Relay-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site seismo.CSS.GOV Posting-Version: version B 2.10.2 9/3/84; site panda.UUCP Path: seismo!harvard!talcott!panda!sources-request From: sources-request@panda.UUCP Newsgroups: mod.sources Subject: help: VMS-style help facility Message-ID: <1046@panda.UUCP> Date: 30 Oct 85 13:34:26 GMT Sender: jpn@panda.UUCP Lines: 1036 Approved: jpn@panda.UUCP Mod.sources: Volume 3, Issue 29 Submitted by: seismo!wucs!nz (Neal Ziring) Here is our VMS-style help utility. ...nz (Neal Ziring @ the American Midwest) ----------------------------------------------------- #! /bin/sh # This is a shell archive, meaning: # 1. Remove everything above the #! /bin/sh line. # 2. Save the resulting text in a file. # 3. Execute the file with /bin/sh (not csh) to create the files: # help/Makefile # help/README # help/help.1 # help/help.h # help/help.c # This archive created: Mon Oct 28 10:21:23 1985 if test -d 'help' then echo shar: directory "'help'" already exists else mkdir 'help' fi # end of directory check export PATH; PATH=/bin:$PATH if test -f 'help/Makefile' then echo shar: will not over-write existing file "'help/Makefile'" else cat << \SHAR_EOF > 'help/Makefile' # # help Makefile, 1.1 7/27/84 # CFLAGS= -g DESTDIR= help: help.o ${CC} ${CFLAGS} help.o -o help install: help install -s -c help $(DESTDIR)/usr/local/bin help.o: help.h /usr/include/sys/file.h /usr/include/sys/param.h clean: rm -f *.o help SHAR_EOF fi # end of overwriting check if test -f 'help/README' then echo shar: will not over-write existing file "'help/README'" else cat << \SHAR_EOF > 'help/README' Help is a facility for climbing a tree structure. Help comes in two types: a help topic with subtopics, or a help topic without any subtopics. Topics that have no subtopics reside in the directory they are a subtopic of, with filenames `topicname.HLP' Topics that have subtopics of their own reside in directories which are subdirectories of the help topic of which they are subtopics, and have filenames `.HLP' Thier subtopics have the filenames `topicname.HLP'. A directory for a subtopic just has the subtopic's name as its directory name. For instance, if you had a help topic "ls" with some subtopics, /usr/help/ls would be a directory, containing the files /usr/help/ls/.HLP main help text /usr/help/ls/.MANUAL shell script for accessing man page /usr/help/ls/options.HLP help for ls options /usr/help/ls/output.HLP help for ls output format If you had a simple help for the who command, "who", it would comprise the following files at the top level: /usr/help/who.HLP help text for who(1) /usr/help/who.MANUAL script for accessing who(1) man page To make help, cd to the directory, and type: % make help or % make install if you are ambitious. The source for help is one C module, help.c, with a header file, help.h. Help uses the more(1) program to read help texts. Make sure that more resides on the path specified by the VIEWPROGRAM defined symbol in help.h. SHAR_EOF fi # end of overwriting check if test -f 'help/help.1' then echo shar: will not over-write existing file "'help/help.1'" else cat << \SHAR_EOF > 'help/help.1' .TH HELP 1 "29 November 1984" .UC 4 .SH NAME help \- user-friendly documentation reader .SH SYNOPSIS .B help [ -dlqCn ] [ topic-path ] .br .SH DESCRIPTION .I Help is a documentation reader based closely on the VMS(tm) help facility. Documentation texts are arranged in a tree, and the users reads them with .I more(1). Help allows documentation to be treated as an n-tree, the