对于用 f77 编写的代码，我在使用 gfortran 编译器时遇到以下错误。我正在使用 -std=legacy 来克服不兼容问题。但是，gfortran 仍然会产生以下错误。我已经尝试了几种方法，但似乎无法摆脱它。非常感谢任何明确的帮助。 这是错误消息，接下来是产生错误的代码。谢谢!
gfortran -ff2c -std=legacy -Ofast -C -o bin/ref_syn erzsol3.f qbessel.f qfcoolr.f qbessel.f:136.29:
COMPLEX FUNCTION BESHS0*16(X, IFAIL)
1
错误:函数定义中预期的形式参数列表位于 (1)
DOUBLE PRECISION FUNCTION BESFJ0(X)
C BESSEL FUNCTION J0(X)
C
DOUBLE PRECISION X
DOUBLE PRECISION SRNAME
DOUBLE PRECISION A, B, C, CX, G, SX, T2, T, TBPI, XBIG, XVSMAL, Y
DOUBLE PRECISION DCOS, DSIN, DSQRT
C ..
DATA SRNAME /8H BESFJ0 /
DATA XBIG,XVSMAL,TBPI/3.4D+15,3.2D-9,6.36619772367581343D-1/
C
T = DABS(X)
C ERROR 1 TEST
IF (T.GT.XBIG) GO TO 60
C X RANGE TEST
IF (T.GT.8.0D0) GO TO 40
C SMALL X
C TEST FOR VERY SMALL X
IF (T.GT.XVSMAL) GO TO 20
BESFJ0 = 1.0D0
GO TO 80
20 T = 3.125D-2*T*T - 1.0D0
T2 = 2.0D0*T
C
C
A= +4.125321000000D-14
B=T2*A -1.943834690000D-12
C=T2*B-A +7.848696314000D-11
A=T2*C-B -2.679253530560D-9
B=T2*A-C +7.608163592419D-8
C=T2*B-A -1.761946907762D-6
A=T2*C-B +3.246032882101D-5
B=T2*A-C -4.606261662063D-4
C=T2*B-A +4.819180069468D-3
A=T2*C-B -3.489376941141D-2
B=T2*A-C +1.580671023321D-1
C=T2*B-A -3.700949938726D-1
A=T2*C-B +2.651786132033D-1
B=T2*A-C -8.723442352852D-3
Y=T*B-A +1.577279714749D-1
C
BESFJ0 = Y
GO TO 80
C
C LARGE X
40 G = T - 0.5D0/TBPI
Y = DSQRT(TBPI/T)
CX = DCOS(G)*Y
SX = -DSIN(G)*Y*8.0D0/T
T = 128.0D0/(T*T) - 1.0D0
C
C
Y= +9.99457276D-1+T*( -5.36366929D-4+T*( +6.13732440D-6+
AT*( -2.06823782D-7+T*( +1.30451717D-8))))
C
G= -1.55551139D-2+T*( +6.83314931D-5+T*( -1.47708215D-6+
AT*( +7.10458046D-8+T*( -5.81753275D-9+T*( +6.75219505D-10)))))
C
C
BESFJ0 = Y*CX + G*SX
GO TO 80
C
60 WRITE(6,61) SRNAME
61 FORMAT(1X,//,'ERROR IN ',A8)
BESFJ0 = DSQRT(TBPI/T)
C
80 RETURN
END
DOUBLE PRECISION FUNCTION BESFJ1(X)
C BESSEL FUNCTION J1(X)
C
DOUBLE PRECISION X
DOUBLE PRECISION SRNAME
DOUBLE PRECISION A, B, C, CX, G, SX, T2, T, TBPI, XBIG, XVSMAL, Y
DOUBLE PRECISION DCOS, DSIN, DSQRT
DATA SRNAME /8H BESFJ1 /
DATA XVSMAL,XBIG,TBPI/3.2D-9,3.4D+15,6.36619772367581343D-1/
C
T = DABS(X)
C ERROR 1 TEST
IF (T.GT.XBIG) GO TO 60
C X RANGE TEST
IF (T.GT.8.0D0) GO TO 40
C SMALL X
Y = 4.0D0
C TEST FOR VERY SMALL X
IF (T.LE.XVSMAL) GO TO 20
T = 3.125D-2*T*T - 1.0D0
T2 = 2.0D0*T
C
A= -5.777404200000D-13
B=T2*A +2.528123664000D-11
C=T2*B-A -9.424212981600D-10
A=T2*C-B +2.949707007278D-8
B=T2*A-C -7.617587805400D-7
C=T2*B-A +1.588701923993D-5
A=T2*C-B -2.604443893486D-4
B=T2*A-C +3.240270182684D-3
C=T2*B-A -2.917552480615D-2
A=T2*C-B +1.777091172397D-1
B=T2*A-C -6.614439341345D-1
C=T2*B-A +1.287994098858D+0
A=T2*C-B -1.191801160541D+0
