htwwwstr.c

www工具包

/*								     HTWWWStr.c
**	WWW RELATED STRING UTILITIES
**
**	(c) COPYRIGHT MIT 1995.
**	Please first read the full copyright statement in the file COPYRIGH.
**	@(#) $Id: HTWWWStr.c,v 2.34 1999/02/22 22:10:12 frystyk Exp $
**
**      Now 13 95	Spwaned from HTString.c
*/

/* Library include files */
#include "wwwsys.h"
#include "WWWUtil.h"
#include "HTParse.h"
#include "HTInet.h"
#include "HTUser.h"
#include "HTWWWStr.h"					 /* Implemented here */

PRIVATE char * months[12] = {
    "Jan","Feb","Mar","Apr","May","Jun","Jul","Aug","Sep","Oct","Nov","Dec"
};

#ifndef HAVE_STRFTIME
PRIVATE char * wkdays[7] = {
    "Sun","Mon","Tue","Wed","Thu","Fri","Sat"
};
#endif

/* ------------------------------------------------------------------------- */

/*	Find next Field
**	---------------
**	Finds the next RFC822 token in a string
**	On entry,
**	*pstr	points to a string containing a word separated
**		by white white space "," ";" or "=". The word
**		can optionally be quoted using <"> or "<" ">"
**		Comments surrrounded by '(' ')' are filtered out
**
** 	On exit,
**	*pstr	has been moved to the first delimiter past the
**		field
**		THE STRING HAS BEEN MUTILATED by a 0 terminator
**
**	Returns	a pointer to the first word or NULL on error
*/
PUBLIC char * HTNextField (char ** pstr)
{
    char * p = *pstr;
    char * start = NULL;
    if (!pstr || !*pstr) return NULL;
    while (1) {
	/* Strip white space and other delimiters */
	while (*p && (isspace((int) *p) || *p==',' || *p==';' || *p=='=')) p++;
	if (!*p) {
	    *pstr = p;
	    return NULL;				   	 /* No field */
	}

	if (*p == '"') {				     /* quoted field */
	    start = ++p;
	    for(;*p && *p!='"'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    break;			    /* kr95-10-9: needs to stop here */
	} else if (*p == '<') {				     /* quoted field */
	    start = ++p;
	    for(;*p && *p!='>'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    break;			    /* kr95-10-9: needs to stop here */
	} else if (*p == '(') {					  /* Comment */
	    for(;*p && *p!=')'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    p++;
	} else {					      /* Spool field */
	    start = p;
	    while(*p && !isspace((int) *p) && *p!=',' && *p!=';' && *p!='=')
		p++;
	    break;						   /* Got it */
	}
    }
    if (*p) *p++ = '\0';
    *pstr = p;
    return start;
}

/*	Find next LWS delimited token
**	-----------------------------
**	On entry,
**	*pstr	points to a string containing a word separated
**		by white white space "," ";" or "=". The word
**		can optionally be quoted using <"> or "<" ">"
**		Comments surrrounded by '(' ')' are filtered out
**
** 	On exit,
**	*pstr	has been moved to the first delimiter past the
**		field
**		THE STRING HAS BEEN MUTILATED by a 0 terminator
**
**	Returns	a pointer to the first word or NULL on error
*/
PUBLIC char * HTNextLWSToken (char ** pstr)
{
    char * p = *pstr;
    char * start = NULL;
    if (!pstr || !*pstr) return NULL;

    /* Strip initial white space  */
    while (*p && (isspace((int) *p))) p++;
    if (!*p) {
	*pstr = p;
	return NULL;				   	 /* No field */
    }

    /* Now search for the next white space */
    start = p;
    while(*p && !isspace((int) *p)) p++;

    if (*p) *p++ = '\0';
    *pstr = p;
    return start;
}

/*	Find next Name-value pair
**	-------------------------
**	This is the same as HTNextField but it does not look for '=' as a
**	separator so if there is a name-value pair then both parts are
**	returned.
**	Returns	a pointer to the first word or NULL on error
*/
PUBLIC char * HTNextPair (char ** pstr)
{
    char * p = *pstr;
    char * start = NULL;
    if (!pstr || !*pstr) return NULL;
    while (1) {
	/* Strip white space and other delimiters */
	while (*p && (*p==',' || *p==';')) p++;
	if (!*p) {
	    *pstr = p;
	    return NULL;				   	 /* No field */
	}

