Changes commited to PARI GIT during the last two weeks

remove thetabrackets: inline into lfunquadmodular*

32	38	src/basemath/lfunquad.c

compute vecRCpol once

24	31	src/basemath/lfunquad.c

comments + missing constbern

3	2	src/basemath/lfunquad.c

typo

1	1	src/functions/modular_forms/mfeisenstein

poleval(,0): 2 -> imin

1	1	src/basemath/gen3.c

poleval: special case y = 0

2	2	src/basemath/gen3.c

lfunquad: vstwist -> vs + reformat comments

27	26	src/basemath/lfunquad.c

simplify Hcol / mfDcoefs

6	8	src/basemath/lfunquad.c

avoid useless computations in Hcol

43	22	src/basemath/lfunquad.c

lfunquad: change variable names

74	74	src/basemath/lfunquad.c

add comments

4	3	src/basemath/lfunquad.c

cleanup thetabrackets

12	14	src/basemath/lfunquad.c

cleanup thetabrackets

18	7	src/basemath/lfunquad.c

lfunquad cleanups

14	18	src/basemath/lfunquad.c

cleanup sigsum / sigsumtwist

16	12	src/basemath/lfunquad.c

document LFEk cost + replace phi(D) by D in usefeq

5	4	src/basemath/lfunquad.c

cleanup lfunquad

9	12	src/basemath/lfunquad.c

Z_XpN_powu only worked for u=1

32	30	src/basemath/ZX.c

ZX_translate: gerepilecopy -> gerepileupto

1	1	src/basemath/ZX.c

use ZX_translate in eulervec + fix bench

7	30	src/basemath/bern.c
15	15	src/test/32/bern

cleanup ratlift code

45	41	src/basemath/nffactor.c

Flx_translate1: use divide and conquer strategy

27	8	src/basemath/Flx.c

ZX_translate: use divide and conquer strategy

59	2	src/basemath/ZX.c

replace lift_to_frac by lift_to_frac_tdenom (FpX_ratlift)

+ use Qdivii

8	19	src/basemath/nffactor.c

retune lfunquad + make sigsum static

5	16	src/basemath/lfunquad.c

Flx_Newton: minor speedup

4	3	src/basemath/Flx.c

Hfind: inverseimage -> QM_gauss

1	1	src/basemath/mftrace.c

typo

1	1	src/basemath/lfunquad.c

Hcolgen: refactor hard-coded threshold + change value: 40 -> 10

7	5	src/basemath/lfunquad.c

\frac -> \dfrac

1	1	src/functions/transcendental/eulerfrac

LFEk: improve

2	1	src/basemath/lfunquad.c

remove usumdivktwist_0_all: k is always odd and the Euler numbers are all 0

5	12	src/basemath/lfunquad.c

18- [libpari] lfunquadneg

1	0	CHANGES
3	0	doc/usersch8.tex

update bench

3	3	src/test/32/lfunquad

17- new GP functions: eulerfrac, eulerpol, eulervec

1	0	CHANGES
1	0	doc/refcard.tex
52	2	src/basemath/bern.c
7	6	src/basemath/lfunquad.c
16	0	src/functions/transcendental/eulerfrac
12	0	src/functions/transcendental/eulerpol
21	0	src/functions/transcendental/eulervec
3	1	src/headers/paridecl.h
52	9	src/test/32/bern
13	3	src/test/in/bern

RgXQX_translate: avoid using GEN *

5	4	src/basemath/RgX.c

Move ZX_translate to ZX.c

0	49	src/basemath/RgX.c
50	0	src/basemath/ZX.c

RgX_translate: do not assume c to be integer

2	2	src/basemath/RgX.c

RgX_translate: avoid using GEN *

8	8	src/basemath/RgX.c

ZX_translate: avoid using GEN *

8	8	src/basemath/RgX.c

lfun(t_QUAD, positive t_INT): fix accuracy problem

2	2	src/basemath/lfun.c
2	2	src/test/32/lfun

lfun(t_LFUN_ZETA,n):  use lfunquad

1	1	src/basemath/lfun.c
2	2	src/test/32/lfun

bnrclassfield: recover if incorrect nf

10	1	src/basemath/kummer.c

update / fix factorint bench

8	3	src/test/32/factorint

cleanup ifac_insert_multiplet + move to avoid forward decl.

82	99	src/basemath/ifactor1.c

cleanup ifac_insert_multiplet

5	5	src/basemath/ifactor1.c

fix "integer class" bug in mpqs [#2165]

3	3	src/modules/mpqs.c
4	0	src/test/in/factorint

parse.y: replace %pure-parser by %define api.pure full

1	1	src/language/parse.y

new libpari functions F2xX_to_F2xC, F2xXV_to_F2xM

1	1	CHANGES
4	0	doc/usersch5.tex
22	0	src/basemath/F2x.c
2	0	src/headers/paridecl.h

remove obsolete comment

0	2	src/modules/mpqs.c

usefeq: remove unused parameter N

3	3	src/basemath/lfunquad.c

improve lfunquadfeq

5	4	src/basemath/lfunquad.c

speedup RCpol (use recurrence)