Y=T*A-C +6.483587706053D-1
C
C
20 BESFJ1 = Y*X*0.125D0
GO TO 80
C
C LARGE X
40 G = T - 1.5D0/TBPI
Y = DSIGN(DSQRT(TBPI/T),X)
CX = DCOS(G)*Y
SX = -DSIN(G)*Y*8.0D0/T
T = 128.0D0/(T*T) - 1.0D0
C
C
Y= +1.00090703D+0+T*( +8.98804504D-4+T*( -7.95959347D-6+
AT*( +2.47105358D-7+T*( -1.49751260D-8))))
C
C
C
G= +4.67768740D-2+T*( -9.62145905D-5+T*( +1.82113970D-6+
AT*( -8.29018528D-8+T*( +6.58335466D-9+T*( -7.49818190D-10)))))
C
C
BESFJ1 = Y*CX + G*SX
GO TO 80
C ERROR 1 EXIT
60 WRITE(6,61) SRNAME
61 FORMAT(1X,//,'ERROR IN ',A8)
BESFJ1 = DSQRT(TBPI/T)
C
80 RETURN
END
COMPLEX FUNCTION BESHS0*16(X, IFAIL)
C COMPLEX*16 BESHS0
C-----------------------------------------------------------------------
C BESSEL FUNCTION HS0(X)
C
C COMPUTED USING CHEBYSHEV REPRESENTATION
C
C-----------------------------------------------------------------------
C
DOUBLE PRECISION X
INTEGER IFAIL
DOUBLE PRECISION SRNAME
DOUBLE PRECISION A, B, C, CX, G, SX, T2, T, TBPI, XBIG, XVSMAL, Y
DOUBLE PRECISION DCOS, DSIN, DSQRT, P, Q, Z
DATA SRNAME /8H BESHS0 /
DATA XBIG,XVSMAL,TBPI/3.4D+15,3.2D-9,6.36619772367581343D-1/
C
T = DABS(X)
C ERROR 1 TEST
IF (T.GT.XBIG) GO TO 60
IFAIL = 0
C X RANGE TEST
IF (T.GT.8.0D0) GO TO 40
C SMALL X
C TEST FOR VERY SMALL X
IF (T.GT.XVSMAL) GO TO 20
BESFJ0 = 1.0D0
GO TO 80
20 T = 3.125D-2*T*T - 1.0D0
T2 = 2.0D0*T
C
C
A= +4.125321000000D-14
B=T2*A -1.943834690000D-12
C=T2*B-A +7.848696314000D-11
A=T2*C-B -2.679253530560D-9
B=T2*A-C +7.608163592419D-8
C=T2*B-A -1.761946907762D-6
A=T2*C-B +3.246032882101D-5
B=T2*A-C -4.606261662063D-4
C=T2*B-A +4.819180069468D-3
A=T2*C-B -3.489376941141D-2
B=T2*A-C +1.580671023321D-1
C=T2*B-A -3.700949938726D-1
A=T2*C-B +2.651786132033D-1
B=T2*A-C -8.723442352852D-3
Y=T*B-A +1.577279714749D-1
C
BESHS0 = DCMPLX(Y,0.0D0)
GO TO 80
C
C LARGE X
40 G = T - 0.5D0/TBPI
Y = DSQRT(TBPI/T)
CX = DCOS(G)*Y
SX = DSIN(G)*Y
Z = 128.0D0/(T*T) - 1.0D0
C
C
P= +9.99457276D-1+Z*( -5.36366929D-4+Z*( +6.13732440D-6+
AZ*( -2.06823782D-7+Z*( +1.30451717D-8))))
C
C
Q= -1.55551139D-2+Z*( +6.83314931D-5+Z*( -1.47708215D-6+
AZ*( +7.10458046D-8+Z*( -5.81753275D-9+Z*( +6.75219505D-10)))))
C
Q = Q*8.0D0/T
C
BESHS0 = DCMPLX(P,Q)*DCMPLX(CX,SX)
GO TO 80
C
C ERROR 1 EXIT
60 WRITE(6,61) SRNAME
61 FORMAT(1X,//,'ERROR IN ',A8)
BESHS0 = DCMPLX(TBPI/T,0.0D0)
C
80 RETURN
END
最佳答案
您的代码不是正确的 Fortran，因此不是 Fortran 77。
最有可能的是正确的语法
COMPLEX*16 FUNCTION BESHS0(X, IFAIL)
这也不是标准的 Fortran 77，但应该被大多数编译器接受。
顺便说一句，Bessel 函数包含在 Fortran 2008 中(不确定是否是这个特定函数)。
关于compilation - gfortran错误: Expected formal argument list in function definition at (1)，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/38181537/