find-patterns/killspam/killspam.c

/* Copyright 2006 Neil Edelman, distributed under the terms of the
 GNU General Public License, see copying.txt */

/* this is a LCC-WIN32 mail programme that goes with Find Patterns */

#include <stdlib.h> /* atoi(), getenv(), malloc(), free() */
#include <string.h> /* memcpy(), strchr() */
#include <stdio.h> /* printf() */
#include <conio.h> /* getch() */
#include <regexp.h> /* I'm lazy so I'll use LCC-WIN32's regexps */
#include "net.h"

#define MAX_REGX_SIZE 0x4000
#define BUF_SIZE 256
#define LOG_FILE_NAME "email"
#define SPAM_FILE_NAME "spamregx"
#define SPAM_ENV_VAR "SPAMREGX"

static const char *programme    = "Vindicator";
static const char *year        = "2006";
static const int versionMajor  = 0;
static const int versionMinor  = 1;

typedef struct tagMailSession {
	enum {
		WAIT_GREET,
		WAIT_USEROK,
		WAIT_PASSOK,
		WAIT_STAT,
		CHECKA_EMAIL,
		WAIT_RETROK,
		PARSE_DATA,
		WAIT_DELEOK,
		RUN_AWAY
	} state;
	int mailPos;
	int mailCount;
	int spamFactor;
	unsigned int lnSize;
	char line[BUF_SIZE];
	unsigned int rpSize;
	char resp[BUF_SIZE];
	FILE *log;
	regexp *subjRe;
	regexp *fromRe;
	regexp *spamRe;
} MailSession;

static MailSession *create_mail_session(void);
static void destroy_mail_session(MailSession *ms);
static int check_email(MailSession *ms, Net *net, char *name, unsigned int port);
static int get_response(MailSession *ms);
static int getstr(char **buf, int size, int echo);

int main(int argc, char **argv) {
	Net *net;
	MailSession *ms;
	int ret, port;
	char *server;

	printf("%s\n\n", programme);
	if(argc <= 1) {
		printf("Arguments: [<server> [<port>]]\n\n");
		printf("Connects to <server> (default localhost) on port <port> (default 110)\n");
		printf("retreives emails, logging them into \"%s\" if it exists. If the\n", LOG_FILE_NAME);
		printf("environment variable \"%s\" is defined or the file \"%s\"\n", SPAM_ENV_VAR, SPAM_FILE_NAME);
		printf("exists, it loads the a regexp from them. Any emails matching the regexp\n");
		printf("can then be annihilated.\n\n");
		fprintf(stderr, "Version %d.%d.\n\n", versionMajor, versionMinor);
		fprintf(stderr, "%s Copyright %s Neil Edelman\n", programme, year);
		fprintf(stderr, "This program comes with ABSOLUTELY NO WARRANTY.\n");
		fprintf(stderr, "This is free software, and you are welcome to redistribute it\n");
		fprintf(stderr, "under certain conditions; see copying.txt.\n\n");
		return 1;
	}
	port = (argc >= 3) ? atoi(argv[2]) : 110;
	server = (argc >= 2) ? argv[1] : "localhost";
	if(!(net = create_net(1024))) {
		printf("Failed creating connection data.\n");
		return 1;
	}
	if(!(ms = create_mail_session())) {
		printf("Failed creating e-mail session data.\n");
		destroy_net(net);
		return 1;
	}
	ret = check_email(ms, net, server, port);
	destroy_mail_session(ms);
	destroy_net(net);
	return ret ? 0 : 1;
}

static MailSession *create_mail_session(void) {
	MailSession *ms;

	/* create and initialize the memory space */
	if(!(ms = malloc(sizeof(MailSession)))) {
		perror("Error allocating memory for e-mail data.\n");
		return 0;
	}
	printf("Created mail session data.\n");
	ms->state = WAIT_GREET;
	ms->mailPos = 1;
	ms->mailCount = 0;
	/* open the log file */
	if(!(ms->log = fopen(LOG_FILE_NAME, "r"))) {
		perror("Didn't find log file");
		printf("To log email, create a file nammed \"%s.\"\n", LOG_FILE_NAME);
	} else {
		fclose(ms->log);
		if(!(ms->log = fopen(LOG_FILE_NAME, "w")))
			perror("Failed opening log file");
		else
			printf("Opened log file, \"%s,\" for writing.\n", LOG_FILE_NAME);
	}
	if(!(ms->subjRe = regcomp("^[Ss]ubject: ")))
		printf("Error compiling subject regexp.\n");
	if(!(ms->fromRe = regcomp("^[Ff]rom: ")))
		printf("Error compiling sender regexp.\n");
	ms->spamRe = 0;
	/* read in the spam regexp */
	{
		FILE *fp = 0;
		int len;
		char regStr[MAX_REGX_SIZE] /* YIKES! */, *regPtr;