structure of the help tree is the structure of the directory tree in which the help files reside. .PP The help program is invoked from the shell like .PP help [ -dlqCn ] topic sub-topic sub-sub-topic ... .PP .I Topic may be any top-level help file name. A .I topic-path is a path of topics and their subtopics, down into the tree. Topic and subtopic names may be abbreviated, but no wild-cards are recognized. If any switches are present they all must follow a single dash, and be the first option on the command line. .PP The options are: .TP 5 .B d Next argument is the help root directory. Default is /usr/help. .TP 5 .B l Use list-only mode. Only the list of subtopics for the command-line help path is printed. Interactive mode is not used. .TP 5 .B q Use text-only mode. Only the text of the help path is printed. Interactive mode is not used. .TP 5 .B C Forces multi-column output of subtopic lists. This is the default for interactive mode. .TP 5 .B n Force one-topic-per-line output of the subtopics lists. This is the default for list mode. .PP The help interactive mode asks the user what help he would like next. The prompt looks like this: .PP Topic? .PP or like this: .PP vi commands deletion subtopic? .PP Where "vi commands deletion" is the .I "help tree path." This path is printed for each prompt, and as a header for each help text. .PP The following commands are meaningful in to the "topic?" prompt: .TP 5 .B ? Print current help and subtopic-list again. .TP 5 .B "." Do a UNIX Programmer's Manual lookup (if possible). If other characters follow the dot, they are taken as the filename for the .MANUAL file. (e.g. command '.who' looks for file who.MANUAL) .TP 5 .B "Return " Exit this level of help. If at top level, goes back to calling program. .TP 5 .B "" Read the documentation for this subtopic, or all subtopics for which this is a legal abbreviation. Wildcards are not permitted, but a short abbreviation matches everything longer. .TP 5 .B "#" List the available subtopics again. This lists the subtopics of this node, and re-prompts the user. The help text is NOT re-printed. .TP 5 .B "$" List some information on the files that make up this topic. Files that match the abbreviation of `topicname' are looked up with .I file(1). .sp 1 .PP .B "HELP FILES" .PP Help files are the text and directories of the .I help documentation tree. There are four kinds of help files: topic directory, help file, subtopic file, and manual file. .PP A topic directory is a directory with the name of the subtopic it describes (e.g. /usr/help/ls/options is a topic directory about the topic of options for .I ls(1)). .PP A help file is a file of formatted text with the name ".HLP". It resides in the topic directory for which it is the description text. .PP A subtopic file has the name "subtopic-name.HLP". It contains the text for a subtopic which does not itself have any subtopics. .PP A manual file contains the commands to be given to .I sh(1) when the user gives the lookup manual command (.). This file has name .MANUAL. Therefore the manual command script for the current node is just ".MANUAL". .br A manual file resides in the same topic directory as the help text whose corresponding manual it accesses. The manual may be a shell script or a binary file, but it must be marked executable by all users. .sp 1 .PP The usual way of calling help is to just type .sp 1 % .I help .PP in the shell. This brings up help interactive mode at the top level of topics. The user can always abort with ctrl/D. .SH AUTHOR Neal Ziring