/***********************************************************/ /* Abstract: */ /* */ /* 'XC' is a cross-reference utility for 'C' programs. */ /* Its has the ability to handle nested include files */ /* to a depth of 8 levels and properly processes nested */ /* comments as supported by BDS C. Option flags support */ /* the following features: */ /* */ /* - Routing of list output to disk */ /* - Cross-referencing of reserved words */ /* - Processing of nested include files */ /* - Generation of listing only */ /* */ /* Usage: xc */ /* */ /* Flags: -i = Enable file inclusion */ /* -l = Generate listing only */ /* -c = Compressed print listing */ /* -r = Cross-ref reserved words */ /* -o = Write output to named file */ /* */ /*************************************************************************/ /* 7-11-85 reworked for CPM 2 David Giunti added conditional DOS to allow compilaton bye Aztec C defining DOS at compilation will give MS DOS code otherwise CPM code changed nl() routine by define to allow Aztec to send a cr after a lf , fputc becomes aputc changed prt_hdr() by define to omit time data for CPM added ffdr(), called by prt_hdr() to save a page on printouts added conditionals MX for Epson and IDS for that printers compressed headings, else the printers default is used */ /* 1-01-84 Enhancements S.R. Jacobson */ /* - changed -e option to -c option and added two strings: */ /* cprint_hdr sets the printer to compressed print mode */ /* cprint_trlr resets the printer to "normal" mode */ /* This makes it easier for non-epson printer users */ /* - removed all the BDS C comments (this makes it easier to modify */ /* in the future. */ /* - changed name of the utility to the "print" utility */ /* - if the -i flag is not used, the second line number will not */ /* appear. */ /* - if a form feed is detected in the source file, a new page is */ /* started. This is useful for structuring the source */ /* code by placing each major section on a separate page. */ /* You do this by embedding a form feed (control L) in a */ /* comment. The text following the form feed will be placed */ /* on the top of the new page. */ /* - tab characters in the file are expanded to every 8 positions. */ /* This may be changed by changing the TABCNT define. */ /* - added the date and time to the printed output. (C-86 only) */ /* - fixed error in hashing algorithm. */ /* - control characters (other than \n, \f, or \t) are stripped */ /* from the input stream (converted to blanks). This */ /* prevents extraneous control chars from interfering with */ /* operation of xc. */ /* - cleaned up the structuring of a number of routines. */ /* - added the file latest update date and time to the printed */ /* output. ***DOS 2.0 ONLY *** */ /*************************************************************************/ /* 9-26-83 Microsoft C 1.04 Conversion WHR */ /* - \t between line numbers and text to fix indenting problem. */ /* - added option -e for output to Epson in condensed print. */ /* - toupper() and isupper() are macros, not functions. */ /* - eliminate side effect in toupper(*++arg) in main(). */ /* - change alloc() to malloc(). */ /* - add #define NAMES that are not in stdio.h */ /* - MS-C requires () in statement A?