		if(!(regPtr = getenv(SPAM_ENV_VAR))
		&& !(fp = fopen(SPAM_FILE_NAME, "r"))) {
			perror("Didn't find spam regexp file");
			printf("To load a spam regexp, write it in a file nammed \"%s\" or set it\n", SPAM_FILE_NAME);
			printf("as an environment variable, \"%s.\" Please enter the regexp with which to\n", SPAM_ENV_VAR);
			printf("smite the vile spam:\n");
			regPtr = regStr;
			if((len = getstr(&regPtr, sizeof(regStr), 1)) < 0) {
				destroy_mail_session(ms);
				return 0;
			}
			putchar('\n');
		} else if(fp) {
			int ch;

			printf("Reading regexp from file \"%s.\"\n", SPAM_FILE_NAME);
			len = 0;
			regPtr = regStr;
			for( ; ; ) {
				if((ch = fgetc(fp)) == EOF) break;
				if(ch && ch != '\n' && ch != '\r') {
					*(regPtr++) = ch;
					if(len++ >= sizeof(regStr)) {
						printf("Regexp size exceeded limit of %u bytes.\n", sizeof(regStr));
						len = 0;
						break;
					}
				}
			}
			fclose(fp);
			*regPtr = 0;
		} else {
			if((len = strlen(regPtr)) >= sizeof(regStr)) {
				printf("Regexp size exceeded limit of %u bytes.\n", sizeof(regStr));
				len = 0;
			} else if(len >= 2
			&& ((regPtr[0] == '\'' && regPtr[len - 1] == '\'')
			|| (regPtr[0] == '\"' && regPtr[len - 1] == '\"'))) {
				memcpy(regStr, regPtr + 1, len - 2);
				regStr[len - 2] = 0;
			} else {
				memcpy(regStr, regPtr, len);
				regStr[len] = 0;
			}
		}
		if(len > 0) {
			if(!(ms->spamRe = regcomp(regStr)))
				printf("Error compiling spam regexp.\n");
			else
				printf("Spam regexp is %s\n", regStr);
		}
	}
	return ms;
}

static void destroy_mail_session(MailSession *ms) {
	/* Email *eml, *nextEml; */

	if(!ms) return;
	if(ms->log) {
		printf("Closing log file.\n");
		fclose(ms->log); ms->log = 0;
	}
	if(ms->subjRe) { free(ms->subjRe); ms->subjRe = 0; }
	if(ms->fromRe) { free(ms->fromRe); ms->fromRe = 0; }
	if(ms->spamRe) { free(ms->spamRe); ms->spamRe = 0; }
	/* for(eml = ms->emailList; eml; eml = nextEml) {
		nextEml = eml->next;
		free(eml);
	} */
	free(ms);
}

static int check_email(MailSession *ms, Net *net, char *name, unsigned int port) {
	unsigned int lnRem, lnEol, svRem, svEol;
	char *lnPtr, *svNl, *svPtr;

	/* connect to the given host */
	if(!set_net_host(net, name, port)
	|| !connect_net(net)) return 0;
	/* loop reading lines from the server */
	lnPtr = ms->line;
	lnRem = sizeof(ms->line) - 1;
	svPtr = 0;
	svRem = 0;
	for( ; ; ) {
		/* if the server string is empty, check for more */
		if(!svRem && !read_net_string(net, &svPtr, &svRem, 100000)) return 0;
		/* how many chars until the end of the line */
		svEol = (svNl = strchr(svPtr, '\n')) ? svNl - svPtr + 1 : svRem;
		lnEol = (svEol > lnRem) ? lnRem : svEol;
		/* copy over a whole line from the server string into the local one */
		memcpy(lnPtr, svPtr, lnEol);
		lnPtr[lnEol] = 0;
		/*printf("svPtr = %s; svRem = %u; lnRem = %u; svEol = %u; lnEol = %u; lnPtr = %s\n", svPtr, svRem, lnRem, svEol, lnEol, lnPtr);*/
		/* shift all positions by this amount */
		svPtr += lnEol;
		svRem -= lnEol;
		svEol -= lnEol;
		lnPtr += lnEol;
		lnRem -= lnEol;
		/* check if there was a new line or the local buffer is full */
		if(svNl || !lnRem) {
			ms->lnSize = sizeof(ms->line) - lnRem - 1;
			ms->rpSize = sizeof(ms->resp);
			if(!get_response(ms)) break;
			if(ms->rpSize && !send_net_string(net, ms->resp, ms->rpSize)) return 0;
			lnPtr = ms->line;
			lnRem = sizeof(ms->line) - 1;
		}
	}
	return 1;
}


