, 15 Feb 2017 11:58:48 +0000 (12:58 +0100)]
Clean up combinatorial helpers.

* Convert all signatures from int to unsiged.
* Renamed partition_generator to partition_with_zero_parts_generator...
* ...and add class partition_generator which does not include zero parts.
*  Rename ::current() to ::get() and make them compute vectors on first use.

2 years agoMove combinatorial helpers from power.cpp to utils.h.
Richard Kreckel [Wed, 15 Feb 2017 11:42:00 +0000 (12:42 +0100)]
Move combinatorial helpers from power.cpp to utils.h.

This way, these can be used by other modules.

2 years agoFinalize 1.7.2 release. release_1-7-2
Richard Kreckel [Tue, 10 Jan 2017 07:16:08 +0000 (08:16 +0100)]
Finalize 1.7.2 release.

2 years agoHappy New Year!
Richard Kreckel [Mon, 9 Jan 2017 22:38:38 +0000 (23:38 +0100)]
Happy New Year!

Update copyright statements.

2 years agoFix memory leaks in operators returning relationals.
Richard Kreckel [Mon, 9 Jan 2017 21:49:11 +0000 (22:49 +0100)]
Fix memory leaks in operators returning relationals.

These memory leaks were introduced in 82df71852.

Thanks to Thomas Luthe for reporting this.

2 years agoGINAC_DECLARE_REGISTERED_CLASS(): add namespace of status_flags.
Stefan Weinzierl [Tue, 3 Jan 2017 20:43:42 +0000 (21:43 +0100)]
GINAC_DECLARE_REGISTERED_CLASS(): add namespace of status_flags.

2 years agoFix MSC compiler warning.
Richard Kreckel [Sun, 6 Nov 2016 18:41:47 +0000 (19:41 +0100)]
Fix MSC compiler warning.

Thanks to Jan Rheinländer for reporting this.

2 years agoClean up some header files and fix compilation with MSC.
Richard Kreckel [Sun, 6 Nov 2016 18:35:46 +0000 (19:35 +0100)]
Clean up some header files and fix compilation with MSC.

On MSC there is no <unistd.h>. It turns out that we need to #include <io.h>
in order to use close(3). This fixes the build failure introduced in 8305ec38.

Thanks to Jan Rheinländer for reporting this.

2 years agoRemove configure check for type 'long long'.
Richard Kreckel [Sun, 6 Nov 2016 17:39:08 +0000 (18:39 +0100)]
Remove configure check for type 'long long'.

This type with at least 64 bits is mandatory since C++11.

2 years agoReplace mktemp with mkstemp.
Till Hofmann [Sat, 8 Oct 2016 18:34:40 +0000 (20:34 +0200)]
Replace mktemp with mkstemp.

As the manpage says, mktemp is deprecated ("Never use mktemp") and gcc
warns about the usage of mktemp. Replace mktemp with its safe variant
mkstemp.

2 years agoFinalize 1.7.1 release. release_1-7-1
Richard Kreckel [Sat, 1 Oct 2016 22:46:08 +0000 (00:46 +0200)]
Finalize 1.7.1 release.

2 years agoSuppress compiler warnings about inconsistent override.
Richard Kreckel [Sat, 1 Oct 2016 22:45:11 +0000 (00:45 +0200)]
Suppress compiler warnings about inconsistent override.

The introduction of the 'override' keyword in GiNaC in d5b86dd10 was
deliberately incomplete: Duplication of code in macros was avoided.
It turns out that some compilers (e.g. g++ -Wsuggest-override or clang++
-Wall) complain if the 'override' keyword is used inconsistently. Let's
make this consistent now, even if it leads to the introduction of two
more macros.

3 years agoFix compilation of programs using __attribute__((deprecated)).
Richard Kreckel [Wed, 14 Sep 2016 20:00:29 +0000 (22:00 +0200)]
Fix compilation of programs using __attribute__((deprecated)).

Ouu, we mustn't redefine 'deprecated' in order for this to work...

Thanks to Alexander Proskurin for reporting this.

3 years agoUpdate ax_cxx_compile_stdcxx.m4.
Richard Kreckel [Wed, 17 Aug 2016 20:08:57 +0000 (22:08 +0200)]
Update ax_cxx_compile_stdcxx.m4.

Serial 3 fixes make CXXFLAGS=..., c.f. the macro's upstream revision history.