(c=B):(c=C) error or not?? */ /* */ /* 4-30-83 Computer Innovations C-86 1.31 Conversion WHR */ /* - #include filename changed to allow a disk drive prefix, D: */ /* - convert if(fprintf(...) == ERROR) lst_err(); to fprintf(..); */ /* - convert if(fopen(...) == ERROR) statements to == NULL. */ /* - C86 requires () in statement A?(c=B):(c=C) error or not?? */ /* - remove getc() == ERROR check in fil_chr(). */ /* - convert file conventions from BDS C to C-86. */ /* - comment out BDS unique statements, mark revised statements. */ /* keep all BDS statements to document conversion effort. */ /* */ /** 4-19-83 BDS C Version file XC.CQ copied from Laurel RCPM WHR */ /*************************************************************************/ /* ** Original Version: ** */ /* */ /* Version 1.0 January, 1982 */ /* */ /* Copyright (c) 1982 by Philip N. Hisley */ /* */ /* Released for non-commercial distribution only */ /* */ /* Please report bugs/fixes/enhancements to: */ /* */ /* Philip N. Hisley */ /* 548H Jamestown Court */ /* Edgewood, Maryland 21040 */ /* (301) 679-4606 */ /* Net Addr: PNH@MIT-AI */ /* */ /* */ /***********************************************************/ /* DEFINITIONS AND VARIABLES */ /*SRJ*/ #include "stdio.h" /* WHR */ #ifndef DOS /* Copyright (C) 1984 by Manx Software Systems */ extern char ctp_[]; #define isalpha(x) (ctp_[(x)+1]&0x03) #define isupper(x) (ctp_[(x)+1]&0x01) #define islower(x) (ctp_[(x)+1]&0x02) #define isdigit(x) (ctp_[(x)+1]&0x04) #define iscntrl(x) (ctp_[(x)+1]&0x20) FILE *fopen() ; /* #include "ctype.h" */ /* include type tests */ /* #undef toupper */ /* undefine Aztecs toupper macro use librarys */ #endif /* ndef DOS */ #define TABCNT 8 /* expand tabs to every 8 characters */ /*SRJ*/ #define NULL 0 #define FALSE 0 /* WHR */ #define TRUE 1 /* WHR */ #define CPMEOF 0x1A /* end of file */ /* WHR */ #define ERROR (-1) /* WHR */ #define MAX_REF 5 /* maximum refs per ref-block */ #define MAX_LEN 20 /* maximum identifier length */ #define MAX_WRD 749 /* maximum number of identifiers */ #define MAX_ALPHA 53 /* maximum alpha chain heads */ #define REFS_PER_LINE 8 /* maximum refs per line */ #define LINES_PER_PAGE 60 struct id_blk { char id_name[MAX_LEN]; struct id_blk *alpha_lnk; struct rf_blk *top_lnk; struct rf_blk *lst_lnk; } oneid; struct rf_blk { int ref_item[MAX_REF]; int ref_cnt; } onerf; struct id_blk *id_vector[MAX_WRD]; struct alpha_hdr { struct id_blk *alpha_top; struct id_blk *alpha_lst; }; struct alpha_hdr alpha_vector[MAX_ALPHA]; struct regval {int ax,bx,cx,dx,si,di,ds,es;} reg; /* structure for sysint */ /*SRJ*/ struct regv {int csx,ssx,dsx,esx;} regx; /* structure for segread */ /*SRJ*/ int y,mo,d; /* year,month and day for date */ /*SRJ*/ int h,mi,s; /* hours, minutes, seconds for time */ /*SRJ*/ char *months[]= /* month names for date */ /*SRJ*/ {"bad month","January","February","March","April","May","June", "July","August","September","October","November","December"}; int ln_chr=0; /* # of characters in the line (for tab) */ /*SRJ*/ unsigned int handle; /* dos file handle */ /*SRJ*/ unsigned int file_d,file_mo,file_y,file_h,file_mi; /* file m/d/y and h:m */ /*SRJ*/ FILE *tb; /* temp fd */ /*SRJ*/ int linum; /* line number */ int edtnum; /* edit line number */ int fil_cnt; /* active file index */ int wrd_cnt; /* token count */ int pagno; /* page number */ int