	if (*p == '"') {				     /* quoted field */
	    start = ++p;
	    for(;*p && *p!='"'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    break;			    /* kr95-10-9: needs to stop here */
	} else if (*p == '<') {				     /* quoted field */
	    start = ++p;
	    for(;*p && *p!='>'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    break;			    /* kr95-10-9: needs to stop here */
	} else if (*p == '(') {					  /* Comment */
	    for(;*p && *p!=')'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    p++;
	} else {					      /* Spool field */
	    start = p;
	    while(*p && *p!=',' && *p!=';')
		p++;
	    break;						   /* Got it */
	}
    }
    if (*p) *p++ = '\0';
    *pstr = p;
    return start;
}

/*	Find next element in a comma separated string
**	---------------------------------------------
**	This is the same as HTNextPair but it does not look for anything
**	else than ',' as separator
**	Returns	a pointer to the first word or NULL on error
*/
PUBLIC char * HTNextElement (char ** pstr)
{
    char * p = *pstr;
    char * start = NULL;
    if (!pstr || !*pstr) return NULL;

    /* Strip white space and other delimiters */
    while (*p && ((isspace((int) *p)) || *p==',')) p++;
    if (!*p) {
	*pstr = p;
	return NULL;					   	 /* No field */
    }
    start = p;
    while (1) {
	if (*p == '"') {				     /* quoted field */
	    for(;*p && *p!='"'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    p++;
	} else if (*p == '<') {				     /* quoted field */
	    for(;*p && *p!='>'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    p++;
	} else if (*p == '(') {					  /* Comment */
	    for(;*p && *p!=')'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    p++;
	} else {					      /* Spool field */
	    while (*p && *p!=',') p++;
	    break;						   /* Got it */
	}
    }
    if (*p) *p++ = '\0';
    *pstr = p;
    return start;
}

/*	Find next "/" delimied segment
**	------------------------------
**	This is the same as HTNextField but it includes "/" as a delimiter.
**	Returns	a pointer to the first segment or NULL on error
*/
PUBLIC char * HTNextSegment (char ** pstr)
{
    char * p = *pstr;
    char * start = NULL;
    if (!pstr || !*pstr) return NULL;
    while (1) {
	/* Strip white space and other delimiters */
	while (*p && (isspace((int) *p) || *p==',' || *p==';' || *p=='=' || *p=='/')) p++;
	if (!*p) {
	    *pstr = p;
	    return NULL;				   	 /* No field */
	}

	if (*p == '"') {				     /* quoted field */
	    start = ++p;
	    for(;*p && *p!='"'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    break;			    /* kr95-10-9: needs to stop here */
	} else if (*p == '<') {				     /* quoted field */
	    start = ++p;
	    for(;*p && *p!='>'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    break;			    /* kr95-10-9: needs to stop here */
	} else if (*p == '(') {					  /* Comment */
	    for(;*p && *p!=')'; p++)
		if (*p == '\\' && *(p+1)) p++;	       /* Skip escaped chars */
	    p++;
	} else {					      /* Spool field */
	    start = p;
	    while(*p && !isspace((int) *p) && *p!=',' && *p!=';' && *p!='=' && *p!='/')
		p++;
	    break;						   /* Got it */
	}
    }
    if (*p) *p++ = '\0';
    *pstr = p;
    return start;
}

/*
**	Find the next s-expression token from a string of characters.
**	We return the name of this expression and the param points to the
**	parameters. 
**
**	NOTE: The string has been mutilated by '/0's
*/
PUBLIC char * HTNextSExp (char ** exp, char ** param)
{
    char * p = *exp;
    char * name = NULL;
    if (!exp || !*exp) return NULL;
    while (*p && isspace((int) *p)) p++;		/* Strip leading white space */
    if (!*p) {
	*exp = p;
	return NULL;					   	 /* No field */
    }
    if (*p == '{') {					     /* Open bracket */
	int cnt = 1;
	/*
	**  Look for name of this expression. If we find a token then search
	**  for the rest of the expression and remove the end '}'
	*/
	p++;
	if ((name = HTNextField(&p)) == NULL) return NULL;
	while (*p && isspace((int) *p)) p++;
	*param = p;
	while (*p) {
	    if (*p == '{') cnt++;
	    if (*p == '}') cnt--;
	    if (!cnt) {
		*p = '\0';
		break;
	    }
	    p++;
	}
    }
    return name;
}

