X-Git-Url: http://pileus.org/git/?a=blobdiff_plain;f=rfc822.c;h=15b88f05ef476312af280de1d2ec7183ed4768db;hb=e4dd196b137223195739b9e0f50ec2a8a02b3534;hp=4fec994178f518ac8f4e31c1d50c9318ad011a24;hpb=28246f879f681d83be8653a5bc7aff8bb16a45eb;p=~andy%2Ffetchmail diff --git a/rfc822.c b/rfc822.c index 4fec9941..15b88f05 100644 --- a/rfc822.c +++ b/rfc822.c @@ -1,38 +1,64 @@ -/* - * rfc822.c -- code for slicing and dicing RFC822 mail headers - * - * Copyright 1997 by Eric S. Raymond - * For license terms, see the file COPYING in this directory. - */ +/***************************************************************************** + +NAME: + rfc822.c -- code for slicing and dicing RFC822 mail headers + +ENTRY POINTS: + nextaddr() -- parse the next address out of an RFC822 header + reply_hack() -- append hostname to local header addresses + +THEORY: + How to parse RFC822 headers in C. This is not a fully conformant +implementation of RFC822 or RFC2822, but it has been in production use +in a widely-deployed MTA (fetcmail) since 1996 without complaints. +Really perverse combinations of quoting and commenting could break it. + +AUTHOR: + Eric S. Raymond , 1997. This source code example +is part of fetchmail and the Unix Cookbook, and are released under the +MIT license. Compile with -DMAIN to build the demonstrator. + +******************************************************************************/ + +#include "config.h" #include #include #include -#if defined(STDC_HEADERS) +#include #include -#endif -#include "config.h" #include "fetchmail.h" +#include "sdump.h" + +#ifndef MAIN #include "i18n.h" +#else +#include +static int verbose; +const char *program_name = "rfc822"; +#endif /* MAIN */ + +#ifndef TRUE +#define TRUE 1 +#define FALSE 0 +#endif #define HEADER_END(p) ((p)[0] == '\n' && ((p)[1] != ' ' && (p)[1] != '\t')) -#ifdef TESTMAIN -static int verbose; -char *program_name = "rfc822"; -#endif /* TESTMAIN */ +#define BEFORE_EOL(s) (strcspn((s), "\r\n")) -unsigned char *reply_hack(buf, host) +char *reply_hack( + char *buf /* header to be hacked */, + const char *host /* server hostname */, + size_t *length) /* hack message headers so replies will work properly */ -unsigned char *buf; /* header to be hacked */ -const unsigned char *host; /* server hostname */ { - unsigned char *from, *cp, last_nws = '\0', *parens_from = NULL; + char *from, *cp, last_nws = '\0', *parens_from = NULL; int parendepth, state, has_bare_name_part, has_host_part; -#ifndef TESTMAIN +#ifndef MAIN int addresscount = 1; -#endif /* TESTMAIN */ +#endif /* MAIN */ if (strncasecmp("From:", buf, 5) && strncasecmp("To:", buf, 3) @@ -52,16 +78,18 @@ const unsigned char *host; /* server hostname */ return(buf); } -#ifndef TESTMAIN - if (outlevel >= O_DEBUG) - report_build(stdout, _("About to rewrite %s"), buf); +#ifndef MAIN + if (outlevel >= O_DEBUG) { + report_build(stdout, GT_("About to rewrite %s...