id_cnt; /* number of unique identifiers */ int rhsh_cnt; /* number of conflict hits */ int filevl; /* file level */ int paglin; /* page line counter */ int prt_ref; char act_fil[MAX_LEN]; char lst_fil[MAX_LEN]; char gbl_fil[MAX_LEN]; FILE *l_buffer; /* WHR */ int i_flg,o_flg,r_flg,l_flg; int debug; int c_flg; /* WHR */ /* The following two strings are the compressed print header and the trailer that returns the printer to the normal mode. These must be changed for each printer. They are currently set up for the IDS Microprism. ** SRJ ** */ #ifdef IDS char cprint_hdr[] = "\033R2$\036"; /* header string for Microprism */ char cprint_trlr[] = "\033R1$\035"; /* trailer string for Microprism */ #else /* Epson header string */ #ifdef MX char cprint_hdr[] = "\x1B\x40\x0F\x1BQ\x84"; /* WHR */ #else /* no printer , or later stack printer defined */ char cprint_hdr[] = "\0" ; char cprint_trlr[] = "\0"; /*DAG*/ #endif /* if def MX or not */ #endif /* IDS not defined */ /* end of printer strings */ /* MAIN PROGRAM */ /*SRJ*/ main(argc,argv) int argc; char **argv; {char *arg; char *strcpy(); /*** int toupper(); ***/ /* Microsoft C macro */ char cc; /* Microsoft C */ /* The the parameters */ if (argc < 2) use_err(); i_flg=r_flg=o_flg=l_flg=FALSE; debug = FALSE; strcpy(gbl_fil,*++argv); --argc; if(gbl_fil[0] == '-') use_err(); while(--argc != 0) { if(*(arg=*++argv) == '-') /*****{ switch( toupper(*++arg) ) *** side effect in Microsoft C */ { switch( cc=*++arg, toupper(cc) ) /* Microsoft C */ { case 'I': i_flg++; break; case 'R': r_flg++; break; case 'L': l_flg++; break; case 'O': {o_flg++; if(--argc == 0) use_err(); strcpy(lst_fil,*++argv); if(lst_fil[0] == '-') use_err(); if(debug) printf("lst_fil=>%s<",lst_fil); break; } case 'D': debug++; break; case 'C': c_flg++; /* SRJ */ /* WHR */ o_flg++; strcpy(lst_fil,"PRN:"); /* SRJ */ break; default: use_err(); } } else use_err(); } if(debug) printf("\ni_flg=%d, r_flg=%d, l_flg=%d", i_flg,r_flg,l_flg); if(debug) printf("\no_flg=%d, debug=%d", o_flg,debug); if (o_flg) {if ( (l_buffer = fopen(lst_fil,"w")) == NULL) /*** output file ***/ {printf("ERROR: Unable to create list file - %s\n",lst_fil); exit(0); } printf("\nXC ....... 'C' Print Utility v1.0\n"); /*SRJ*/ if (c_flg) fprintf(l_buffer,"%s",cprint_hdr); } else /*SRJ*/ l_buffer=stdout; /* if no output file, use stdout */ /*SRJ*/ /* THE FOLLOWING CODE GETS THE DATE AND THE TIME FROM DOS. IT USES THE ROUTINES SYSINT AND SEGREAD PROVIDED BY THE CI C-86 LIBRARY. THIS IS CI C-86 DEPENDENT CODE AND MUST BE CHANGED FOR OTHER COMPILERS */ /*SRJ*/ #ifdef DOS segread(®x); /* get the segment registers */ /*SRJ*/ reg.ds=regx.dsx; /* save the ds in the sysint data */ /*SRJ*/ reg.es=regx.esx; /* save the es in the sysint data */ /*SRJ*/ /* get the date */ /*SRJ*/ reg.ax=0x2A00; /* date request code */ /*SRJ*/ sysint(0x21,®,®); /* DOS interrupt */ /*SRJ*/ y=reg.cx; /* store year */ /*SRJ*/ mo=reg.dx>>8; /* store month */ /*SRJ*/ d=reg.dx & 0xFF; /* get the day */ /*SRJ*/ /* get the time */ /*SRJ*/ reg.ax=0x2C00; /* time request code */ /*SRJ*/ sysint(0x21,®,®); /* DOS interrupt */ /*SRJ*/ h=reg.cx>>8; /* get hours */ /*SRJ*/ mi=reg.cx & 0xFF; /* get minutes */ /*SRJ*/ s=reg.dx>>8; /* get seconds */ /*SRJ*/ #endif /* def DOS */ /* NOW WE GET THE FILE LATEST UPDATE DATE AND TIME */ /* first try to open the file */ if ((tb = fopen(gbl_fil,"r")) == NULL) /*open the input file */ /*SRJ*/ {printf("\nERROR: Unable to open input file: %s\n",gbl_fil); /*SRJ*/ exit(10); /*SRJ*/ } fclose(tb); /* close the file */ /*SRJ*/ /* note-we have no way to get an error indication back from the dos call below (file open), so we have to open it to see if it is there before we can use the proceedure below */ /*SRJ*/ /* now get the file date and time */ /*SRJ*/ #ifdef DOS /* open the file - we assume the file exists */ /*SRJ*/ reg.ax=0x3D00; /* open request code */ /*SRJ*/ reg.dx=gbl_fil; /* address of file name */ /*SRJ*/ sysint(0x21,®,®); /* DOS interrupt */ /*SRJ*/ handle=reg.ax; /* store file handle */ /*SRJ*/ /* get the date and time */ /*SRJ*/ reg.ax=0x5700; /* time request code */ /*SRJ*/ reg.bx=handle; /* file handle in bx */ /*SRJ*/ sysint(0x21,®,®); /* DOS interrupt */ /*SRJ*/ /* unpack the file time */ /*SRJ*/ file_h=(reg.cx >> 11) & 0x1F; /* get hours */ /*SRJ*/ file_mi=(reg.cx >> 5) & 0x3f; /* get minutes */ /*SRJ*/ /* unpack the file date */ /*SRJ*/ file_d=reg.dx & 0x1F; /* get the day */ /*SRJ*/ file_mo= (reg.dx >> 5) & 0x0F; /* get the month */ /*SRJ*/ file_y=( (reg.dx >> 9) & 0x7F )+80; /* get the month */ /*SRJ*/ /* close the file */ /*SRJ*/ reg.bx=handle; /*SRJ*/ reg.ax=0x3D00; /*SRJ*/ sysint(0x21,®,®); /* DOS interrupt */ /*SRJ*/ #endif /* def DOS */ /* END OF CI C-86 DEPENDENT CODE */ /*SRJ*/ prt_ref = FALSE; for(linum=0;linum < MAX_WRD;linum++) id_vector[linum] = NULL; for(linum=0;linum < MAX_ALPHA;linum++) alpha_vector[linum].alpha_top = alpha_vector[linum].alpha_lst = NULL; fil_cnt = wrd_cnt = linum = 0; filevl=paglin=pagno=edtnum=0; id_cnt=rhsh_cnt=0; proc_file(gbl_fil); if(!l_flg) {prnt_tbl(); printf("\nAllowable Symbols: %d\n",MAX_WRD); printf("Unique Symbols: %d\n",id_cnt); } if(o_flg) {nl(); fprintf(l_buffer,"\nAllowable Symbols: %d\n",MAX_WRD); /* WHR */ fprintf(l_buffer,"Unique Symbols: %d\n",id_cnt); /* WHR */ fprintf(l_buffer,"\f%c",CPMEOF); /* WHR */ fprintf(l_buffer,"%s",cprint_trlr); /* output trailer */ /* SRJ */ fflush(l_buffer); fclose(l_buffer); } } /* ERROR MESSAGE PRINT ROUTINES */ /*SRJ*/ lst_err() {printf("\nERROR: Write error on list output file - %s\n",lst_fil); exit(0); } use_err() {printf("\nERROR: Invalid parameter specification\n\n"); printf("Usage: xc \n\n"); printf("Flags: -i = Enable file inclusion\n"); printf(" -l = Generate listing only\n"); printf(" -c = Compressed print option\n"); /*SRJ*/ printf(" -r = Cross-reference reserved words\n"); printf(" -o = Write output to named file\n"); exit(0); } /* PROCESS THE FILE */ /*SRJ*/ proc_file(filnam) char *filnam; {FILE *buffer; /* allocated buffer pointer */ /* WHR */ char token[MAX_LEN]; /* token buffer */ int eof_flg; /* end-of-file indicator */ int tok_len; /* token length */ int incnum; /* included line number */ char *strcpy(); strcpy(act_fil,filnam); if ((buffer = fopen(filnam,"r")) == NULL) /*** input file ***/ /* WHR */ {printf("\nERROR: Unable to open input file: %s\n",filnam); exit(0); } if(filevl++ == 0) {prt_hdr(); nl(); } eof_flg = FALSE; do {if(get_token(buffer,token,&tok_len,&eof_flg,0)) {if (debug) printf("token: %s length: %d\n",token,tok_len); if (chk_token(token)) /* #include processing changed to accept drive: WHR */ {if (strcmp(token,"#include") == 0) {if (get_token(buffer,token,&tok_len,&eof_flg,1)) {if (debug) printf("**token: %s length: %d\n", token,tok_len); if (!i_flg) continue; else {incnum=edtnum; edtnum=0; nl(); proc_file(token); edtnum=incnum; strcpy(act_fil,filnam); continue; } } } put_token(token,linum); } } } while (!eof_flg); filevl -= 1; fclose(buffer); return; } /* GET A TOKEN */ /*SRJ*/ get_token(g_buffer,g_token,g_toklen,g_eoflg,g_flg) FILE *g_buffer; char *g_token; int *g_toklen; int *g_eoflg; int g_flg; /* * 'getoken' returns the next valid identifier or * reserved word from a given file along with the * character length of the token and an end-of-file * indicator * */ { int c; char *h_token; char tmpchr; char tmpchr2; /* WHR fix for B:filename.ext */ h_token = g_token; gtk: /* top of loop, get new token */ *g_toklen = 0; g_token = h_token; /* * Scan and discard any characters until an alphabetic or * '_' (underscore) character is encountered or an end-of-file * condition occurs */ while( (!isalpha(*g_token = rdchr(g_buffer,g_eoflg,g_flg))) && !*g_eoflg && *g_token != '_' && *g_token != '0' && *g_token != '#' ); if(*g_eoflg) return(FALSE); *g_toklen += 1; /* * Scan and collect identified alpanumeric token until * a non-alphanumeric character is encountered or and * end-of-file condition occurs */ if(g_flg) { tmpchr = '.'; tmpchr2 = ':'; /* WHR fix for B:filename.ext */ } else { tmpchr = '_'; tmpchr2 = '_'; /* WHR fix for B:filename.ext */ } while( (isalpha(c=rdchr(g_buffer,g_eoflg,g_flg)) || isdigit(c) || c == '_' || c == tmpchr || c == tmpchr2) && !*g_eoflg) /* WHR fix for B:filename.ext */ { if(*g_toklen < MAX_LEN) { *++g_token = c; *g_toklen += 1; } } /* * Check to see if a numeric hex or octal constant has * been encountered ... if so dump it and try again */ if (*h_token == '0') goto gtk; /* * Tack a NULL character onto the end of the token */ *++g_token = NULL; /* * Screen out all #token strings except #include */ if (*h_token == '#' && strcmp(h_token,"#include")) goto gtk; return (TRUE); } /* READ A CHARACTER FROM THE FILE AND PROCESS IT */ /*SRJ*/ rdchr(r_buffer,r_eoflg,rd_flg) int *r_eoflg; FILE *r_buffer; int rd_flg; /* 'rdchr' returns the next valid character in a file and an end-of-file indicator. A valid character is defined as any which does not appear in either a commented or a quoted string ... 'rdchr' will correctly handle comment tokens which appear within a quoted string */ { int c; int q_flg; /* double quoted string flag */ int q1_flg; /* single quoted string flag */ int cs_flg; /* comment start flag */ int ce_flg; /* comment end flag */ int c_cnt; /* comment nesting level */ int t_flg; /* transparency flag */ q_flg = FALSE; q1_flg = FALSE; cs_flg = FALSE; ce_flg = FALSE; t_flg = FALSE; c_cnt = 0; rch: /* * Fetch character from file */ c = fil_chr(r_buffer,r_eoflg) & 0x7F; /* read only 7 bit ascii */ /*SRJ*/ if (*r_eoflg) return(c); /* EOF encountered */ switch(c) /* test the returned character */ /*SRJ*/ {case '\n': /* carriage return */ /*SRJ*/ nl(); /* output new line and line numbers */ /*SRJ*/ ln_chr=0; /* set to first char in line */ /*SRJ*/ c=' '; /* return a space */ /*SRJ*/ break; case '\f': /* form feed */ /*SRJ*/ prt_hdr(); /* output header and start new page*/ /*SRJ*/ fprintf(l_buffer," "); /* space over line #s */ /*SRJ*/ ln_chr=0; /* set to first char in line */ /*SRJ*/ c=' '; /* set char to blank */ /*SRJ*/ break; case '\t': /* tab character */ /*SRJ*/ do /* expand the tabs */ /*SRJ*/ fputc(' ',l_buffer); /*SRJ*/ while( (++ln_chr % TABCNT) != 0); /*SRJ*/ c=' '; /* set char to blank */ /*SRJ*/ break; default: ln_chr++; /* move to next char pos */ /*SRJ*/ if( iscntrl(c) ) /* insure no control chars */ /*SRJ*/ c=' '; fputc(c,l_buffer); /* and output next char */ /*SRJ*/ } if (rd_flg) return(c); if (t_flg) { t_flg = !t_flg; goto rch; } if (c == '\\') { t_flg = TRUE; goto rch; } /* If the character is not part of a quoted string check for and process commented strings... nested comments are handled correctly but unbalanced comments are not ... the assumption is made that the syntax of the program being xref'd is correct */ if (!q_flg && !q1_flg) { if (c == '*' && c_cnt && !cs_flg) { ce_flg = TRUE; goto rch; } if (c == '/' && ce_flg) { c_cnt -= 1; ce_flg = FALSE; goto rch; } ce_flg = FALSE; if (c == '/') { cs_flg = TRUE; goto rch; } if (c == '*' && cs_flg) { c_cnt += 1; cs_flg = FALSE; goto rch; } cs_flg = FALSE; if (c_cnt) goto rch; } /* Check for and process quoted strings */ if ( c == '"' && !q1_flg) { /* toggle quote flag */ q_flg = !q_flg; if(debug) printf("q_flg toggled to: %d\n" ,q_flg); goto rch; } if (q_flg) goto rch; if (c == '\'') { /* toggle quote flag */ q1_flg = !q1_flg; if(debug) printf("q1_flg toggled to: %d\n" ,q1_flg); goto rch; } if (q1_flg) goto rch; /* Valid character ... return to caller */ return (c); }/*rdchr. -----------------------------------------------*/ fil_chr(f_buffer,f_eof) FILE *f_buffer; int *f_eof; { int fc; fc=getc(f_buffer); if (fc == CPMEOF || fc == EOF) { *f_eof = TRUE; fc = NULL; } return(fc); } /* CHECK THE TOKEN */ /*SRJ*/ chk_token(c_token) char *c_token; { char u_token[MAX_LEN]; int i; { if (r_flg) return(TRUE); i = 0; do { u_token[i] = toupper(c_token[i]); } while (c_token[i++] != NULL); switch(u_token[0]) { case 'A': if (strcmp(u_token,"AUTO") == 0) return(FALSE); break; case 'B': if (strcmp(u_token,"BREAK") == 0) return(FALSE); break; case 'C': if (strcmp(u_token,"CHAR") == 0) return (FALSE); if (strcmp(u_token,"CONTINUE") == 0) return (FALSE); if (strcmp(u_token,"CASE") == 0) return (FALSE); break; case 'D': if(strcmp(u_token,"DOUBLE") == 0) return(FALSE); if(strcmp(u_token,"DO") == 0) return(FALSE); if(strcmp(u_token,"DEFAULT") == 0) return(FALSE); break; case 'E': if(strcmp(u_token,"EXTERN") == 0) return(FALSE); if(strcmp(u_token,"ELSE") == 0) return(FALSE); if(strcmp(u_token,"ENTRY") == 0) return(FALSE); break; case 'F': if(strcmp(u_token,"FLOAT") == 0) return(FALSE); if(strcmp(u_token,"FOR") == 0) return(FALSE); break; case 'G': if(strcmp(u_token,"GOTO") == 0) return(FALSE); break; case 'I': if(strcmp(u_token,"INT") == 0) return(FALSE); if(strcmp(u_token,"IF") == 0) return(FALSE); break; case 'L': if(strcmp(u_token,"LONG") == 0) return(FALSE); break; case 'R': if(strcmp(u_token,"RETURN") == 0) return(FALSE); if(strcmp(u_token,"REGISTER") == 0) return(FALSE); break; case 'S': if(strcmp(u_token,"STRUCT") == 0) return(FALSE); if(strcmp(u_token,"SHORT") == 0) return(FALSE); if(strcmp(u_token,"STATIC") == 0) return(FALSE); if(strcmp(u_token,"SIZEOF") == 0) return(FALSE); if(strcmp(u_token,"SWITCH") == 