with George Robbert, Washington University in St. Louis. .SH FILES /usr/local/bin/help .br /usr/help* .br /usr/ucb/more .br /bin/sh .SH "SEE ALSO" more(1), man(1), sh(1), system(2), access(2) .SH DIAGNOSTICS Error messages are printed whenever a request is made to get documentation that does not exists, or is not accessible. .PP Signals are not trapped. .SH BUGS The file naming conventions are somewhat strict. .PP No wildcards allowed in topic names. (Could have made them reg. exp. but that would be confusing to novices.) .PP Terminal type is not checked, the display is assumed to be at least 76 columns wide. .PP There is no way for the user to specify another program for viewing files ("more -d" is always used.) This is coded into the header file. .PP The algorithm used for selecting amoung ambiguous help paths is non-obvious, and can be confusing when too-short abbreviations are used. SHAR_EOF fi # end of overwriting check if test -f 'help/help.h' then echo shar: will not over-write existing file "'help/help.h'" else cat << \SHAR_EOF > 'help/help.h' /* * help.h: header file for the VMS-help emulator */ #include #include #include #include #include #define iswhite(_c) (_c ==' ' || _c ==' ' || _c =='\n' || _c =='\r') #define PROMPT " Topic? " #define SUBPROMPT "subtopic? " #define HELPEX ".HLP" /* help extension */ #define MANEX ".MANUAL" /* manual filename*/ #define MAN_SUBEX "/.MANUAL" /* other man filename */ #define INFOHEAD "file " #define INFOTAIL " | sed 's/^/ /'" #define MAXLINELEN 128 #define MAXNAMELEN 20 /* max length of NAME of topix */ #define MAXNAMES 256 /* maximum number of subtopics */ #define COLUMNSPACE 4 #define TERMWID 76 #define HELPDIR "/usr/help" #define VIEWPROGRAM "/usr/ucb/more" /* program to look at text */ #define VIEWPROGOPTS1 "-d" #define VIEWPROGOPTS2 "-18" #define SHELLPROG "/bin/sh" /* program to execute text */ #define SHELLOPTS "-c" char progname[MAXNAMELEN]; char olddir[MAXNAMELEN + 68]; char newdir[MAXNAMELEN + 68]; char *helpdir; char **environ; /* environment, for forks */ int dumb_flag; int list_flag; int col_flag; int frst_flag; char *helpcmds[] = { "Return - Exit this level of help", "* - Print this message", "? - Reprint this help and its subtopics", "# - Reprint just subtopics of this help", ". - Look at current manual page, if any", "$ - Get information on topic files", ". - Look at topic manual page, if any", " - Look at help for subtopic `topic'" , NULL }; SHAR_EOF fi # end of overwriting check if test -f 'help/help.c' then echo shar: will not over-write existing file "'help/help.c'" else cat << \SHAR_EOF > 'help/help.c' #include "help.h" #define MAXMATCHES 24 /* ************************************************************ * H E L P !! * ========================= * * This program emulates the VMS help facility in the UNIX * environment. The main routine HELP1 looks up help and * subtopics for help. The help texts for various topix * are tree-structured. A directory is defined as a "help" * directory, it has four types of files in it: * main help text: .HLP * manual page name: .MANUAL * subtopic texts: .HLP * subtopic directories: * * Subtopic names must start with an alphanumeric * character. Preferably all subtopics will start with * lowercase letters. * * The routine help1 is recursive, it descends the tree * structure. */ main(argc, argv) int argc; char *argv[]; { int i,j,k; int longlen; char yesno[10]; char *opts; strcpy(progname,argv[0]); helpdir = NULL; getwd(olddir); dumb_flag = 0; list_flag = 0; frst_flag = 1; col_flag = 1; argc--; if (*(*++argv) == '-') { /* must be options */ for (opts = *argv; *opts != '\0'; opts++) { switch(*opts) { case '-' : break; case 'd' : if (argc > 1) { helpdir = *++argv; argc--; } break; case 'l' : list_flag = 1; col_flag = 0; break; case 'q' : dumb_flag = 1; break; case 'C' : col_flag = 1; break; case 'n' : col_flag = 0; default: fprintf(stderr,"%s: %c: bad option.