\n"), (cp = sdump(buf, BEFORE_EOL(buf)))); + xfree(cp); + } /* make room to hack the address; buf must be malloced */ for (cp = buf; *cp; cp++) - if (*cp == ',' || isspace(*cp)) + if (*cp == ',' || isspace((unsigned char)*cp)) addresscount++; - buf = (unsigned char *)xrealloc(buf, strlen(buf) + addresscount * strlen(host) + 1); -#endif /* TESTMAIN */ + buf = (char *)xrealloc(buf, strlen(buf) + addresscount * (strlen(host) + 1) + 1); +#endif /* MAIN */ /* * This is going to foo up on some ill-formed addresses. @@ -73,13 +101,13 @@ const unsigned char *host; /* server hostname */ has_host_part = has_bare_name_part = FALSE; for (from = buf; *from; from++) { -#ifdef TESTMAIN +#ifdef MAIN if (verbose) { printf("state %d: %s", state, buf); - printf("%*s^\n", from - buf + 10, " "); + printf("%*s^\n", (int)(from - buf + 10), " "); } -#endif /* TESTMAIN */ +#endif /* MAIN */ if (state != 2) { if (*from == '(') @@ -97,11 +125,11 @@ const unsigned char *host; /* server hostname */ break; case 1: /* we've seen the colon, we're looking for addresses */ - if (!isspace(*from)) + if (!isspace((unsigned char)*from)) last_nws = *from; if (*from == '<') state = 3; - else if (*from == '@') + else if (*from == '@' || *from == '!') has_host_part = TRUE; else if (*from == '"') state = 2; @@ -116,12 +144,12 @@ const unsigned char *host; /* server hostname */ && last_nws != ';') { int hostlen; - unsigned char *p; + char *p; p = from; if (parens_from) from = parens_from; - while (isspace(*from) || (*from == ',')) + while (isspace((unsigned char)*from) || (*from == ',')) --from; from++; hostlen = strlen(host); @@ -139,7 +167,7 @@ const unsigned char *host; /* server hostname */ { parens_from = from; } - else if (!isspace(*from)) + else if (!isspace((unsigned char)*from)) has_bare_name_part = TRUE; break; @@ -158,9 +186,9 @@ const unsigned char *host; /* server hostname */ break; case 3: /* we're in a <>-enclosed address */ - if (*from == '@') + if (*from == '@' || *from == '!') has_host_part = TRUE; - else if (*from == '>' && from[-1] != '<') + else if (*from == '>' && (from > buf && from[-1] != '<')) { state = 1; if (!has_host_part) @@ -182,30 +210,34 @@ const unsigned char *host; /* server hostname */ /* * If we passed a comma, reset everything. */ - if (from[-1] == ',' && !parendepth) { + if ((from > buf && from[-1] == ',') && !parendepth) { has_host_part = has_bare_name_part = FALSE; parens_from = NULL; } } -#ifndef TESTMAIN - if (outlevel >= O_DEBUG) - report_complete(stdout, _("Rewritten version is %s\n"), buf); -#endif /* TESTMAIN */ +#ifndef MAIN + if (outlevel >= O_DEBUG) { + report_complete(stdout, GT_("...rewritten version is %s.\n"), + (cp = sdump(buf, BEFORE_EOL(buf)))); + xfree(cp) + } + +#endif /* MAIN */ + *length = strlen(buf); return(buf); } -unsigned char *nxtaddr(hdr) +char *nxtaddr(const char *hdr /* header to be parsed, NUL to continue previous hdr */) /* parse addresses in succession out of a specified RFC822 header */ -const unsigned char *hdr; /* header to be parsed, NUL to continue previous hdr */ { - static unsigned char address[POPBUFSIZE+1]; - static int tp; - static const unsigned char *hp; + static char address[BUFSIZ]; + static size_t tp; + static const char *hp; static int state, oldstate; -#ifdef TESTMAIN - static const unsigned char *orighdr; -#endif /* TESTMAIN */ +#ifdef MAIN + static const char *orighdr; +#endif /* MAIN */ int parendepth = 0; #define START_HDR 0 /* before header colon */ @@ -222,21 +254,23 @@ const unsigned char *hdr; /* header to be parsed, NUL to continue previous hdr * { hp = hdr; state = START_HDR; -#ifdef TESTMAIN +#ifdef MAIN orighdr = hdr; -#endif /* TESTMAIN */ +#endif /* MAIN */ tp = 0; } + if (!hp) return NULL; + for (; *hp; hp++) { -#ifdef TESTMAIN +#ifdef MAIN if (verbose) { printf("state %d: %s", state, orighdr); - printf("%*s^\n", hp - orighdr + 10, " "); + printf("%*s^\n", (int)(hp - orighdr + 10), " "); } -#endif /* TESTMAIN */ +#endif /* MAIN */ if (state == ENDIT_ALL) /* after last address */ return(NULL); @@ -245,13 +279,13 @@ const unsigned char *hdr; /* header to be parsed, NUL to continue previous hdr * state = ENDIT_ALL; if (tp) { - while (isspace(address[--tp])) - continue; - address[++tp] = '\0'; + while (tp > 0 && isspace((unsigned char)address[tp - 1])) + tp--; + address[tp] = '\0'; tp = 0; return (address); } - return((unsigned char *)NULL); + return(NULL); } else if (*hp == '\\') /* handle RFC822 escaping */ { @@ -286,7 +320,7 @@ const unsigned char *hdr; /* header to be parsed, NUL to continue previous hdr * state = INSIDE_BRACKETS; tp = 0; } - else if (*hp != ',' && !isspace(*hp)) + else if (*hp != ',' && !isspace((unsigned char)*hp)) { --hp; state = BARE_ADDRESS; @@ -315,18 +349,20 @@ const unsigned char *hdr; /* header to be parsed, NUL to continue previous hdr * state = INSIDE_BRACKETS; tp = 0; } - else if (!isspace(*hp)) /* just take it, ignoring whitespace */ + else if (*hp == '"') /* quoted word, copy verbatim */ + { + oldstate = state; + state = INSIDE_DQUOTE; + address[NEXTTP()] = *hp; + } + else if (!isspace((unsigned char)*hp)) /* just take it, ignoring whitespace */ address[NEXTTP()] = *hp; break; case INSIDE_DQUOTE: /* we're in a quoted string, copy verbatim */ - if (*hp != '"') - address[NEXTTP()] = *hp; - else - { - address[NEXTTP()] = *hp; + address[NEXTTP()] = *hp; + if (*hp == '"') state = oldstate; - } break; case INSIDE_PARENS: /* we're in a parenthesized comment, ignore */ @@ -364,30 +400,31 @@ const unsigned char *hdr; /* header to be parsed, NUL to continue previous hdr * return(NULL); } -#ifdef TESTMAIN -static void parsebuf(unsigned char *longbuf, int reply) +#ifdef MAIN +static void parsebuf(char *longbuf, int reply) { - unsigned char *cp; + char *cp; + size_t dummy; if (reply) { - reply_hack(longbuf, "HOSTNAME.NET"); - printf("Rewritten buffer: %s", longbuf); + reply_hack(longbuf, "HOSTNAME.NET", &dummy); + printf("Rewritten buffer: %s", (char *)longbuf); } else - if ((cp = nxtaddr(longbuf)) != (unsigned char *)NULL) + if ((cp = nxtaddr(longbuf)) != (char *)NULL) do { - printf("\t-> \"%s\"\n", cp); + printf("\t-> \"%s\"\n", (char *)cp); } while - ((cp = nxtaddr((unsigned char *)NULL)) != (unsigned char *)NULL); + ((cp = nxtaddr((char *)NULL)) != (char *)NULL); } -main(int argc, char *argv[]) +int main(int argc, char *argv[]) { - unsigned char buf[MSGBUFSIZE], longbuf[BUFSIZ]; - int ch, reply; + char buf[BUFSIZ], longbuf[BUFSIZ]; + int ch, reply; verbose = reply = FALSE; while ((ch = getopt(argc, argv, "rv")) != EOF) @@ -402,16 +439,18 @@ main(int argc, char *argv[]) break; } + longbuf[0] = '\0'; + while (fgets(buf, sizeof(buf)-1, stdin)) { if (buf[0] == ' ' || buf[0] == '\t') - strcat(longbuf, buf); + strlcat(longbuf, buf, sizeof(longbuf)); else if (!strncasecmp("From: ", buf, 6) || !strncasecmp("To: ", buf, 4) || !strncasecmp("Reply-", buf, 6) || !strncasecmp("Cc: ", buf, 4) || !strncasecmp("Bcc: ", buf, 5)) - strcpy(longbuf, buf); + strlcpy(longbuf, buf, sizeof(longbuf)); else if (longbuf[0]) { if (verbose) @@ -426,7 +465,8 @@ main(int argc, char *argv[]) fputs(longbuf, stdout); parsebuf(longbuf, reply); } + exit(0); } -#endif /* TESTMAIN */ +#endif /* MAIN */ /* rfc822.c end */