3 years agoImprove performance of clifford_to_lst().
Vladimir Kisil [Wed, 17 Aug 2016 20:05:13 +0000 (22:05 +0200)]
Improve performance of clifford_to_lst().

Rewrite get_clifford_comp() so it is less obscure and more efficient.

3 years agoFix convergence of Li_projection at ±I.
Stefan Weinzierl [Thu, 11 Aug 2016 21:55:59 +0000 (23:55 +0200)]
Fix convergence of Li_projection at ±I.

For arguments near ±I, direct summation does not practically converge.
Use the Bernoulli transformation instead.

Thanks to Peter Banks for reporting this:
<http://www.cebix.net/pipermail/ginac-list/2016-August/002085.html>

3 years agoAdd proper functions to make clifford_bar() and clifford_star().
Vladimir Kisil [Thu, 11 Aug 2016 18:47:08 +0000 (20:47 +0200)]
Add proper functions to make clifford_bar() and clifford_star().

Previously operations clifford_bar() and clifford_star() called
conjugate() method. This results in reversion of all non-commutative
entries (not only Clifford units) and produced complex conjugation
of all non-real items. The new routine operates on Clifford units only.

3 years ago[bugfix] Fix crash in basic::subs().
Richard Kreckel [Thu, 21 Jul 2016 06:55:44 +0000 (08:55 +0200)]
[bugfix] Fix crash in basic::subs().

A regression in 1.7.0 was introduced with 1b8bcb06 in function
basic::subs_one_level(): implicitly casting *this to an ex first for
finding *this in m and later in the function's return statement caused
a crash in the second cast because *this was deleted in the first one.
After all, *this was dynamically allocated in basic::subs().

This bug was reported and hunted down by Mario Prausa.

3 years agoFinalize 1.7.0 release. release_1-7-0
Richard Kreckel [Sat, 30 Apr 2016 20:39:42 +0000 (22:39 +0200)]
Finalize 1.7.0 release.

3 years agoMore shortcuts for trivial cases in mul::combine_*_to_pair().
Richard Kreckel [Fri, 29 Apr 2016 05:31:20 +0000 (07:31 +0200)]
More shortcuts for trivial cases in mul::combine_*_to_pair().

3 years agoUpdate reference to Robert C. Martin's "Acyclic Visitor" paper.
Richard Kreckel [Mon, 18 Apr 2016 20:18:20 +0000 (22:18 +0200)]
Update reference to Robert C. Martin's "Acyclic Visitor" paper.

The source at objectmentor.com has been unavailble for a long while.

3 years agoAdd fderivative::derivatives() method.
Richard Kreckel [Thu, 14 Apr 2016 19:29:37 +0000 (21:29 +0200)]
Add fderivative::derivatives() method.

3 years agoFix typos in comments.
Richard Kreckel [Wed, 13 Apr 2016 06:50:41 +0000 (08:50 +0200)]
Fix typos in comments.

3 years agoUpdate compiler list and compile times.
Richard Kreckel [Wed, 13 Apr 2016 06:15:44 +0000 (08:15 +0200)]
Update compiler list and compile times.

3 years agoPrefer C library wrappers over C headers, where possible.
Richard Kreckel [Wed, 13 Apr 2016 06:10:22 +0000 (08:10 +0200)]
Prefer C library wrappers over C headers, where possible.

3 years agoMerge branch 'c++11' in preparation for version 1.7.0.
Richard Kreckel [Tue, 5 Apr 2016 20:51:12 +0000 (22:51 +0200)]
Merge branch 'c++11' in preparation for version 1.7.0.

3 years ago[build] Install compiler.h header file. c++11
Richard Kreckel [Mon, 4 Apr 2016 18:42:12 +0000 (20:42 +0200)]
[build] Install compiler.h header file.

Some header files need the deprecated macro. (Looking forward to C++14.)

3 years agoInstall ginac-excompiler in $LIBEXECDIR, not in $BINDIR.
Richard Kreckel [Mon, 4 Apr 2016 07:05:50 +0000 (09:05 +0200)]
Install ginac-excompiler in $LIBEXECDIR, not in $BINDIR.

...and make the GiNaC library aware of where it is installed.

The ginac-excompiler script is only invoked by GiNaC::compile_ex(...)
and serves no purpose on its own. In compliance with the FHS, it should
be installed in $LIBEXECDIR, not in $BINDIR. This also disburdens
distribution packagers from having to provide a manpage (which may be
required for all programs in $BINDIR).