16	19	src/basemath/lfunquad.c

23- direct formulas for quadratic Dirichlet L-functions at integers

1	0	CHANGES
31	3	src/basemath/alglin1.c
29	0	src/basemath/lfun.c
752	0	src/basemath/lfunquad.c
2	2	src/basemath/mftrace.c
6	0	src/headers/paridecl.h
472	0	src/test/32/lfunquad
15	0	src/test/in/lfunquad

whitespace edit

2	2	src/basemath/trans1.c

16- better formula for Catalan

1	0	CHANGES
12	12	src/basemath/trans1.c

missing 'static'

1	1	src/basemath/alglin1.c

QXQ_inv_filter: removed, use simpler code

7	34	src/basemath/polarit3.c

24- division by 0 in Flx_extresultant when one input is 0

1	0	CHANGES
3	3	src/basemath/Flx.c

23- modular symbols package: better type checks

1	0	CHANGES
2	1	src/basemath/modsym.c
10	7	src/test/32/modsym
5	2	src/test/in/modsym

QXQ_inv_filter: fix pathological case

24	13	src/basemath/polarit3.c
16	1	src/test/32/polmod
5	0	src/test/in/polmod

QXQ_inv: detect when A, B not coprime

2	0	src/basemath/polarit3.c
2	2	src/test/32/div
2	4	src/test/32/pol

remove QM_ker_r

3	5	src/basemath/modsym.c

ZX_gcd_filter: simplify interface

8	12	src/basemath/QX_factor.c
8	0	src/test/32/gcdext
4	0	src/test/in/gcdext

typo in QM_inv (wrong result)

+ remove code duplication + better take denominator into account in QM_inv

1	0	CHANGES
38	42	src/basemath/alglin1.c
3	3	src/test/32/mat

msfromell: use a smaller prime for Flm_ker

1	1	src/basemath/modsym.c

new libpari function gen_ZpM_Dixon with simpler interface

2	1	CHANGES
3	1	COMPAT
23	0	src/basemath/Hensel.c
3	0	src/headers/paridecl.h

rename gen_ZpM_Dixon to gen_ZpM_Dixon_Wiedemann

1	0	CHANGES
1	0	COMPAT
1	1	doc/usersch5.tex
3	3	src/basemath/FpV.c
1	1	src/headers/paridecl.h

New libpari function F2xn_inv

1	1	CHANGES
4	0	doc/usersch5.tex
66	0	src/basemath/F2x.c
4	1	src/basemath/RgX.c
1	0	src/headers/paridecl.h

New libpari function F2xn_red

1	0	CHANGES
2	0	doc/usersch5.tex
15	0	src/basemath/F2x.c
1	0	src/headers/paridecl.h

ZlM_gauss_ratlift: use better algorithm

22	18	src/basemath/alglin1.c

ZM_mul_i: fix tuning

1	1	src/basemath/ZV.c

ZM_mul: add fast parallel CRT implementation

95	10	src/basemath/ZV.c
6	0	src/functions/symbolic_operators/mul
1	0	src/headers/paripriv.h

fix #2164

2	1	src/basemath/buch2.c
2	1	src/test/32/nf
4	0	src/test/in/nf

14- parsum: removed optional argument [allows grouping of subsums => faster]

1	0	CHANGES
1	0	COMPAT
18	7	doc/parallel.tex
1	1	doc/refcard.tex
7	5	src/functions/programming/parsum
1	1	src/headers/paridecl.h
38	25	src/language/eval.c

cleanup polmodular0_ZM

5	5	src/basemath/polmodular.c

improve polmodular_worker use

- avoid potential crash when looking up dinf->primes[i] for i>= nprimes
- avoid allocating vectors + utoipos

21	18	src/basemath/polmodular.c
1	1	src/functions/polynomials/polmodular
1	1	src/headers/paripriv.h

QXQ_inv: use asymptocally fast algorithm

94	38	src/basemath/polarit3.c
6	0	src/functions/symbolic_operators/div
1	0	src/headers/paripriv.h
4	2	src/test/32/pol

Missing function nxCV_chinese_center

4	0	doc/usersch5.tex
20	0	src/basemath/arith1.c
1	0	src/headers/paridecl.h

mtsingle_err_recover: fix warning

1	0	src/mt/mt.c

ZX_gcd_slice: use gerepileupto instead of gerepileall

1	2	src/basemath/QX_factor.c

mt single: improve mt_is_thread emulation

In particular, avoid _*_worker to appear in error message

2	0	src/headers/paripriv.h
1	0	src/headers/paristio.h
8	0	src/language/eval.c
15	0	src/mt/mt.c
1	0	src/mt/mt.h
1	1	src/mt/single.c

new libpari function ZpM_invlift

1	1	CHANGES
8	0	doc/usersch5.tex
20	0	src/basemath/Hensel.c
1	0	src/headers/paridecl.h

new libpari function gen_ZpM_Newton

1	0	CHANGES
4	0	doc/usersch5.tex
40	0	src/basemath/Hensel.c
3	0	src/headers/paridecl.h