\n", progname,*opts); break; } } argc--; argv++; } if (helpdir == NULL) helpdir = HELPDIR; if (chdir(helpdir) < 0) { fprintf(stderr,"%s: %s: help directory not found.\n", progname,helpdir); exit(1); } if (argc >= 1) { /* treat vector as a help path */ #ifdef DEBUG fprintf(stderr," help path vector, argc=%d, *argv=%s.\n", argc,*argv); #endif help1( "", argv, 0); } else help1( "", NULL, 0); exit(0); } /* ************************************************************** * printhelp: given a string, pop .HLP on the end on do * a more on it. * * This routine sends a help file to more(1). A string * is passed in, which is the name of a help. If the string * is nil, then just use the name HLP. * The return value is -1 if no help file is accessible, * 0 if the more(1) command was called okay with fkoff(); */ printhelp(hs, path) char *hs; { char filename[MAXNAMELEN], comm[MAXNAMELEN + 20]; if (hs == NULL) strcpy(filename,HELPEX); else if (strlen(hs) < 1) strcpy(filename, HELPEX); else { strcpy(filename, hs); strcat(filename,HELPEX); } if ( access(filename, R_OK) < 0 ) { printf("\n Sorry, no help text for %s.\n", (hs==NULL)?"this topic":hs ); return(-1); } if (path != NULL) printf("\n HELP: %s\n",path); fkoff(VIEWPROGRAM,VIEWPROGOPTS1,VIEWPROGOPTS2,filename,NULL); return(0); } /* ************************************************************* * printtopix: print the topics available in this directory * in a nice format. * * This routine does a directory of help options, * and prints them out in a manner similar to that of ls(1). * All filenames which start with anything other than numbers * or letters are not kept. Extensions are stripped off. * * The number of subtopics found is returned, along with * a pointer to a null-terminated vector of string pointers. * If the mode is non-zero, it means just use the strings stored * in topix. If mode==0, then re-allocate space and re-check * the directory. If mode */ printtopix( topix, mode, supress) char *topix[]; int mode, supress; { int i,j,k,l; int longlen; int namewidth, totalnames, rowcnt, colcnt; char *malloc(), *nbuf; char row[TERMWID+1]; char *thisname, *index(); char *s1, *s2, **nxtname, *nxcnt; struct direct *readdir(), *filedat; DIR *dirp, *opendir(); if ( mode ) { for(nxtname=topix, i=0; *nxtname++ != NULL; i++ ); totalnames = i; goto inputtopix; } /* if mode is zero, then allocate space and search the directory */ nbuf = malloc( MAXNAMES * MAXNAMELEN ); nxcnt=nbuf; dirp = opendir("."); if (dirp == NULL) { fprintf(stderr,"%s: Cannot open help directory.\n",progname); return(-1); } for(nxtname=topix, i=0; i < (MAXNAMES-1) ; ) { filedat = readdir(dirp); if (filedat == NULL) break; thisname = filedat->d_name; if ( !