The location for $LIBEXECDIR defaults to ${prefix}/libexec/. It may be
overwritten at configuration time. (Many distributions want to set it to
${prefix}/lib/ginac/.)

3 years ago[build] Ship FindLibDL.cmake, it's needed by CMakeLists.txt.
Richard Kreckel [Sun, 3 Apr 2016 15:46:47 +0000 (17:46 +0200)]
[build] Ship FindLibDL.cmake, it's needed by CMakeLists.txt.

3 years agoRemove ex::to_rational(lst) and ex::to_polynomial(lst).
Richard Kreckel [Sat, 12 Mar 2016 23:02:57 +0000 (00:02 +0100)]
Remove ex::to_rational(lst) and ex::to_polynomial(lst).

Those were GiNaC 1.1 compatibility functions. Use the corresponding
functions with exmap arguments instead.

3 years agoFinalize 1.6.7 release. release_1-6-7
Richard Kreckel [Sat, 6 Feb 2016 11:51:39 +0000 (12:51 +0100)]
Finalize 1.6.7 release.

3 years agoUse C++11 'auto' and range-based for loops where possible.
Richard Kreckel [Thu, 25 Feb 2016 21:12:28 +0000 (22:12 +0100)]
Use C++11 'auto' and range-based for loops where possible.

3 years agoUse C++11 'nullptr' where applicable.
Richard Kreckel [Mon, 22 Feb 2016 21:51:40 +0000 (22:51 +0100)]
Use C++11 'nullptr' where applicable.

3 years ago[build] Force C++11 compilation when using CMake.
Richard Kreckel [Thu, 11 Feb 2016 22:10:55 +0000 (23:10 +0100)]
[build] Force C++11 compilation when using CMake.

3 years ago[build] Fix CMake build.
Richard Kreckel [Thu, 11 Feb 2016 22:08:12 +0000 (23:08 +0100)]
[build] Fix CMake build.

3 years agominor optimization in power::expand_add
Richard Kreckel [Sat, 6 Feb 2016 22:19:35 +0000 (23:19 +0100)]
minor optimization in power::expand_add

3 years agoFix a comment.
Richard Kreckel [Sat, 6 Feb 2016 11:53:14 +0000 (12:53 +0100)]
Fix a comment.

3 years agospeed up multiply_lcm(e, lcm) a bit
Richard Kreckel [Sat, 6 Feb 2016 00:35:07 +0000 (01:35 +0100)]
speed up multiply_lcm(e, lcm) a bit

and comment it a little

3 years ago[bugfix] fix elusive bug in quo, rem,...
Richard Kreckel [Fri, 5 Feb 2016 23:47:08 +0000 (00:47 +0100)]
[bugfix] fix elusive bug in quo, rem,...

The power of two rational numeric objects needs not be rational. As a
result, some care is required when transforming (b^e)*l -> (b*l^(1/e))^e
for some rational e and l. This is a common transformation in order to
convert a polynomial over Q into a polynomial over Z when computing
their quotient, remainder, etc. Failure to be careful can potentially
introduce spurious non-rational numbers into rational polynomials and
make those operations fail. This patch avoids this transformation when
l^(1/e) is not a rational number.

3 years ago[bugfix] fix elusive bug in quo, rem,...
Richard Kreckel [Fri, 5 Feb 2016 23:47:08 +0000 (00:47 +0100)]
[bugfix] fix elusive bug in quo, rem,...

The power of two rational numeric objects needs not be rational. As a
result, some care is required when transforming (b^e)*l -> (b*l^(1/e))^e
for some rational e and l. This is a common transformation in order to
convert a polynomial over Q into a polynomial over Z when computing
their quotient, remainder, etc. Failure to be careful can potentially
introduce spurious non-rational numbers into rational polynomials and
make those operations fail. This patch avoids this transformation when
l^(1/e) is not a rational number.

3 years ago[bugfix] fix sqrfree(poly) for zero polynomials in disguise.
Richard Kreckel [Wed, 3 Feb 2016 07:13:58 +0000 (08:13 +0100)]
[bugfix] fix sqrfree(poly) for zero polynomials in disguise.

Yun's algorithm now handles polynomials which would become zero after
expanding.

3 years ago[bugfix] fix sqrfree(poly) for zero polynomials in disguise.
Richard Kreckel [Wed, 3 Feb 2016 07:13:58 +0000 (08:13 +0100)]
[bugfix] fix sqrfree(poly) for zero polynomials in disguise.

