X-Git-Url: http://git.marmaro.de/?p=mmh;a=blobdiff_plain;f=sbr%2Fm_getfld.c;h=6931b2eb25d1f41fbe3cc6b83741a7c7587c735d;hp=548b3eaf2c4f7c049da7e6f339e811a10dcd24dd;hb=10eff860a28b96582526eb739fd0a55441669938;hpb=5e8e1811bd39ba43dda7a3446694474ab16ab960 diff --git a/sbr/m_getfld.c b/sbr/m_getfld.c index 548b3ea..6931b2e 100644 --- a/sbr/m_getfld.c +++ b/sbr/m_getfld.c @@ -8,53 +8,22 @@ #include #include +#include +#include /* -** This module has a long and checkered history. First, it didn't burst -** maildrops correctly because it considered two CTRL-A:s in a row to be -** an inter-message delimiter. It really is four CTRL-A:s followed by a -** newline. Unfortunately, MMDF will convert this delimiter *inside* a -** message to a CTRL-B followed by three CTRL-A:s and a newline. This -** caused the old version of m_getfld() to declare eom prematurely. The -** fix was a lot slower than +** This module has a long and checkered history. ** -** c == '\001' && peekc (iob) == '\001' +** [ Here had been some history of delimiter problems in MMDF maildrops ... ] ** -** but it worked, and to increase generality, MBOX style maildrops could -** be parsed as well. Unfortunately the speed issue finally caught up with -** us since this routine is at the very heart of MH. +** Unfortunately the speed issue finally caught up with us since this +** routine is at the very heart of MH. To speed things up considerably, the +** routine Eom() was made an auxilary function called by the macro eom(). +** Unless we are bursting a maildrop, the eom() macro returns FALSE saying +** we aren't at the end of the message. ** -** To speed things up considerably, the routine Eom() was made an auxilary -** function called by the macro eom(). Unless we are bursting a maildrop, -** the eom() macro returns FALSE saying we aren't at the end of the -** message. +** [ ... and here had been some more of it. ] ** -** After mhl was made a built-in in msh, m_getfld() worked just fine -** (using m_unknown() at startup). Until one day: a message which was -** the result of a bursting was shown. Then, since the burst boundaries -** aren't CTRL-A:s, m_getfld() would blinding plunge on past the boundary. -** Very sad. The solution: introduce m_eomsbr(). This hook gets called -** after the end of each line (since testing for eom involves an fseek()). -** This worked fine, until one day: a message with no body portion arrived. -** Then the -** -** while (eom(c = getc(iob), iob)) -** continue; -** -** loop caused m_getfld() to return FMTERR. So, that logic was changed to -** check for (*eom_action) and act accordingly. -** -** [ Note by meillo 2011-10: -** as msh was removed from mmh, m_eomsbr() became irrelevant. ] -** -** This worked fine, until one day: someone didn't use four CTRL:A's as -** their delimiters. So, the bullet got bit and we read mts.h and -** continue to struggle on. It's not that bad though, since the only time -** the code gets executed is when inc (or msh) calls it, and both of these -** have already called mts_init(). -** -** [ Note by meillo 2012-02: -** MMDF-style maildrops (4x ^A) and mts_init() were removed. ] ** ** ------------------------ ** (Written by Van Jacobson for the mh6 m_getfld, January, 1986): @@ -76,9 +45,8 @@ ** If you hack on this and slow it down, I, my children and my ** children's children will curse you. ** -** This routine gets used on three different types of files: normal, -** single msg files, "packed" unix or mmdf mailboxs (when used by inc) -** and packed, directoried bulletin board files (when used by msh). +** This routine gets used on two different types of files: normal, +** single msg files and "packed" unix mailboxs (when used by inc). ** The biggest impact of different file types is in "eom" testing. The ** code has been carefully organized to test for eom at appropriate ** times and at no other times (since the check is quite expensive). @@ -171,7 +139,7 @@ int ismbox = FALSE; ** is used in m_Eom because the first character of the string ** has been read and matched before m_Eom is called. */ -char *msg_delim = ""; +static char *msg_delim = ""; static unsigned char *fdelim; static unsigned char *delimend; @@ -188,14 +156,6 @@ static int (*eom_action)(int) = NULL; # define DEFINED__FILBUF_TO_SOMETHING_SPECIFIC #endif -#ifdef SCO_5_STDIO -# define _ptr __ptr -# define _cnt __cnt -# define _base __base -# define _filbuf(fp) ((fp)->__cnt = 0, __filbuf(fp)) -# define DEFINED__FILBUF_TO_SOMETHING_SPECIFIC -#endif - #ifndef DEFINED__FILBUF_TO_SOMETHING_SPECIFIC extern int _filbuf(FILE*); #endif @@ -205,8 +165,8 @@ int m_getfld(int state, unsigned char *name, unsigned char *buf, int bufsz, FILE *iob) { - register unsigned char *bp, *cp, *ep, *sp; - register int cnt, c, i, j; + unsigned char *bp, *cp, *ep, *sp; + int cnt, c, i, j; if ((c = getc(iob)) < 0) { msg_count = 0; @@ -227,8 +187,6 @@ m_getfld(int state, unsigned char *name, unsigned char *buf, } switch (state) { - case FLDEOF: - case BODYEOF: case FLD: if (c == '\n' || c == '-') { /* we hit the header/body separator */ @@ -465,8 +423,8 @@ m_getfld(int state, unsigned char *name, unsigned char *buf, *cp++ = j = *(iob->_ptr + c); c = _filbuf(iob); #endif - if (c == EOF || - ((j == '\0' || j == '\n') && c != ' ' && c != '\t')) { + if (c == EOF || ((j == '\0' || j == '\n') + && c != ' ' && c != '\t')) { if (c != EOF) { #ifdef LINUX_STDIO --iob->_IO_read_ptr; @@ -604,7 +562,7 @@ m_getfld(int state, unsigned char *name, unsigned char *buf, break; default: - adios(NULL, "m_getfld() called with bogus state of %d", state); + adios(EX_SOFTWARE, NULL, "m_getfld() called with bogus state of %d", state); } finish: *cp = 0; @@ -612,16 +570,20 @@ finish: return (state); } -static char fromline[BUFSIZ] = ""; void -m_unknown(FILE *iob) +thisisanmbox(FILE *iob) { - register int c; - register long pos; + int c; char text[10]; - register char *cp; - register char *delimstr; + char *cp; + char *delimstr; + + c = getc(iob); + if (feof(iob)) { + return; + } + ungetc(c, iob); /* ** Figure out what the message delimitter string is for this @@ -635,22 +597,19 @@ m_unknown(FILE *iob) ** abort. */ - pos = ftell(iob); if (fread(text, sizeof(*text), 5, iob) != 5) { - adios(NULL, "Read error"); + adios(EX_IOERR, NULL, "Read error"); } if (strncmp(text, "From ", 5)!=0) { - adios(NULL, "No Unix style (mbox) maildrop."); + adios(EX_USAGE, NULL, "No Unix style (mbox) maildrop."); } ismbox = TRUE; delimstr = "\nFrom "; - cp = fromline; - while ((c = getc(iob)) != '\n' && cp - fromline < BUFSIZ - 1) - *cp++ = c; - *cp = '\0'; - + while ((c = getc(iob)) != '\n' && c >= 0) { + continue; + } c = strlen(delimstr); - fdelim = (unsigned char *) mh_xmalloc((size_t) (c + 3)); + fdelim = (unsigned char *) mh_xcalloc((size_t) (c + 3), sizeof(char)); *fdelim++ = '\0'; *fdelim = '\n'; msg_delim = (char *)fdelim+1; @@ -660,14 +619,14 @@ m_unknown(FILE *iob) strcpy(msg_delim, delimstr); delimend = (unsigned char *)msg_delim + edelimlen; if (edelimlen <= 1) - adios(NULL, "maildrop delimiter must be at least 2 bytes"); + adios(EX_DATAERR, NULL, "maildrop delimiter must be at least 2 bytes"); /* ** build a Boyer-Moore end-position map for the matcher in m_getfld. ** N.B. - we don't match just the first char (since it's the newline ** separator) or the last char (since the matchc would have found it ** if it was a real delim). */ - pat_map = (unsigned char **) calloc(256, sizeof(unsigned char *)); + pat_map = (unsigned char **) mh_xcalloc(256, sizeof(unsigned char *)); for (cp = (char *) fdelim + 1; cp < (char *) delimend; cp++ ) pat_map[(unsigned char)*cp] = (unsigned char *) cp; @@ -681,10 +640,9 @@ m_unknown(FILE *iob) static int m_Eom(int c, FILE *iob) { - register long pos = 0L; - register int i; + long pos = 0L; + int i; char text[10]; - register char *cp; pos = ftell(iob); if ((i = fread(text, sizeof *text, edelimlen, iob)) != edelimlen || @@ -702,51 +660,23 @@ m_Eom(int c, FILE *iob) } if (ismbox) { - cp = fromline; - while ((c = getc(iob)) != '\n' && c >= 0 && cp - fromline < BUFSIZ - 1) - *cp++ = c; - *cp = '\0'; + while ((c = getc(iob)) != '\n' && c >= 0) { + continue; + } } return 1; } -/* -** Return the Return-Path and Delivery-Date header information. -** -** Currently, I'm assuming that the "From " line takes the following form: -** "From" sender@host date (sendmail delivery) -*/ -int -get_returnpath(char *rp, int rplen, char *dd, int ddlen) -{ - char *cp; - - if (!(cp = strchr(fromline, ' '))) { - return 0; - } - /* Extract the Return-Path. */ - snprintf(rp, rplen, "%.*s\n", (int)(cp - fromline), fromline); - /* Advance over the whitespace. */ - while (*cp == ' ' || *cp == '\t') { - cp++; - } - /* Extract the Delivery-Date. */ - snprintf(dd, ddlen, "%.*s\n", 24, cp); - *fromline = '\0'; - return 1; -} - - static unsigned char * matchc(int patln, char *pat, int strln, char *str) { - register char *es = str + strln - patln; - register char *sp; - register char *pp; - register char *ep = pat + patln; - register char pc = *pat++; + char *es = str + strln - patln; + char *sp; + char *pp; + char *ep = pat + patln; + char pc = *pat++; for(;;) { while (pc != *str++)