(isalnum(*thisname)) ) /* if not in [0-9A-Za-z] */ continue; /* do the next one. */ if ( (s1 = index(thisname,'.')) != NULL) { if (strcmp(s1, HELPEX) == 0) *s1 = '\0'; else if ( strcmp(s1, MANEX) == 0) continue; } if (strlen(thisname) >= MAXNAMELEN) *(thisname+MAXNAMELEN-1) = '\0'; /* copy in data from this loop */ strcpy(nxcnt,thisname); *nxtname++ = nxcnt; /* update pointers for next loop */ nxcnt += strlen(thisname) + 1; i++; } *nxtname++ = NULL; totalnames = i; closedir(dirp); if (totalnames == 0) return(0); /* sort the names in ascending order with exchange algorithm */ for(i=0; i < totalnames-1; i++) for(j=i+1; j 0) { thisname = topix[i]; topix[i] = topix[j]; topix[j] = thisname; } inputtopix: if (supress) return(totalnames); longlen = 0; for(i=0; i < totalnames; i++ ) longlen = ((k=strlen(topix[i]))>longlen)?k:longlen; /* here print the names out in nice columns */ namewidth = longlen + COLUMNSPACE; rowcnt = TERMWID / namewidth; colcnt = (totalnames + (rowcnt-1)) / rowcnt ; if (colcnt <= 0) colcnt = 1; if (col_flag && rowcnt > 1) { printf("\n Subtopics:\n\n"); for(i=0; i < colcnt ; i++ ) { for(k=0; k < TERMWID; row[k++] = ' '); row[k] = '\0'; for(j=0, s1 = row; (i+j) < totalnames; j += colcnt) { row[strlen(row)] = ' '; strcpy(s1,topix[i+j]); s1 = s1 + namewidth; } printf(" %s\n",row); } printf("\n"); } else { for(i=0; i < totalnames; i++) printf("%s\n",topix[i]); } return(totalnames); } /* **************************************************************** * help1: descend recursive help tree and provide some * user services. * * This routine is the heart of the new UNIX help facility. * It climbs recursively around an n-tree of documentation files * and directories. The routine printhelp() outputs a file of * help text, the routine printtopix prints out all subtopics. * * This routine can operate in interactive mode, or not. The * basic cycle of operation is: * if (not list-only-mode) print help. * if (not quiet-mode) print list. * if (not interactive) return. * print prompt and do commands. * return. * * There are a number of commands available in the * interactive mode, they are: * * blank line: up recursive level. * subtopic name: recurse to subtopic. * ?: list topics again. * *: get list of commands * $: get info on topix * ^D: quit help program. * .: man page (if any). * # list topics again * */ help1(ppt,svec,skip) char *ppt; /* prompt */ char *svec[]; int skip; { int i,j,k, err; int no_subs; char answer[MAXLINELEN]; char fullpath[MAXNAMELEN + 11]; char yesno[10]; char *topix[MAXNAMES], *mx[MAXMATCHES]; int matchv(); char *index(), *getenv(); char *s1, *s2, *s3, *rest; char *wvec[MAXMATCHES]; char cmdbuf[90]; if ( svec != NULL && *svec != NULL) { printtopix( topix, 0, 1); if ( (k = matchv(*svec, topix, mx)) == 0) { printf(" sorry, no help path to %s %s\n\n",ppt,*svec); return(-1); } for(i=0, err = -1; mx[i] != NULL; ) { if (i > 0) j = takeit(" Next help path: %s %s\nTry it?",ppt,mx[i]); if (j == 1) { i++; continue; } if (j < 0) break; strcpy(fullpath,ppt); strcat(fullpath," "); strcat(fullpath,mx[i]); if ( chdir(mx[i]) < 0) { if ( *(svec + 1) != NULL ) { i++; continue; } if (list_flag) return(-1); else printhelp( mx[i],fullpath); printf("\n"); if (!dumb_flag) help1(ppt,NULL,1); err = 0; } else { if ( help1(fullpath,(svec+1),0) >= 0) { err = 0; chdir(".."); printf("\n"); if (!dumb_flag) help1(ppt,NULL,1); } else chdir(".."); } i++; } return(err); } if ( !list_flag && !skip) if (printhelp(NULL, ppt) < 0) { return(-1); } if ( !list_flag && !dumb_flag && ( access(MANEX, R_OK) >= 0) && !skip) printf("\n Manual page IS available.\n"); if ( !dumb_flag ) { if ( printtopix( topix, 0, skip) <= 0 ) { no_subs = 1; } else no_subs= 0; } if ( dumb_flag || list_flag ) return(0); while ( 1 ) { if (frst_flag) { printf(" (type '*' for commands)\n"); frst_flag = 0; } if ( strlen(ppt) < 1 ) printf("%s",PROMPT); else printf(" %s %s",ppt,SUBPROMPT); if ( fgets(answer, MAXLINELEN-1, stdin) == NULL ) { printf("\nbye...