Yun's algorithm now handles polynomials which would become zero after
expanding.

3 years agoRemove extra cases for missing <stdint.h>.
Richard Kreckel [Sun, 31 Jan 2016 09:15:50 +0000 (10:15 +0100)]
Remove extra cases for missing <stdint.h>.

Use <cstdint> unconditionally -- it's been standardized in C++11.

3 years ago[C++17] Remove inheritance from std::binary_function, std::unary_function.
Richard Kreckel [Fri, 29 Jan 2016 22:45:21 +0000 (23:45 +0100)]
[C++17] Remove inheritance from std::binary_function, std::unary_function.

This inheritance is unneeded since C++11 and the two classes may be
removed in C++-17.

3 years ago[C++17] Replace std::bind2nd with a lambda.
Richard Kreckel [Fri, 29 Jan 2016 21:57:52 +0000 (22:57 +0100)]
[C++17] Replace std::bind2nd with a lambda.

3 years agoRemove global variable max_recursion_level.
Richard Kreckel [Thu, 28 Jan 2016 23:11:14 +0000 (00:11 +0100)]
Remove global variable max_recursion_level.

With the removal of the last 'level' parameter, this variable has
become obsolete.

3 years agoRemove 'level' argument of normal().
Richard Kreckel [Thu, 28 Jan 2016 21:45:56 +0000 (22:45 +0100)]
Remove 'level' argument of normal().

The 'level' argument was modeled after that of the eval() methods
(removed in 6c946d4c). It has never been very useful except for
confusing developers and it hasn't been documented in the tutorial.
Moreover, I have no indication that it has ever been used.

3 years agoRemove 'level' argument of evalf().
Richard Kreckel [Thu, 28 Jan 2016 21:11:46 +0000 (22:11 +0100)]
Remove 'level' argument of evalf().

The 'level' argument was modeled after that of the eval() methods
(removed in 6c946d4c). It has never been very useful except for
confusing developers. Moreover, I have no indication that it has
ever been used.

3 years agoAvoid x^0 and Order(x^0) terms together in series expansion.
Richard Kreckel [Wed, 6 Jan 2016 21:34:23 +0000 (22:34 +0100)]
Avoid x^0 and Order(x^0) terms together in series expansion.

At the branch cut, the series expansions of log(), atan(), and atanh()
assembled pseries objects which contained both x^0 and Order(x^0) terms.
This patch removes the extra x^0 term if the order is 0. It also adds
a GINAC_ASSERT for these kinds of invariants to the pseries ctor and
simplifies the loops in pseries' degree(), ldegree(), eval() and evalf()
member functions.

3 years agoHappy New Year!
Richard Kreckel [Fri, 1 Jan 2016 10:49:48 +0000 (11:49 +0100)]
Happy New Year!

3 years agoHappy New Year!
Richard Kreckel [Fri, 1 Jan 2016 01:05:06 +0000 (02:05 +0100)]
Happy New Year!

3 years agoDe-document eval() in ginsh help and manpage.
Richard Kreckel [Fri, 1 Jan 2016 01:02:49 +0000 (02:02 +0100)]
De-document eval() in ginsh help and manpage.

This was forgotten in 6c946d4c.

3 years agoMake relational::rhs() and lhs() nonvirtual.
Richard Kreckel [Thu, 31 Dec 2015 18:34:54 +0000 (19:34 +0100)]
Make relational::rhs() and lhs() nonvirtual.

I am unable to see a point in ever overriding these two member functions.

3 years agoClarification on symmetries of metric of clifford object.
Vladimir V. Kisil [Thu, 31 Dec 2015 15:44:16 +0000 (16:44 +0100)]
Clarification on symmetries of metric of clifford object.

The metric of a clifford object may be non-symmetric. Even if
a metric is defined by a symmetric tensor, clifford object may
not be aware of the symmetry and it needs to be explicitly declared.

Signed-off-by: Vladimir V. Kisil <kisilv@maths.leeds.ac.uk>
3 years agoAdd specific dbgprinttree() method to clifford.
Vladimir V. Kisil [Thu, 31 Dec 2015 15:41:48 +0000 (16:41 +0100)]
Add specific dbgprinttree() method to clifford.

Metric is an important element of a clifford object, thus it is useful
to have the metric to be printed in the detailed debug printing.

