X-Git-Url: https://code.citadel.org/?a=blobdiff_plain;f=libcitadel%2Flib%2Ftools.c;h=df2f1b42d1086e434a3c254ebca21fe217a3047a;hb=1aa2da0249792527f99172681aecc77e0ad086af;hp=60168fe91e0ca0cb3cd8c21cc38d775623b43515;hpb=0b4754a432708fec0328414fe4adff19c70c3855;p=citadel.git diff --git a/libcitadel/lib/tools.c b/libcitadel/lib/tools.c index 60168fe91..df2f1b42d 100644 --- a/libcitadel/lib/tools.c +++ b/libcitadel/lib/tools.c @@ -1,6 +1,22 @@ /* * A basic toolset containing miscellaneous functions for string manipluation, * encoding/decoding, and a bunch of other stuff. + * + * Copyright (c) 1987-2011 by the citadel.org team + * + * This program is open source software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 3 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ @@ -371,7 +387,6 @@ size_t CtdlEncodeBase64(char *dest, const char *source, size_t sourcelen, int li dest[dpos++] = '\r'; dest[dpos++] = '\n'; dest[dpos] = 0; - thisline = 0; } return(dpos); @@ -430,9 +445,9 @@ int CtdlDecodeBase64(char *dest, const char *source, size_t length) /* * if we send out non ascii subjects, we encode it this way. */ -char *rfc2047encode(char *line, long length) +char *rfc2047encode(const char *line, long length) { - char *AlreadyEncoded; + const char *AlreadyEncoded; char *result; long end; #define UTF8_HEADER "=?UTF-8?B?" @@ -456,39 +471,77 @@ char *rfc2047encode(char *line, long length) return result; } - /* - * Strip leading and trailing spaces from a string + * removes double slashes from pathnames + * allows / disallows trailing slashes */ -void striplt(char *buf) +void StripSlashes(char *Dir, int TrailingSlash) { - size_t len; - int a; + char *a, *b; + + a = b = Dir; + + while (!IsEmptyStr(a)) { + if (*a == '/') { + while (*a == '/') + a++; + *b = '/'; + b++; + } + else { + *b = *a; + b++; a++; + } + } + if ((TrailingSlash) && (*(b - 1) != '/')){ + *b = '/'; + b++; + } + *b = '\0'; - if (buf==NULL) return; - if (IsEmptyStr(buf)) return; - len = strlen(buf); - while ((!IsEmptyStr(buf)) && (isspace(buf[len - 1]))) - buf[--len] = 0; - if (IsEmptyStr(buf)) return; - a = 0; - while ((!IsEmptyStr(buf)) && (isspace(buf[a]))) - a++; - if (a > 0) - memmove(buf, &buf[a], len - a + 1); } +/* + * Strip leading and trailing spaces from a string + */ +size_t striplt(char *buf) { + char *first_nonspace = NULL; + char *last_nonspace = NULL; + char *ptr; + size_t new_len = 0; + + if ((buf == NULL) || (*buf == '\0')) { + return 0; + } + + for (ptr=buf; *ptr!=0; ++ptr) { + if (!isspace(*ptr)) { + if (!first_nonspace) { + first_nonspace = ptr; + } + last_nonspace = ptr; + } + } + if ((!first_nonspace) || (!last_nonspace)) { + buf[0] = 0; + return 0; + } + new_len = last_nonspace - first_nonspace + 1; + memmove(buf, first_nonspace, new_len); + buf[new_len] = 0; + return new_len; +} /** * \brief check for the presence of a character within a string (returns count) * \param st the string to examine * \param ch the char to search - * \return the position inside of st + * \return the number of times ch appears in st */ -int haschar(const char *st,int ch) +int haschar(const char *st, int ch) { const char *ptr; int b; @@ -513,40 +566,19 @@ int haschar(const char *st,int ch) */ void fmt_date(char *buf, size_t n, time_t thetime, int seconds) { struct tm tm; - int hour; + char *teh_format = NULL; - /* Month strings for date conversions ... this needs to be localized eventually */ - char *fmt_date_months[12] = { - "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec" - }; - - strcpy(buf, ""); + *buf = '\0'; localtime_r(&thetime, &tm); - hour = tm.tm_hour; - if (hour == 0) hour = 12; - else if (hour > 12) hour = hour - 12; - if (seconds) { - snprintf(buf, n, "%s %d %4d %d:%02d:%02d%s", - fmt_date_months[tm.tm_mon], - tm.tm_mday, - tm.tm_year + 1900, - hour, - tm.tm_min, - tm.tm_sec, - ( (tm.tm_hour >= 12) ? "pm" : "am" ) - ); - } else { - snprintf(buf, n, "%s %d %4d %d:%02d%s", - fmt_date_months[tm.tm_mon], - tm.tm_mday, - tm.tm_year + 1900, - hour, - tm.tm_min, - ( (tm.tm_hour >= 12) ? "pm" : "am" ) - ); + teh_format = "%F %R:%S"; } + else { + teh_format = "%F %R"; + } + + strftime(buf, n, teh_format, &tm); } @@ -555,7 +587,7 @@ void fmt_date(char *buf, size_t n, time_t thetime, int seconds) { * Determine whether the specified message number is contained within the * specified sequence set. */ -int is_msg_in_sequence_set(char *mset, long msgnum) { +int is_msg_in_sequence_set(const char *mset, long msgnum) { int num_sets; int s; char setstr[128], lostr[128], histr[128]; @@ -642,13 +674,71 @@ char *memreadlinelen(char *start, char *buf, int maxlen, int *retlen) } +/** + * \brief Utility function to "readline" from memory + * \param start Location in memory from which we are reading. + * \param buf the buffer to place the string in. + * \param maxlen Size of string buffer + * \return Pointer to the source memory right after we stopped reading. + */ +const char *cmemreadline(const char *start, char *buf, int maxlen) +{ + char ch; + const char *ptr; + int len = 0; /**< tally our own length to avoid strlen() delays */ + + ptr = start; + + while (1) { + ch = *ptr++; + if ((len + 1 < (maxlen)) && (ch != 13) && (ch != 10)) { + buf[len++] = ch; + } + if ((ch == 10) || (ch == 0)) { + buf[len] = 0; + return ptr; + } + } +} + + +/** + * \brief Utility function to "readline" from memory + * \param start Location in memory from which we are reading. + * \param buf the buffer to place the string in. + * \param maxlen Size of string buffer + * \param retlen the length of the returned string + * \return Pointer to the source memory right after we stopped reading. + */ +const char *cmemreadlinelen(const char *start, char *buf, int maxlen, int *retlen) +{ + char ch; + const char *ptr; + int len = 0; /**< tally our own length to avoid strlen() delays */ + + ptr = start; + + while (1) { + ch = *ptr++; + if ((len + 1 < (maxlen)) && (ch != 13) && (ch != 10)) { + buf[len++] = ch; + } + if ((ch == 10) || (ch == 0)) { + buf[len] = 0; + *retlen = len; + return ptr; + } + } +} + + /* * Strip a boundarized substring out of a string (for example, remove * parentheses and anything inside them). */ -void stripout(char *str, char leftboundary, char rightboundary) { +int stripout(char *str, char leftboundary, char rightboundary) { int a; int lb = (-1); int rb = (-1); @@ -660,12 +750,14 @@ void stripout(char *str, char leftboundary, char rightboundary) { if ( (lb > 0) && (rb > lb) ) { strcpy(&str[lb - 1], &str[rb + 1]); + return 1; } else if ( (lb == 0) && (rb > lb) ) { strcpy(str, &str[rb + 1]); + return 1; } - + return 0; } @@ -673,19 +765,30 @@ void stripout(char *str, char leftboundary, char rightboundary) { * Reduce a string down to a boundarized substring (for example, remove * parentheses and anything outside them). */ -void stripallbut(char *str, char leftboundary, char rightboundary) { - int a; - - for (a = 0; a < strlen(str); ++ a) { - if (str[a] == leftboundary) strcpy(str, &str[a+1]); - } - - for (a = 0; a < strlen(str); ++ a) { - if (str[a] == rightboundary) str[a] = 0; +long stripallbut(char *str, char leftboundary, char rightboundary) { + long len = 0; + + char *lb = NULL; + char *rb = NULL; + + lb = strrchr(str, leftboundary); + if (lb != NULL) { + ++lb; + rb = strchr(str, rightboundary); + if ((rb != NULL) && (rb >= lb)) { + *rb = 0; + fflush(stderr); + len = (long)rb - (long)lb; + memmove(str, lb, len); + str[len] = 0; + return(len); + } } + return (long)strlen(str); } + char *myfgets(char *s, int size, FILE *stream) { char *ret = fgets(s, size, stream); char *nl; @@ -711,7 +814,7 @@ void urlesc(char *outbuf, size_t oblen, char *strbuf) int a, b, c, len, eclen, olen; char *ec = " +#&;`'|*?