Changes commited to PARI GIT during the last two weeks

lfunqf: add support for odd dimensional lattices

This commit changes k from a logn to a GEN. t_INT and t_FRAC are
supported.

2	2	doc/usersch8.tex
47	43	src/basemath/lfun.c
30	29	src/basemath/lfunutils.c
2	2	src/basemath/mftrace.c
13	1	src/functions/l_functions/lfunqf
2	2	src/headers/paridecl.h

eint'(2) did not work

This patch works this way:
1) add a new opcode OCpackargs that takes the arity of the function and
return a t_VEC with the arguments and a t_VECSMALL which indicates
whether the argument was actually present.
2) add a new function closurecallgendefprec which take both vectors
to dispatch the call which the correct default value.

8	0	doc/usersch5.tex
3	3	src/functions/sums/derivnum
2	0	src/headers/paridecl.h
1	1	src/headers/paripriv.h
3	10	src/language/compile.c
50	0	src/language/eval.c
1	1	src/language/opcode.h
5	3	src/language/sumiter.c

fix diagnostic

1	1	src/basemath/base2.c

86- lfunzeros(f, [0, b]) wasn't allowed

1	0	CHANGES
1	1	src/basemath/lfun.c
2	1	src/test/32/lfun
1	0	src/test/in/lfun

cleanup closure_derivn: write f''' [n quotes] as derivn(f, n) of f' for n = 1

1	1	doc/usersch5.tex
21	32	src/language/compile.c
21	6	src/test/32/deriv
4	1	src/test/in/deriv

typo in lfunparams: maxss -> maxdd

1	1	src/basemath/lfun.c
4	4	src/test/32/lfun

85- n-th derivative of zetahurwitz used 'seriesprecision' instead of n [#2130]

+ imprecise value when n large [#2130]

2	0	CHANGES
2	2	src/basemath/trans3.c
3	1	src/test/32/zetahurwitz
1	0	src/test/in/zetahurwitz

doc: fix typo (from Jacques Gélinas)

1	1	doc/develop.tex

doc: adress->address (from CG)

1	1	doc/develop.tex

Doc: ZM_pseudoinv: fix from Loic

1	1	doc/usersch5.tex

50- ??? include double quotes when needed (e.g. ???conductor) [#2122]

1	0	CHANGES
5	1	doc/gphelp.in

84- matdet(mat with t_RFRAC entries): wrong result [#2128]

1	0	CHANGES
4	12	src/basemath/alglin1.c
2	1	src/test/32/det
2	0	src/test/in/det

improve doc

6	4	src/functions/number_theoretical/ffgen

move and cleanup ZV_union_shallow* + uniformize with merge_sort_uniq

The latter remains 25% slower.

39	46	src/basemath/bibli2.c

fix C++ compilation

1	1	src/basemath/bibli2.c

fix bnrclassfield: don't corrupt bnr

1	1	src/basemath/kummer.c
3	1	src/test/32/bnrclassfield
6	0	src/test/in/bnrclassfield

doc typo [fix gphtml translation]

2	2	src/functions/number_fields/bnrclassfield

doc typos

1	1	src/functions/programming/version
2	2	src/functions/sums/intnumromb

cleanups bnrclassfield

84	112	src/basemath/kummer.c

cleanup bnrclassfield_tower: use listP

9	11	src/basemath/kummer.c

83- nfhnfmod: wrong result

Fix nfcleanmod

1	0	CHANGES
11	9	src/basemath/base5.c

add bnrclassfield

2	1	doc/refcard-nf.tex

setunion: call merge_sort_uniq

3	19	src/basemath/bibli2.c

fix bnrclassfield bench (32-bit)

3	8	src/test/32/bnrclassfield
10	4	src/test/in/bnrclassfield

MPQS_DEBUG: check_root: fix signedness

1	1	src/modules/mpqs.c

bnrclassfield_tower: fix C89 compat

1	1	src/basemath/kummer.c

sqrtn(Mod(3,19),4) -> no error [#2127]

1	0	CHANGES
1	1	src/basemath/arith1.c
3	2	src/test/32/sqrtn
1	0	src/test/in/sqrtn

kill warnings

2	2	src/basemath/base5.c
2	2	src/basemath/kummer.c

new function bnrclassfield

1	0	CHANGES
366	0	src/basemath/kummer.c
32	0	src/functions/number_fields/bnrclassfield
2	2	src/functions/number_fields/rnfkummer
1	0	src/headers/paridecl.h
241	0	src/test/32/bnrclassfield
79	0	src/test/in/bnrclassfield

fix rnfpseudobasis doc

1	1	CHANGES
12	13	src/functions/number_fields/rnfpseudobasis

49- rnfdisc(nf, [T,B]) allow B a list of primes and maximal ideals

(same for rnfbasis, rnfinit, rnfconductor)