/*
**	Returns a Message-ID string including the open '<' and the closing '>'.
**	The format of the string is:
**
**		"<" time-in-sec "Z" process-id "@" FQDN ">"
**
**	or on systems that doesn't have process-id:
**
**		"<" time-in-sec "Z" user "@" FQDN ">"
**
**	Returns a pointer to the MessageID
*/
PUBLIC const char * HTMessageIdStr (HTUserProfile * up)
{
    static char buf[80];
    time_t sectime = time(NULL);
#ifdef HAVE_GETPID
    const char * address = HTUserProfile_fqdn(up);
#else
    const char * address = HTUserProfile_email(up);
#endif /* HAVE_GETPID */
    if (!address) address = tmpnam(NULL);
    if ((!address || !*address) && sectime < 0) {
	HTTRACE(CORE_TRACE, "MessageID...  Can't make a unique MessageID\n");
	return "";
    }
#ifdef HAVE_GETPID
    sprintf(buf, "<%ldZ%ld@%s>", (long)sectime, (long)getpid(), address ? address : "@@@");
#else
    sprintf(buf, "<%ldZ%s>", sectime, address ? address : "@@@");
#endif /* HAVE_GETPID */

    *(buf+79) = '\0';
    return buf;
}

/*	Date and Time Parser
**	--------------------
**	These functions are taken from the server written by Ari Luotonen
*/
#if 0
PRIVATE int make_num (const char *  s)
{
    if (*s >= '0' && *s <= '9')
	return 10 * (*s - '0') + *(s+1) - '0';
    else
	return *(s+1) - '0';
}
#endif
PRIVATE int make_month (char * s, char ** ends)
{
    char * ptr = s;
    while (!isalpha((int) *ptr)) ptr++;
    if (*ptr) {
	int i;
	*ends = ptr+3;		
	for (i=0; i<12; i++)
	    if (!strncasecomp(months[i], ptr, 3)) return i;
    }
    return 0;
}

/*
**	Parse a str in GMT format to a local time time_t representation
**	Four formats are accepted:
**
**		Wkd, 00 Mon 0000 00:00:00 GMT		(rfc1123)
**		Weekday, 00-Mon-00 00:00:00 GMT		(rfc850)
**		Wkd Mon 00 00:00:00 0000 GMT		(ctime)
**		1*DIGIT					(delta-seconds)
*/
PUBLIC time_t HTParseTime (const char * str, HTUserProfile * up, BOOL expand)
{
    char * s;
    struct tm tm;
    time_t t;

    if (!str) return 0;

    if ((s = strchr(str, ','))) {	 /* Thursday, 10-Jun-93 01:29:59 GMT */
	s++;				/* or: Thu, 10 Jan 1993 01:29:59 GMT */
	while (*s && *s==' ') s++;
	if (strchr(s,'-')) {				     /* First format */
	    HTTRACE(CORE_TRACE, "Format...... Weekday, 00-Mon-00 00:00:00 GMT\n");
	    if ((int)strlen(s) < 18) {
		HTTRACE(CORE_TRACE, "ERROR....... Not a valid time format \"%s\"\n" _ s);
		return 0;
	    }
	    tm.tm_mday = strtol(s, &s, 10);
	    tm.tm_mon = make_month(s, &s);
	    tm.tm_year = strtol(++s, &s, 10);
	    tm.tm_hour = strtol(s, &s, 10);
	    tm.tm_min = strtol(++s, &s, 10);
	    tm.tm_sec = strtol(++s, &s, 10);

	} else {					    /* Second format */
	    HTTRACE(CORE_TRACE, "Format...... Wkd, 00 Mon 0000 00:00:00 GMT\n");
	    if ((int)strlen(s) < 20) {
		HTTRACE(CORE_TRACE, "ERROR....... Not a valid time format \"%s\"\n" _ s);
		return 0;
	    }
	    tm.tm_mday = strtol(s, &s, 10);
	    tm.tm_mon = make_month(s, &s);
	    tm.tm_year = strtol(s, &s, 10) - 1900;
	    tm.tm_hour = strtol(s, &s, 10);
	    tm.tm_min = strtol(++s, &s, 10);
	    tm.tm_sec = strtol(++s, &s, 10);
	}
    } else if (isdigit((int) *str)) {

	if (strchr(str, 'T')) {		        /* ISO (limited format) date string */