-~<>^()[]{}/$\"\\"; - strcpy(outbuf, ""); + *outbuf = '\0'; len = strlen(strbuf); eclen = strlen(ec); olen = 0; @@ -749,12 +852,41 @@ char *strcpy(char *dest, const char *src) { * Generate a new, globally unique UID parameter for a calendar etc. object */ void generate_uuid(char *buf) { - static int seq = 0; + static int seq = (-1); + static int no_kernel_uuid = 0; + + /* If we are running on Linux then we have a kernelspace uuid generator available */ + + if (no_kernel_uuid == 0) { + FILE *fp; + fp = fopen("/proc/sys/kernel/random/uuid", "rb"); + if (fp) { + int rv; + rv = fread(buf, 36, 1, fp); + fclose(fp); + if (rv == 1) { + buf[36] = 0; + return; + } + } + } - sprintf(buf, "%lx-%lx-%x", - time(NULL), + /* If the kernel didn't provide us with a uuid, we generate a pseudo-random one */ + + no_kernel_uuid = 1; + + if (seq == (-1)) { + seq = (int)rand(); + } + ++seq; + seq = (seq % 0x0FFF) ; + + sprintf(buf, "%08lx-%04lx-4%03x-a%03x-%012lx", + (long)time(NULL), (long)getpid(), - (seq++) + seq, + seq, + (long)rand() ); } @@ -765,12 +897,65 @@ void generate_uuid(char *buf) { * The code is roughly based on the strstr() replacement from 'tin' written * by Urs Jannsen. */ -char *bmstrcasestr(char *text, char *pattern) { +inline static char *_bmstrcasestr_len(char *text, size_t textlen, const char *pattern, size_t patlen) { register unsigned char *p, *t; register int i, j, *delta; register size_t p1; int deltaspace[256]; + + if (!text) return(NULL); + if (!pattern) return(NULL); + + /* algorithm fails if pattern is empty */ + if ((p1 = patlen) == 0) + return (text); + + /* code below fails (whenever i is unsigned) if pattern too long */ + if (p1 > textlen) + return (NULL); + + /* set up deltas */ + delta = deltaspace; + for (i = 0; i <= 255; i++) + delta[i] = p1; + for (p = (unsigned char *) pattern, i = p1; --i > 0;) + delta[tolower(*p++)] = i; + + /* + * From now on, we want patlen - 1. + * In the loop below, p points to the end of the pattern, + * t points to the end of the text to be tested against the + * pattern, and i counts the amount of text remaining, not + * including the part to be tested. + */ + p1--; + p = (unsigned char *) pattern + p1; + t = (unsigned char *) text + p1; + i = textlen - patlen; + while(1) { + if (tolower(p[0]) == tolower(t[0])) { + if (strncasecmp ((const char *)(p - p1), (const char *)(t - p1), p1) == 0) { + return ((char *)t - p1); + } + } + j = delta[tolower(t[0])]; + if (i < j) + break; + i -= j; + t += j; + } + return (NULL); +} + +/* + * bmstrcasestr() -- case-insensitive substring search + * + * This uses the Boyer-Moore search algorithm and is therefore quite fast. + * The code is roughly based on the strstr() replacement from 'tin' written + * by Urs Jannsen. + */ +char *bmstrcasestr(char *text, const char *pattern) { size_t textlen; size_t patlen; @@ -780,6 +965,33 @@ char *bmstrcasestr(char *text, char *pattern) { textlen = strlen (text); patlen = strlen (pattern); + return _bmstrcasestr_len(text, textlen, pattern, patlen); +} + +char *bmstrcasestr_len(char *text, size_t textlen, const char *pattern, size_t patlen) { + return _bmstrcasestr_len(text, textlen, pattern, patlen); +} + + + + +/* + * bmstrcasestr() -- case-insensitive substring search + * + * This uses the Boyer-Moore search algorithm and is therefore quite fast. + * The code is roughly based on the strstr() replacement from 'tin' written + * by Urs Jannsen. + */ +inline static const char *_cbmstrcasestr_len(const char *text, size_t textlen, const char *pattern, size_t patlen) { + + register unsigned char *p, *t; + register int i, j, *delta; + register size_t p1; + int deltaspace[256]; + + if (!text) return(NULL); + if (!pattern) return(NULL); + /* algorithm fails if pattern is empty */ if ((p1 = patlen) == 0) return (text); @@ -821,7 +1033,29 @@ char *bmstrcasestr(char *text, char *pattern) { return (NULL); } +/* + * bmstrcasestr() -- case-insensitive substring search + * + * This uses the Boyer-Moore search algorithm and is therefore quite fast. + * The code is roughly based on the strstr() replacement from 'tin' written + * by Urs Jannsen. + */ +const char *cbmstrcasestr(const char *text, const char *pattern) { + size_t textlen; + size_t patlen; + + if (!text) return(NULL); + if (!pattern) return(NULL); + + textlen = strlen (text); + patlen = strlen (pattern); + return _cbmstrcasestr_len(text, textlen, pattern, patlen); +} + +const char *cbmstrcasestr_len(const char *text, size_t textlen, const char *pattern, size_t patlen) { + return _cbmstrcasestr_len(text, textlen, pattern, patlen); +} /* * Local replacement for controversial C library function that generates @@ -847,7 +1081,7 @@ void CtdlMakeTempFileName(char *name, int len) { * Determine whether the specified message number is contained within the specified set. * Returns nonzero if the specified message number is in the specified message set string. */ -int is_msg_in_mset(char *mset, long msgnum) { +int is_msg_in_mset(const char *mset, long msgnum) { int num_sets; int s; char setstr[SIZ], lostr[SIZ], histr[SIZ]; /* was 1024 */ @@ -881,10 +1115,8 @@ int is_msg_in_mset(char *mset, long msgnum) { /* - * \brief searches for a paternn within asearch string - * \param search the string to search - * \param patn the pattern to find in string - * \returns position in string + * searches for a pattern within a search string + * returns position in string */ int pattern2(char *search, char *patn) { @@ -921,97 +1153,6 @@ void stripltlen(char *buf, int *len) } } -/** - * \brief detect whether this char starts an utf-8 encoded char - * \param Char character to inspect - * \returns yes or no - */ -inline int Ctdl_IsUtf8SequenceStart(char Char) -{ -/** 11??.???? indicates an UTF8 Sequence. */ - return ((Char & 0xC0) != 0); -} - -/** - * \brief evaluate the length of an utf8 special character sequence - * \param Char the character to examine - * \returns width of utf8 chars in bytes - */ -inline int Ctdl_GetUtf8SequenceLength(char Char) -{ - int n = 1; - char test = (1<<7); - - while ((n < 8) && ((test & Char) != 0)) { - test = test << 1; - n ++; - } - if (n > 6) - n = 1; - return n; -} - -/** - * \brief measure the number of glyphs in an UTF8 string... - * \param str string to measure - * \returns the length of str - */ -int Ctdl_Utf8StrLen(char *str) -{ - int n = 0; - int m = 0; - char *aptr; - - if (str == NULL) - return n; - aptr = str; - while (*aptr != '\0') { - if (Ctdl_IsUtf8SequenceStart(*aptr)){ - m = Ctdl_GetUtf8SequenceLength(*aptr); - while ((m-- > 0) && (*aptr++ != '\0')) - n ++; - } - else { - n++; - aptr++; - } - - } - return n; -} - -/** - * \brief cuts a string after maxlen glyphs - * \param str string to cut to maxlen glyphs - * \param maxlen how long may the string become? - * \returns pointer to maxlen or the end of the string - */ -char *Ctdl_Utf8StrCut(char *str, int maxlen) -{ - int n = 0, m = 0; - char *aptr; - - if (str == NULL) - return NULL; - aptr = str; - while (*aptr != '\0') { - if (Ctdl_IsUtf8SequenceStart(*aptr)){ - m = Ctdl_GetUtf8SequenceLength(*aptr); - while ((m-- > 0) && (*aptr++ != '\0')) - n ++; - } - else { - n++; - aptr++; - } - if (n > maxlen) { - *aptr = '\0'; - return aptr; - } - } - return aptr; -} - /* * Convert all whitespace characters in a supplied string to underscores @@ -1032,59 +1173,20 @@ void convert_spaces_to_underscores(char *str) } -/* - * Parse a URL into host, port number, and resource identifier. - * (This is used by various functions which might need to fetch web pages.) +/* + * check whether the provided string needs to be qp encoded or not */ -int parse_url(char *url, char *hostname, int *port, char *identifier) +int CheckEncode(const char *pch, long len, const char *pche) { - char protocol[1024]; - char scratch[1024]; - char *ptr = NULL; - char *nptr = NULL; - - strcpy(scratch, url); - ptr = (char *)strchr(scratch, ':'); - if (!ptr) { - return(1); /* no protocol specified */ - } - - strcpy(ptr, ""); - strcpy(protocol, scratch); - if (strcmp(protocol, "http")) { - return(2); /* not HTTP */ - } - - strcpy(scratch, url); - ptr = (char *) strstr(scratch, "//"); - if (!ptr) { - return(3); /* no server specified */ - } - ptr += 2; - - strcpy(hostname, ptr); - nptr = (char *)strchr(ptr, ':'); - if (!nptr) { - *port = 80; /* default */ - nptr = (char *)strchr(hostname, '/'); - } - else { - sscanf(nptr, ":%d", port); - nptr = (char *)strchr(hostname, ':'); - } - - if (nptr) { - *nptr = '\0'; - } - - nptr = (char *)strchr(ptr, '/'); - - if (!nptr) { - return(4); /* no url specified */ + if (pche == NULL) + pche = pch + len; + while (pch < pche) { + if (((unsigned char) *pch < 32) || + ((unsigned char) *pch > 126)) { + return 1; + } + pch++; } - - strcpy(identifier, nptr); - return(0); + return 0; } -