2	0	CHANGES
19	8	src/basemath/base2.c
25	0	src/basemath/base4.c
38	11	src/basemath/base5.c
3	4	src/basemath/buch3.c
5	0	src/functions/number_fields/rnfbasis
3	4	src/functions/number_fields/rnfconductor
2	3	src/functions/number_fields/rnfdisc
14	6	src/functions/number_fields/rnfinit
1	0	src/headers/paridecl.h
3	3	src/headers/paripriv.h
27	1	src/test/32/rnf
5	0	src/test/in/rnf

kill warning

1	1	src/language/init.c

fix documentation of primecert*, change human-readable format: a4 -> a

1	1	src/basemath/ecpp.c
8	7	src/functions/number_theoretical/primecert
6	3	src/functions/number_theoretical/primecertexport
7	7	src/test/32/isprime

cleanup extra accuracy code in mfgaexpansion and mfsymbol_i

12	11	src/basemath/mftrace.c

fix catastrophic cancellation in mf_eisendec

8	3	src/basemath/mftrace.c
3	2	src/test/32/mf
5	1	src/test/in/mf

bid_get_fact2: clearer doc

4	4	doc/usersch6.tex

51- [libpari] bid_get_fact2

1	0	CHANGES
5	0	doc/usersch6.tex

50- new GP function idealismaximal

1	1	CHANGES
1	0	doc/refcard-nf.tex
83	0	src/basemath/base4.c
33	0	src/functions/number_fields/idealismaximal
1	0	src/headers/paridecl.h
7	1	src/test/32/ideal
7	0	src/test/in/ideal

add arity

1	0	doc/refcard.tex

50- new GP function idealdown

1	0	CHANGES
1	0	doc/refcard-nf.tex
20	0	src/basemath/base4.c
21	0	src/functions/number_fields/idealdown
1	0	src/headers/paridecl.h
6	1	src/test/32/ideal
5	0	src/test/in/ideal

whitespace edits

1	4	src/basemath/base3.c

improve doc of galpol functions

4	3	src/functions/number_fields/galoisgetgroup
11	9	src/functions/number_fields/galoisgetname
5	3	src/functions/number_fields/galoisgetpol

81- (f(x) = x*'y); f'(1) -> error

1	0	CHANGES
2	1	src/language/sumiter.c
3	0	src/test/32/deriv
5	0	src/test/in/deriv

80- ispower(2, 2^64) -> error

1	0	CHANGES
4	1	src/basemath/arith1.c
3	1	src/test/32/ispower
3	0	src/test/in/ispower

fix poldisc and polresultant wrt auxiliary variable

78- poldisc(y^2/x + 1, y) -> error [also affects polresultant]
79- poldisc(x^2/y + 1, y) -> 0 [also affects polresultant]

2	0	CHANGES
32	31	src/basemath/polarit2.c
6	1	src/test/32/disc
2	1	src/test/32/resultant
5	0	src/test/in/disc
2	0	src/test/in/resultant

rewrite addsub_frac. No icopy in overlapping ranges (use icopy_avma)

37	66	src/basemath/gen1.c

ff: add test for large matrices (from Peter Bruin)

8	5	src/test/32/ff
10	3	src/test/in/ff

Add gen_matmul_sw (from Peter Bruin)

204	7	src/basemath/alglin1.c

gen_ker: add missing reduction

2	1	src/basemath/alglin1.c

CHANGES: F56: fix description, add missing a=

1	1	CHANGES

ellnf_rootno_global: handle curves over nf in x correctly

13	5	src/basemath/elliptic.c
4	4	src/test/in/ellnf

gen_Gauss_pivot: add missing reduction

1	0	CHANGES
1	1	src/basemath/alglin1.c

allocatemem: typo in doc

5	5	src/functions/programming/allocatemem

gen_det: reduce after addition

2	1	src/basemath/alglin1.c

gen_CUP_gauss: rename to gen_CUP_basecase to avoid confusion

3	3	src/basemath/Flv.c
3	3	src/basemath/alglin1.c

Flm_CUP_gauss: use Fl_addmul_pre

3	2	src/basemath/Flv.c

whitespace edits

15	15	src/basemath/alglin1.c

GEN (*mul)(...) functions: remove variable names in argument types

15	15	src/basemath/alglin1.c

prevent coeff explosion in gen_CUP_gauss and gen_det

2	2	src/basemath/alglin1.c

Flm_CUP: use get_Fl_red

93	83	src/basemath/Flv.c

New libpari function Flm_mul_pre

4	0	doc/usersch5.tex
9	0	src/basemath/FpV.c
1	0	src/headers/paridecl.h