CD-Section: Parse 'KDM' Keyworded Data C A routine is available to examine input lines for the presece of KDM C data; if such data are found, the data will be checked for validity and C stored in the Keyword Data Module. Any non-KDM data lines will be returned C to the calling program for interpretation by that program and thus both KDM C and other control or parameters data may be included in the file being C read. Another routine enables additional parameter checking to be C carried out. CD-end C LAST UPDATE: C+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ C C************************** C** KWD_PARSEKDM ** C************************** C CD-Routine: Parse Line Containing KDM Data - KWD_PARSEKDM C This routine parses a given line of data, examining it for the presence C of a requested set of KDM data. Any such data found is automatically C checked and, if valid, stored in the appropriate Keyword Data Module. C If the line does not contain KDM data from the requested set, C then it is merely passed back to the calling program to allow further C interpretation by that program. KDM and non-KDM data (or data items C from different KDM sets) must not be given C in the same line of data and a line of data containing any valid KDM C keyword for the KDM set being examined will be assumed to be a KDM C parameter line. The calling program C should also take account of the fact that incorrectly specified KDM C keyords may result in the line being returned to the calling program. CD-end CD-Fortran: SUBROUTINE KWD_PARSEKDM (KDX, LINE, IFLAG, BADTOK, ERRSTR) CD-end INCLUDE 'implicit_none.finc' C C Purpose: Parse line for KDM data and store any found. C C Author: John W. Campbell, June 1995 C C Arguments: C INTEGER KDX CHARACTER*(*) LINE INTEGER IFLAG CHARACTER*(*) BADTOK CHARACTER*(*) ERRSTR CD-Parameters: C KDX i (R) Index to required KDM data set. C LINE c (R) Character string containing line to be parsed C (Assume comments & continuations already removed) C IFLAG i (W) Return flag = 0, KDM data found & OK C = 1, Not KDM data of requested set C = 2, No non-blank tokens C < 0, KDM data but error(s) C =-1, Invalid or inappropriate set C specification C =-2, Invalid or inappropriate subset C specification C =-3, Syntax error in set/subset C specification C =-4, Invalid value C =-5, Non-KDM data mixed with KDM data C or mixed KDM datasets. C (Possibly miss-spelt keyword) C -100, Coding error in KWD routines, C information supposedly already C checked is found later to be invalid! C BADTOK c (W) Returns the token string for an invalid token C ERRSTR c (W) Error string (max 80 chars) if IFLAG<0 CD-end C C====== Parameter statements C C C====== GLOBALS: C C C====== FUNCTION TYPES: C INTEGER NXTTOK INTEGER LENSTR INTEGER KDMSTR C C====== LOCALS: C INTEGER I INTEGER LL INTEGER IL INTEGER M1 INTEGER M2 INTEGER II INTEGER MDUM INTEGER ITYP INTEGER IERR INTEGER ISET INTEGER NVALS INTEGER ISUBSET INTEGER ISS INTEGER IPAR INTEGER IVAL(2) REAL FVAL(2) CHARACTER*50 KEYWORD CHARACTER*50 FULLNAME CHARACTER*4 SEP C INTEGER I !Return flag from NXTTOK routine C INTEGER LL !Length of input line C INTEGER IL !Last character accessed in line being parsed C INTEGER M1 !First character in token field C INTEGER M2 !Final character in token field C INTEGER II !Temp/loop variable C INTEGER MDUM !Dummy character position index C INTEGER ITYP !KDM parameter type flag from KDMF_CHECKPAR C INTEGER IERR !Error flag C INTEGER ISET !Set number decoded from KDM keyword (0=all sets) C INTEGER NVALS !No. of integer/float values C INTEGER ISUBSET !Subset number decoded from KDM keyword C (0=all subsets) C INTEGER ISS !Dataset/set/subset specific parameter flag 0/1/2 C INTEGER IPAR !Parameter no. in KDM data set C INTEGER IVAL(2) !Dummy integer value array C REAL FVAL(2) !Dummy real value array C CHARACTER*50 KEYWORD !Keyword string C CHARACTER*50 FULLNAME!Full parameter name C CHARACTER*4 SEP !Token separator characters DATA SEP /' ,='/ C C------------------------------------------------------------------------------- C IFLAG = 2 SEP(2:2) = CHAR(9) BADTOK = ' ' IL = 0 LL = LEN(LINE) C C====== Get next token C 100 I = NXTTOK (LINE, SEP, IL, M1, M2) IF (I.EQ.-1) RETURN IF (I.EQ.0) GO TO 100 C C====== See if KDM name C CALL KWD_PARSEITEM (LINE(M1:M2), KEYWORD, ISET, ISUBSET, ISS, + IERR) IF (IERR.EQ.0) CALL KDMF_CHECKPAR (KDX, KDMSTR(KEYWORD), + LENSTR(KEYWORD), + ISET, ISUBSET, ISS, + KDMSTR(FULLNAME), 50, + ITYP, NVALS, IPAR, IERR) IF (IERR.LE.0) THEN IFLAG = 0 IF (IERR.LT.0) THEN BADTOK = LINE(M1:M2) IFLAG = IERR IF (IERR.EQ.-1) + ERRSTR = 'Invalid or inappropriate set specification' IF (IERR.EQ.-2) + ERRSTR = 'Invalid or inappropriate subset specification' IF (IERR.EQ.-3) + ERRSTR = 'Syntax error in set/subset specification' RETURN END IF I = NXTTOK (LINE, SEP, IL, M1, M2) IF (I.LE.0) THEN CALL KDMF_RESET (KDX, KDMSTR(FULLNAME), LENSTR(FULLNAME), + ISET, ISUBSET, KDMSTR(ERRSTR), LEN(ERRSTR), + IERR) IF (IERR.LT.0) THEN IFLAG = IERR C C====== Note: IERR=-3 should not occur as keyword already checked C IF (IERR.EQ.-3) IFLAG = -100 RETURN END IF IF (I.LT.0) RETURN ELSE IF (ITYP.GT.9) THEN M2 = LL IL = LL END IF IF (ITYP.LT.7.AND.NVALS.GT.1) THEN DO 120 II = 2, NVALS I = NXTTOK (LINE, SEP, IL, MDUM, M2) IF (I.LE.0) THEN ERRSTR = 'Too few values given for the parameter' IFLAG = -4 RETURN END IF 120 CONTINUE END IF CALL KDMF_SETVALUE (KDX, KDMSTR(FULLNAME), LENSTR(FULLNAME), + ISET, ISUBSET, KDMSTR(LINE(M1:M2)), + M2-M1+1, IVAL, FVAL, 0, + KDMSTR(ERRSTR), LEN(ERRSTR), IERR) IF (IERR.GT.0) THEN BADTOK = LINE(M1:M2) IFLAG = - 4 RETURN END IF END IF ELSE C C====== Non-KDM keyword found C IF (IFLAG.EQ.0) THEN BADTOK = LINE(M1:M2) IFLAG = -5 ERRSTR = 'Mixed KDM and non-KDM data' RETURN ELSE IFLAG = 1 RETURN END IF END IF GO TO 100 END