0) return(FALSE); break; case 'T': if(strcmp(u_token,"TYPEDEF") == 0) return(FALSE); break; case 'U': if(strcmp(u_token,"UNION") == 0) return(FALSE); if(strcmp(u_token,"UNSIGNED") == 0) return(FALSE); break; case 'W': if(strcmp(u_token,"WHILE") == 0) return(FALSE); break; } } return (TRUE); }/*chk_token. ---------------------------------------------*/ /* STORE THE TOKEN */ /*SRJ*/ /* * Install parsed token and line reference in linked structure */ put_token(p_token,p_ref) char *p_token; int p_ref; { int hsh_index; int i; unsigned long j; /*SRJ*/ int d; int found; struct id_blk *idptr; struct rf_blk *rfptr; struct id_blk *alloc_id(); struct rf_blk *alloc_rf(); struct rf_blk *add_rf(); if (l_flg) return; j=0; for (i=0; p_token[i] != NULL; i++) /* Hashing algorithm is far from */ j = j * 10 + p_token[i]; /* memory-bound index vector! */ hsh_index = j % MAX_WRD; /*SRJ*/ if(debug) printf("hash index=%d ",hsh_index); found = FALSE; d = 1; do {idptr = id_vector[hsh_index]; if (idptr == NULL) {id_cnt++; idptr = id_vector[hsh_index] = alloc_id(p_token); chain_alpha(idptr,p_token); idptr->top_lnk = idptr->lst_lnk = alloc_rf(p_ref); found = TRUE; if(debug) printf("empty cell in vector\n"); } else if (strcmp(p_token,idptr->id_name) == 0) {idptr->lst_lnk = add_rf(idptr->lst_lnk,p_ref); if(debug) printf("duplicate token\n"); found = TRUE; } else {hsh_index += d; if(debug) printf("hash clash hash index=%d\n",hsh_index); d += 2; rhsh_cnt++; if (hsh_index >= MAX_WRD) hsh_index -= MAX_WRD; if (d == MAX_WRD) {printf("\nERROR: Symbol table overflow\n"); exit(0); } } } while (!found); return; }/*put_token. --------------------------------------------*/ /* BUILD THE CHAINS */ /*SRJ*/ chain_alpha(ca_ptr,ca_token) struct id_blk *ca_ptr; char *ca_token; { char c; int f; struct id_blk *cur_ptr; struct id_blk *lst_ptr; /*** int isupper(); ****/ /** Microsoft C macro **/ c = ca_token[0]; if (c == '_') c = 0; else /** isupper(c) ? c=1+((c-'A')*2) : c=2+((c-'a')*2) ; error or not?? ** A good one for the puzzle book! Is the () required around (c=..)? ** C86 and Microsoft C both req the ()'s, BDS C did not. ** Is it required because = has lower precedence than ?: ???? **/ isupper(c) ? (c=1+((c-'A')*2)) : (c=2+((c-'a')*2)) ; if(alpha_vector[c].alpha_top == NULL) { alpha_vector[c].alpha_top = alpha_vector[c].alpha_lst = ca_ptr; ca_ptr->alpha_lnk = NULL; return; } /* check to see if new id_blk should be inserted between * the alpha_vector header block and the first id_blk in * the current alpha chain */ if(strcmp(alpha_vector[c].alpha_top->id_name,ca_token) >0) { ca_ptr->alpha_lnk=alpha_vector[c].alpha_top; alpha_vector[c].alpha_top=ca_ptr; return; } if(strcmp(alpha_vector[c].alpha_lst->id_name,ca_token) < 0) { alpha_vector[c].alpha_lst->alpha_lnk = ca_ptr; ca_ptr->alpha_lnk = NULL; alpha_vector[c].alpha_lst=ca_ptr; return; } cur_ptr = alpha_vector[c].alpha_top; while(strcmp(cur_ptr->id_name,ca_token) < 0) { lst_ptr = cur_ptr; cur_ptr = lst_ptr->alpha_lnk; } lst_ptr->alpha_lnk = ca_ptr; ca_ptr->alpha_lnk = cur_ptr; return; }/*chain_alpha. -----------------------------------------*/ /* GET MEMORY FOR TOKEN TREE */ /*SRJ*/ struct id_blk *alloc_id(aid_token) char *aid_token; { int ai; struct id_blk *aid_ptr; char *malloc(); /* Microsoft C */ /*** if((aid_ptr = alloc(sizeof(oneid))) == 