Signed-off-by: Vladimir V. Kisil <kisilv@maths.leeds.ac.uk>
3 years agocache pseries coeff accesses in pseries::mul_series.
Ralf Stephan [Thu, 31 Dec 2015 15:36:46 +0000 (16:36 +0100)]
cache pseries coeff accesses in pseries::mul_series.

Coeff is too general as long as we have only ints as exponents.
This patch fixes a potentially cubic slowdown.

3 years agoFinalize 1.6.6 release. ginac_1-6-6
Richard Kreckel [Sun, 20 Dec 2015 20:29:55 +0000 (21:29 +0100)]
Finalize 1.6.6 release.

3 years agoRemove reference to ftpthep.physik.uni-mainz.de.
Richard Kreckel [Sun, 20 Dec 2015 16:56:09 +0000 (17:56 +0100)]
Remove reference to ftpthep.physik.uni-mainz.de.

Higgs is down, alas...

3 years agoRemove reference to ftpthep.physik.uni-mainz.de.
Richard Kreckel [Sun, 20 Dec 2015 16:56:09 +0000 (17:56 +0100)]
Remove reference to ftpthep.physik.uni-mainz.de.

Higgs is down, alas...

3 years agoHoist typeid(*this) from loops.
Richard Kreckel [Sun, 20 Dec 2015 16:50:07 +0000 (17:50 +0100)]
Hoist typeid(*this) from loops.

The compilers don't seem to reliably recognize the invariance of this.

3 years agoRemove iter_swap specializations.
Richard Kreckel [Sun, 20 Dec 2015 16:30:34 +0000 (17:30 +0100)]
Remove iter_swap specializations.

It's a myth that iter_swap helps the compiler produce better code.
Let's just use specialized swap instead.

3 years agoAdd support for power::info(info_flags::real).
Richard Kreckel [Sun, 20 Dec 2015 03:44:01 +0000 (04:44 +0100)]
Add support for power::info(info_flags::real).

(And, while at it, clean up the surrounding code a bit.)

3 years agoRemove info_flags::algebraic.
Richard Kreckel [Sun, 20 Dec 2015 03:36:42 +0000 (04:36 +0100)]
Remove info_flags::algebraic.

It doesn't work, never has, and it is quire unclear how to fix it.

3 years agoRemove expairseq::construct_from_2_ex_via_exvector() member function.
Richard Kreckel [Thu, 17 Dec 2015 23:06:31 +0000 (00:06 +0100)]
Remove expairseq::construct_from_2_ex_via_exvector() member function.

This function is much slower than the optimized construct_from_2_ex()
member function, which explains why it wasn't used.

3 years agoinclude <algorithm> for use of std::min().
Richard Kreckel [Thu, 17 Dec 2015 21:57:04 +0000 (22:57 +0100)]
include <algorithm> for use of std::min().

Thanks to Jan Rheinländer for reporting this.

Cherry-picked from a56196f3.

3 years agoinclude <algorithm> for use of std::min().
Richard Kreckel [Thu, 17 Dec 2015 21:57:04 +0000 (22:57 +0100)]
include <algorithm> for use of std::min().

Thanks to Jan Rheinländer for reporting this.

3 years agoRemove useless code in add::eval().
Richard Kreckel [Thu, 17 Dec 2015 21:45:50 +0000 (22:45 +0100)]
Remove useless code in add::eval().

With commit ae6c004b, we have actually a more rigorous solution for the
bug fixed first (and wrongly) on September 23 2010 with commit 89d5356b.
This patch removes the now useless code and adds the new regression check
from master, just in case.

3 years ago[bugfix] Make sure add::eval() collects all numeric terms *correctly*.
Richard Kreckel [Thu, 17 Dec 2015 21:30:13 +0000 (22:30 +0100)]
[bugfix] Make sure add::eval() collects all numeric terms *correctly*.

This fixes a bug introduced with in 2010 with 89d5356b.

3 years agoAdd rvalue reference ctors from epvectors for add, mul.
Richard Kreckel [Thu, 17 Dec 2015 08:32:15 +0000 (09:32 +0100)]
Add rvalue reference ctors from epvectors for add, mul.

3 years agoMerge some cosmetic patches.
Richard Kreckel [Wed, 16 Dec 2015 21:02:25 +0000 (22:02 +0100)]
Merge some cosmetic patches.

This is a collection of some minor optimizations and indentation fixes.

3 years agoMake .eval() evaluate top-level only.
Richard Kreckel [Wed, 16 Dec 2015 20:22:36 +0000 (21:22 +0100)]
Make .eval() evaluate top-level only.