static int get_response(MailSession *ms) {
	char *rpPtr;

	if(ms->state != PARSE_DATA) {
		printf("Mr. Mail Deamon: %s", ms->line);
		if(!ms->lnSize || ms->line[ms->lnSize - 1] != '\n') putchar('\n');
	}
	switch(ms->state) {
	case WAIT_GREET:
		if(*ms->line != '+') {
			printf("If this is a POP3 mail server, it seems to be having an identity crisis.\n");
			return 0;
		}
		printf("The server is glad to see you. It wants your user name: ");
		rpPtr = ms->resp + sprintf(ms->resp, "user ");
		if(getstr(&rpPtr, ms->rpSize + ms->resp - rpPtr, 1) < 0) return 0;
		printf("\n"); rpPtr += sprintf(rpPtr, "\n");
		ms->rpSize = rpPtr - ms->resp;
		printf("%s: %s", programme, ms->resp);
		ms->state = WAIT_USEROK;
		return 1;
	case WAIT_USEROK:
		if(*ms->line == '-') break;
		if(*ms->line != '+') return 1;
		printf("That user name seems acceptable. Now for a password: ");
		rpPtr = ms->resp + sprintf(ms->resp, "pass ");
		if(getstr(&rpPtr, ms->rpSize + ms->resp - rpPtr, 0) < 0) return 0;
		printf("\n"); rpPtr += sprintf(rpPtr, "\n");
		ms->rpSize = rpPtr - ms->resp;
		printf("%s: pass <secret ninja-encrypted password>\n", programme/*, ms->resp*/);
		ms->state = WAIT_PASSOK;
		return 1;
	case WAIT_PASSOK:
		if(*ms->line == '-') break;
		if(*ms->line != '+') return 1;
		printf("The server has accepted you! It thinks you are Special.\n");
		sprintf(ms->resp, "stat\n"); ms->rpSize = 5;
		printf("%s: %s", programme, ms->resp);
		ms->state = WAIT_STAT;
		return 1;
	case WAIT_STAT:
		if(*ms->line == '-') break;
		if(*ms->line != '+') return 1;
		{
			char *numStart;

			rpPtr = ms->line;
			while(*rpPtr && *rpPtr != ' ') rpPtr++;
			while(*rpPtr == ' ') rpPtr++;
			if(!*(numStart = rpPtr)) { printf("That doesn't make any sense!\n"); break; }
			while(*rpPtr && *rpPtr != ' ') rpPtr++;
			*rpPtr = 0;
			ms->mailCount = atoi(numStart);
		}
		if(ms->mailPos <= ms->mailCount) {
			rpPtr = ms->resp + sprintf(ms->resp, "retr %u\n", ms->mailPos);
			ms->rpSize = rpPtr - ms->resp;
			ms->state = WAIT_RETROK;
		} else {
			sprintf(ms->resp, "quit\n"); ms->rpSize = 5;
			ms->state = RUN_AWAY;
		}
		printf("%s: %s", programme, ms->resp);
		return 1;
	case WAIT_RETROK:
		if(*ms->line == '-') break;
		if(*ms->line != '+') return 1;
		*ms->resp = 0;
		ms->rpSize = 0;
		/* keep it simple . . . (who wants confirmation later anyway?)
		{
			Email *eml;