\n"); exit(1); } /* first remove any leading blanks or tabs */ for(s1=answer; *s1 == ' ' || *s1 == ' '; s1++); /* chop off all of answer after first word. */ s2 = index(s1, ' '); if (s2 == 0) s2 = index(s1, '\n'); if (s2) { *s2 = '\0'; rest = s2+1; } else rest = s1 + (strlen(s1) - 1); makewvec(rest,wvec); if ( strlen(s1) == 0 ) /* on blank line, */ break; /* pop up one level*/ switch (*s1) { case '?': /* ?: print stuff again */ printhelp(NULL, ppt); if ( access(MANEX, R_OK) >= 0 ) printf("\n Manual page IS available.\n"); if (!no_subs) printtopix(topix, 1,0); else printf("\n Sorry, no subtopics.\n\n"); break; case '#': if (!no_subs) printtopix( topix, 1,0); else printf("\n Sorry, no subtopics.\n\n"); break; case '*': /* *: list commands */ printf("\n"); for(i=0; helpcmds[i] != NULL; i++) { printf(" %s\n",helpcmds[i]); } printf("\n"); break; case '$': /* $: find out about files */ s2 = s1 + 1; if (no_subs) { printf("\n Sorry, no subtopics.\n\n"); break; } strcpy(cmdbuf, INFOHEAD); strcat(cmdbuf,s2); strcat(cmdbuf,"* "); strcat(cmdbuf, INFOTAIL); printf("\n File information: \n"); system(cmdbuf); printf("\n"); break; case '.': /* .: do manpage if any */ s2 = s1 + 1; if (no_subs) { printf("\n Sorry, no subtopics.\n\n"); break; } if ( *s2 == '\0' ) { mx[0] = ""; mx[1] = NULL; k = 1; } else if ( (k = matchv( s2, topix, mx)) == 0 ) { printf("\n Sorry, no topics match %s\n",s2); printf(" (list cmnds with '*', topix with '#')\n"); } for( i=0; i < k; i++ ) { strcpy(cmdbuf,mx[i]); strcat(cmdbuf,MANEX); if (access(cmdbuf, R_OK|X_OK ) < 0) { strcpy(cmdbuf,mx[i]); strcat(cmdbuf,MAN_SUBEX); if (access(cmdbuf,R_OK|X_OK) < 0) { printf("\n Sorry, %s for %s.\n\n", "No manual reference available", (strlen(s2)==0)?ppt:mx[i]); continue; } } if (i > 0) { k= takeit(" Next man page: %s.\n Take it? ", mx[i]); if (k == 1) continue; if (k == -1) break; } fkoff(SHELLPROG,SHELLOPTS,cmdbuf,NULL); /* source contents of .MANUAL file */ } break; default: /* must be a topic spec */ if ( no_subs ) { printf("\n Sorry, no subtopics.\n\n"); break; } if ( (k = matchv( s1, topix, mx)) == 0 ) { printf("\n Sorry, no help for %s.\n",s1); printf(" (list cmnds with '*', topix with '#')\n"); break; /* leave switch */ } #ifdef DEBUG fprintf(stderr,"diag: wvec= '%s' '%s' ...\n", wvec[0],(wvec[0]!=NULL)?(wvec[1]):NULL); #endif for(i=0; i < k; i++) { s3 = mx[i]; if (i > 0) { j=takeit("\nNext %s subtopic: %s\nTake it?", ppt,s3); if (j == 1) continue; if (j == -1) break; } if ( chdir(s3) >= 0 ) { /* directory subtopic */ strcpy(fullpath,ppt); strcat(fullpath," "); strcat(fullpath,s3); help1(fullpath,wvec,0); /* recurse */ if ( strcmp(getwd(newdir),helpdir) ) chdir(".."); } /* else text subtopic */ else { strcpy(fullpath, ppt); strcat(fullpath, " "); strcat(fullpath, s3); printhelp(s3, fullpath); } } } /* end of switch */ } /* end of while(1) */ if (!no_subs && (*topix != NULL) ) free( *topix ); return(0); } /* end of help1 */ /* **************************************************************** * takeit: ask user whether to take next topic, return t or f * * This routine takes a message, with format, and asks the * user whether he wants to do the action, not do the action, * or quit the cycle of actions. * y - return 0 * n - return 1 * q - return -1 * ? - tell what y, n, and q do. * * - tell what y, n, and q do. * - carriage return * * Naturally, case is insignificant. */ takeit(fmt,m1,m2,m3) char *fmt, *m1, *m2, *m3; { char ans[40]; int done, ret; char *fgets(); for(done = 0; !done; ) { if (fmt != NULL) printf(fmt, m1, m2, m3); printf(" [ynq] "); if (fgets(ans, 39, stdin) == NULL) exit(1); /* abort */ isupper(*ans)?