With the previous patch, some old workarounds have become unnecessary:
If all expressions are evaluated, any object which is an aggregate of
expressions will only ever have to evaluate the top level. As such, it
has become pointless to evaluate child objects of an expression prior
to doing its own term-rewriting. This patch removes the evaluation of
children from all GiNaC objects. It also removes the now superfluous
parameter 'level' of the eval methods.

3 years agoMake add::eval(), mul::eval() work without compromise.
Richard Kreckel [Wed, 16 Dec 2015 12:00:30 +0000 (13:00 +0100)]
Make add::eval(), mul::eval() work without compromise.

If a user asks to evaluate an object, the expectation is that the
library returns an evaluated, canonical expression. Everything else is
a bug. (It doesn't matter whether the user asks explicitly or by
assigning to an ex.) It turns out that it was possible to construct
objects which didn't fully evaluate. This patch fixes this by making
eval() a little bit more careful.

This obsoletes the need to go through combine_ex_with_coeff_to_pair()
when constructing an epvector that is then used to construct an add
or mul. (Alas, this was omitted frequently enough...)

3 years agoImprove method of setting status_flags::dynallocated.
Richard Kreckel [Mon, 30 Nov 2015 22:06:53 +0000 (23:06 +0100)]
Improve method of setting status_flags::dynallocated.

There seems to be no way to obsolete the need to mark an object derived
from basic and handled by ex as being 'on the heap', at least none that
doesn't have significant performance impact. Having said that, this
patch aims at making this process simpler and more intuitive.

Where, before, one would return from a function returning an ex with
        return (new mul(a, b))->setflag(status_flags::dynallocated);
this patch lets us return with
        return dynallocate<mul>(a, b);
which should be much clearer. In any case, it involves less typing.

The two points where the status_flags::dynallocated are set are now
  * the dynallocate<B>(args...) template function and
  * the virtual duplicate() member functions.

This patch rolls out the new functionality throughout the library.

3 years agoRemove accidentally pushed binary.
Richard Kreckel [Sun, 29 Nov 2015 11:42:59 +0000 (12:42 +0100)]
Remove accidentally pushed binary.

In commit 79cad108, a compiled binary from the test suite was accidentally
added and pushed. Sorry.

3 years agoRename array of precomputed data in test suite.
Richard Kreckel [Sun, 29 Nov 2015 11:33:59 +0000 (12:33 +0100)]
Rename array of precomputed data in test suite.

Reason: C++17 may introduce a std::data<> template. Right now, the GCC 6.0
prerelease bails out at this code, when compiler with -std=c++17.
<http://en.cppreference.com/w/cpp/iterator/data>

3 years agoUse neseted initializer lists to construct matrix objects.
Richard Kreckel [Sat, 28 Nov 2015 14:43:46 +0000 (15:43 +0100)]
Use neseted initializer lists to construct matrix objects.

Add constructor of initializer_list<initializer_list<ex>> to matrix.
Use this syntax where, previously, ctor from comma-separated list of
elements was used. Deprecate the ctor from comma-separated list.

Note: The output format '[[a,b],[c,d]]' and ginsh syntax are
unchanged because lists are printed '{a,b,c}' and a matrix is not a
list of lists.

3 years ago[build] Update scripts for correct rpath for linking with CLN.
Richard Kreckel [Fri, 27 Nov 2015 10:49:16 +0000 (11:49 +0100)]
[build] Update scripts for correct rpath for linking with CLN.

Use up-to-date files from GNU gettext 0.19.6.

3 years agoUse initializer lists to construct container<>, lst.
Richard Kreckel [Thu, 26 Nov 2015 20:36:46 +0000 (21:36 +0100)]
Use initializer lists to construct container<>, lst.

Add constructor of initializer_list<ex> to container<C<ex>>. In
particular, this means that we can finally declare lst objects as
lst{a,2*b,2}. Convert GiNaC to this syntax throughout. Deprecate
the old constructors taking 1..16 ex parameters and the ones from
comma seaparated lists (without braces).

3 years agoInfrastructure: Update baseline language check M4 macros.
Richard Kreckel [Thu, 26 Nov 2015 14:04:52 +0000 (15:04 +0100)]
Infrastructure: Update baseline language check M4 macros.

Use the more general ax_cxx_compile_stdcxx.m4 macro from gnu.org
instead of the old ax_cxx_compile_stdcxx_11.m4.