			if(!(eml = malloc(sizeof(Email))))
				perror("Failed allocating memory to store e-mail info.");
			else {
				eml->next = ms->emailList;
				eml->subject[0] = 0;
				eml->sender[0] = 0;
				eml->number = ms->mailPos;
				eml->spamFactor = 0;
				ms->emailList = eml;
			}
		} */
		ms->spamFactor = 0;
		/* printf("[ Scouring message %u for unholy spam . . . ]\n", ms->mailPos); */
		ms->state = PARSE_DATA;
		return 1;
	case PARSE_DATA:
		if(ms->line[0] == '.' && (ms->line[1] == '\r' || ms->line[1] == '\n')) {
			if(ms->log) fprintf(ms->log, ".\n[ End of Message %u. ]\n", ms->mailPos);
			if(ms->spamFactor) {
				rpPtr = ms->resp + sprintf(ms->resp, "dele %u\n", ms->mailPos);
				ms->rpSize = rpPtr - ms->resp;
				ms->state = WAIT_DELEOK;
			} else if(++ms->mailPos <= ms->mailCount) {
				rpPtr = ms->resp + sprintf(ms->resp, "retr %u\n", ms->mailPos);
				ms->rpSize = rpPtr - ms->resp;
				ms->state = WAIT_RETROK;
			} else {
				sprintf(ms->resp, "quit\n"); ms->rpSize = 5;
				ms->state = RUN_AWAY;
			}
			printf("%s: %s", programme, ms->resp);
		} else {
			if(ms->log) fprintf(ms->log, "%s", ms->line);
			/* if(ms->emailList
			&& ms->emailList->number == ms->mailPos) {
				if(!ms->emailList->subject[0]
				&& ms->subjRe && regexec(ms->subjRe, ms->line)) {
					strncpy(ms->emailList->subject, ms->line, sizeof(ms->emailList->subject));
					printf("%s", ms->line);
				} else if(!ms->emailList->sender[0]
				&& ms->fromRe && regexec(ms->fromRe, ms->line)) {
					strncpy(ms->emailList->subject, ms->line, sizeof(ms->emailList->subject));
					printf("%s", ms->line);
				}
				if(!ms->emailList->spamFactor
				&& ms->spamRe && regexec(ms->spamRe, ms->line)) {
					ms->emailList->spamFactor++;
					sm->spamFactor++;
					printf("EVIL SPAM!->%s", ms->line);
				}
			} */
			if(ms->subjRe && regexec(ms->subjRe, ms->line))
				printf("%s", ms->line);
			else if(ms->fromRe && regexec(ms->fromRe, ms->line))
				printf("%s", ms->line);
			if(ms->spamRe && regexec(ms->spamRe, ms->line)) {
				ms->spamFactor++;
				printf("---EVIL SPAM!--->%s", ms->line);
			}
			*ms->resp = 0;
			ms->rpSize = 0;
		}
		return 1;
	case WAIT_DELEOK:
		if(*ms->line == '-') break;
		if(*ms->line != '+') return 1;
		if(++ms->mailPos <= ms->mailCount) {
			rpPtr = ms->resp + sprintf(ms->resp, "retr %u\n", ms->mailPos);
			ms->rpSize = rpPtr - ms->resp;
			ms->state = WAIT_RETROK;
		} else {
			sprintf(ms->resp, "quit\n"); ms->rpSize = 5;
			ms->state = RUN_AWAY;
		}
		printf("%s: %s", programme, ms->resp);
		return 1;
	case RUN_AWAY:
		if(*ms->line != '+') printf("The server is very confused. ");
		printf("Bye server.\n");
		return 0;
	}
	printf("The server is confused!\n");
	sprintf(ms->resp, "quit\n"); ms->rpSize = 5;
	ms->state = RUN_AWAY;
	return 1;
}

static int getstr(char **buf, int size, int echo) {
	int ch;
	int pos = 0;
	char *str = *buf;

	if(size < 1) return -1;
	*str = 0;
	for( ; ; ) {
		ch = getch();
		if(ch <= 0) return -1;
		if((ch & 0x60) == 0x00) {
			switch((ch & 0x1F) | 0x40) {
			case '@': /* ^@ - null */
			case 'C': /* ^C - end of text */
			case 'D': /* ^D - end of transmission */
			case '[': /* ^[ - escape */
				*buf = str;
				return -1;
			case 'U':
				if(echo) { while(pos--) printf("\b \b"); }
				pos = 0; *(str = *buf) = 0;
				break;
			case 'H':
				if(pos) {
					pos--; *(--str) = 0;
					if(echo) printf("\b \b");
				}
				break;
			case 'J': /* ^J - LF (NL) new line */
			case 'M': /* ^M - CR return */
				*buf = str;
				return pos;
			}
		} else if(ch == 0x7F) {
		} else {
			if(pos + 1 < size) {
				if(echo) putchar(ch);
				str[0] = ch;
				str[1] = 0;
				if(pos + 2 < size) { pos++; str++; }
			}
		}
	}
}