(*ans=tolower(*ans)):0; if (*ans == 'n') { done = 1; ret = 1; } else if (*ans == 'y' || *ans == '\n') { done = 1; ret = 0; } else if (*ans == 'q') { done = 1; ret = -1; } else printf(" Answer y to get next subtopic, n to skip, q to quit.\n"); } return(ret); } /* *************************************************************** * matchv: return all matches of a string in a vector. * * This routine accepts a string and a vector of strings. * The string is supposed to be an abbreviation of one or more * strings in the vector. The full versions of the strings are * placed in a another vector (which is in a static area) and * a pointer to that vector returned. Note that the input * vector of pointers must have NULL as its terminating element. * The output vector will also terminate with NULL. * NOTE: The vector returned contains pointers into the input * vector. Do not, therefore, mess with the contents of the output * vector. * * If NULL is returned, nothing matched, or there was some other * error. */ matchv( src, vec, mx) char *src; char *vec[]; char *mx[]; { char *m[MAXMATCHES]; char *s1, *s2; int i,j, slen; if ( (slen = strlen(src)) == 0 ) return(NULL); for(i=0, j=0; vec[i] != NULL && j < MAXMATCHES; i++) if ( strncmp(src,vec[i],slen) == 0 && strlen(vec[i]) >= slen ) m[j++] = vec[i]; m[j] = NULL; for(i=0; i <= j; mx[i] = m[i], i++); return(j); } /* ***************************************************************** * fkoff: fork a process and return the exit status of the process. * * This routine takes a command line separated into words, and * uses vfork(2) and execve(2) to quickly run the program. * * This is a simplified version of the fkoff() routine from dcon(8). * Here, a program name and up to four arguments may be passed in. * If one of them is null, FINE, but the fifth * */ fkoff(prg,arg1,arg2,arg3,arg4) char *prg, *arg1, *arg2, *arg3, *arg4; { char *command, *malloc(), *argvec[6]; int pid, stat, i; for(i=0; i < 6; argvec[i++] = NULL); argvec[0] = prg; argvec[1] = arg1; argvec[2] = arg2; argvec[3] = arg3; argvec[4] = arg4; if (argvec[0] == NULL) return(-1); command = malloc( strlen(argvec[0]) + 1); strcpy(command,argvec[0]); pid = vfork(); /* fork 2 copies of us */ if (pid == 0) { /* we are child, execve the program. */ pid = execve(command,argvec,environ); _exit(1); } else { /* we are parent, wait for child */ pid = wait(&stat); if (pid == -1) return(pid); stat = stat / 0400; /* get hi byte of status */ return(stat); } } /* ************************************************************** * makewvec: make a vector of words, up to 7 of them * * This routine uses index to simply parse a string * made up of (possibly) several blank-separated words. * The words are stored into the slots of a vector, as pointers * into the original string. Therefore, the original string * is destroyed. */ makewvec(sstr, rvec) char *sstr; char *rvec[]; { int mcnt = 0; int done = 0; char *s1, *s2, *index(); if (strlen(sstr) == 0) { rvec[0] = NULL; } else { /* skip leading whitespace */ for(s1=sstr; iswhite(*s1); s1++); for(mcnt = 0, done = 0; !done; mcnt++) { s2 = index(s1,' '); if (s2 == NULL) s2 = index(s1,' '); if (s2 == NULL) s2 = index(s1,'\n'); if (s2 != NULL) *s2 = '\0'; rvec[mcnt] = s1; if (s2 == NULL) done = 1; else { /* skip more white space */ for(s1 = s2+1; iswhite(*s1); s1++); if (*s1 == '\0') done = 1; } } rvec[mcnt] = NULL; } return(mcnt); } SHAR_EOF fi # end of overwriting check # End of shell archive exit 0