0) { ***/ /* Microsoft C */ if((aid_ptr = (struct id_blk *) malloc(sizeof(oneid))) == 0) { /* MS C */ printf("\nERROR: Unable to allocate identifier block\n"); exit(0); } ai=0; do { aid_ptr->id_name[ai] = aid_token[ai]; } while (aid_token[ai++] != NULL); return (aid_ptr); }/*id_blk. -----------------------------------------*/ struct rf_blk *alloc_rf(arf_ref) int arf_ref; { int ri; struct rf_blk *arf_ptr; char * malloc(); /* Microsoft C */ /** if((arf_ptr = alloc(sizeof(onerf))) == 0) { **/ /* Microsoft C */ if((arf_ptr = (struct rf_blk *) malloc(sizeof(onerf))) == 0) { /* MS C */ printf("\nERROR: Unable to allocate reference block\n"); exit(0); } arf_ptr->ref_item[0] = arf_ref; arf_ptr->ref_cnt = 1; for (ri=1;riref_item[ri] = NULL; return (arf_ptr); }/*alloc_rf. ------------------------------------------*/ /* STORE REFERENCE IN TREE */ /*SRJ*/ struct rf_blk *add_rf(adr_ptr,adr_ref) struct rf_blk *adr_ptr; int adr_ref; { struct rf_blk *tmp_ptr; tmp_ptr = adr_ptr; if(adr_ptr->ref_cnt == MAX_REF) { /*** tmp_ptr = adr_ptr->ref_cnt = alloc_rf(adr_ref); Microsoft C **/ adr_ptr->ref_cnt = alloc_rf(adr_ref); /* Microsoft C **/ tmp_ptr = (struct rf_blk *) adr_ptr->ref_cnt; /* Microsoft C **/ } else { adr_ptr->ref_item[adr_ptr->ref_cnt++] = adr_ref; } return (tmp_ptr); }/*rf_blk. ------------------------------------------*/ /* PRINT THE CROSS REFERENCE TABLE */ /*SRJ*/ prnt_tbl() { int prf_cnt; int pti; int pref; int lin_cnt; struct id_blk *pid_ptr; struct rf_blk *ptb_ptr; prt_ref = TRUE; prt_hdr(); nl(); for (pti=0;ptiid_name);/*WHR*/ ptb_ptr=pid_ptr->top_lnk; lin_cnt=prf_cnt=0; do {if(prf_cnt == MAX_REF) {prf_cnt=0; /*** ptb_ptr = ptb_ptr->ref_cnt; Microsoft C **/ ptb_ptr = (struct rf_blk *)ptb_ptr->ref_cnt; } if(ptb_ptr > MAX_REF) {if((pref=ptb_ptr->ref_item[prf_cnt++]) != 0) {fprintf(l_buffer,"%-4d ",pref); /*WHR*/ if (++lin_cnt == REFS_PER_LINE) {nl(); fprintf(l_buffer,"%16s",": "); lin_cnt=0; } } } else pref=0; } while (pref); nl(); } while ((pid_ptr=pid_ptr->alpha_lnk) != NULL); } } }/*prnt_tbl. ---------------------------------------*/ /* TOP OF PAGE AND PRINT THE HEADER */ /*SRJ*/ prt_hdr() { #ifdef DOS ffdr() ; /* do a formfeed if necessary */ fprintf(l_buffer, "\r%-10s [file d/t= %d/%02d/%02d %d:%02d] [clock d/t= %d %s %d %d:%02d:%02d] Page %d\n\n", gbl_fil,file_mo,file_d,file_y,file_h,file_mi, d,months[mo],y,h,mi,s,++pagno); /*SRJ*/ #else /* it's cpm or has no clocked files */ ffder(); /* print a form feed if necessary */ fprintf(l_buffer, "\r%-10s Page %d\n\n",gbl_fil,++pagno); /*DAG*/ #endif /* defed DOS or not */ paglin =3; return; } /* FORMFEED to save a page */ ffder() /* print a form feed if necessary */ { if (pagno) /* isn't the first page */ fprintf(l_buffer,"\r\f\n"); } /* NEW LINE */ /*SRJ*/ nl() { #ifdef DOS fputc('\n',l_buffer); /* output cr/lf */ /*SRJ*/ #else aputc('\n',l_buffer); /* output cr/lf */ /*SRJ*/ #endif /* def DOS or not */ if (++paglin >= LINES_PER_PAGE) prt_hdr(); if(!prt_ref) {fprintf(l_buffer,"%4d ",++linum); /*SRJ*/ if(i_flg) /*SRJ*/ fprintf(l_buffer," %4d: ",++edtnum); /*SRJ*//*WHR*/ } /* print the status on the console */ if(o_flg) if(linum % 60 == 1) printf("\n<%d>\t",linum); else printf("."); return; }/*nl. -------------------------------------------*/ /*============= end of file xc.c ==========================*/