]> code.citadel.org Git - citadel.git/blobdiff - libcitadel/lib/stringbuf.c
* zero-safe StrtoI/L
[citadel.git] / libcitadel / lib / stringbuf.c
index e28e9f5aff0e39e00cb23f8c0ab5d5c8dbdea482..67a50ece57b97c64f1de39dcb550eedc57d5b4ca 100644 (file)
@@ -78,7 +78,7 @@ static int IncreaseBuf(StrBuf *Buf, int KeepOriginal, int DestSize)
                return -1;
                
        if (DestSize > 0)
-               while (NewSize < DestSize)
+               while (NewSize <= DestSize)
                        NewSize *= 2;
 
        NewBuf= (char*) malloc(NewSize);
@@ -226,6 +226,8 @@ StrBuf* _NewConstStrBuf(const char* StringConstant, size_t SizeOfStrConstant)
  */
 int FlushStrBuf(StrBuf *buf)
 {
+       if (buf == NULL)
+               return -1;
        if (buf->ConstBuf)
                return -1;       
        buf->buf[0] ='\0';
@@ -269,6 +271,8 @@ void HFreeStrBuf (void *VFreeMe)
  */
 long StrTol(const StrBuf *Buf)
 {
+       if (Buf == NULL)
+               return 0;
        if(Buf->BufUsed > 0)
                return atol(Buf->buf);
        else
@@ -280,7 +284,9 @@ long StrTol(const StrBuf *Buf)
  */
 int StrToi(const StrBuf *Buf)
 {
-       if(Buf->BufUsed > 0)
+       if (Buf == NULL)
+               return 0;
+       if (Buf->BufUsed > 0)
                return atoi(Buf->buf);
        else
                return 0;
@@ -339,6 +345,7 @@ void StrBufAppendBuf(StrBuf *Buf, const StrBuf *AppendBuf, size_t Offset)
 void StrBufAppendBufPlain(StrBuf *Buf, const char *AppendBuf, long AppendSize, size_t Offset)
 {
        long aps;
+       long BufSizeRequired;
 
        if ((AppendBuf == NULL) || (Buf == NULL))
                return;
@@ -348,8 +355,9 @@ void StrBufAppendBufPlain(StrBuf *Buf, const char *AppendBuf, long AppendSize, s
        else
                aps = AppendSize - Offset;
 
-       if (Buf->BufSize < Buf->BufUsed + aps)
-               IncreaseBuf(Buf, (Buf->BufUsed > 0), Buf->BufUsed + aps);
+       BufSizeRequired = Buf->BufUsed + aps + 1;
+       if (Buf->BufSize <= BufSizeRequired)
+               IncreaseBuf(Buf, (Buf->BufUsed > 0), BufSizeRequired);
 
        memcpy(Buf->buf + Buf->BufUsed, 
               AppendBuf + Offset, 
@@ -476,7 +484,7 @@ long StrEscAppend(StrBuf *Target, const StrBuf *Source, const char *PlainIn, int
                        bptr += 5;
                        Target->BufUsed += 5;
                }
-               else if (*aptr == '\"') {
+               else if (*aptr == '"') {
                        memcpy(bptr, "&quot;", 6);
                        bptr += 6;
                        Target->BufUsed += 6;
@@ -1139,6 +1147,47 @@ void StrBufCutRight(StrBuf *Buf, int nChars)
        Buf->buf[Buf->BufUsed] = '\0';
 }
 
+/**
+ * \brief Cut the string after n Chars
+ * \param Buf Buffer to modify
+ * \param AfternChars after how many chars should we trunkate the string?
+ * \param At if non-null and points inside of our string, cut it there.
+ */
+void StrBufCutAt(StrBuf *Buf, int AfternChars, const char *At)
+{
+       if (At != NULL){
+               AfternChars = At - Buf->buf;
+       }
+
+       if ((AfternChars < 0) || (AfternChars >= Buf->BufUsed))
+               return;
+       Buf->BufUsed = AfternChars;
+       Buf->buf[Buf->BufUsed] = '\0';
+}
+
+
+/*
+ * Strip leading and trailing spaces from a string; with premeasured and adjusted length.
+ * buf - the string to modify
+ * len - length of the string. 
+ */
+void StrBufTrim(StrBuf *Buf)
+{
+       int delta = 0;
+       if ((Buf == NULL) || (Buf->BufUsed == 0)) return;
+
+       while ((Buf->BufUsed > delta) && (isspace(Buf->buf[delta]))){
+               delta ++;
+       }
+       if (delta > 0) StrBufCutLeft(Buf, delta);
+
+       if (Buf->BufUsed == 0) return;
+       while (isspace(Buf->buf[Buf->BufUsed - 1])){
+               Buf->BufUsed --;
+       }
+       Buf->buf[Buf->BufUsed] = '\0';
+}
+
 
 void StrBufUpCase(StrBuf *Buf) 
 {
@@ -1153,6 +1202,19 @@ void StrBufUpCase(StrBuf *Buf)
 }
 
 
+void StrBufLowerCase(StrBuf *Buf) 
+{
+       char *pch, *pche;
+
+       pch = Buf->buf;
+       pche = pch + Buf->BufUsed;
+       while (pch < pche) {
+               *pch = tolower(*pch);
+               pch ++;
+       }
+}
+
+
 /**
  * \brief unhide special chars hidden to the HTML escaper
  * \param target buffer to put the unescaped string in
@@ -1488,8 +1550,9 @@ void StrBufReplaceChars(StrBuf *buf, char search, char replace)
  * tocode      Target encoding
  * fromcode    Source encoding
  */
-static iconv_t ctdl_iconv_open(const char *tocode, const char *fromcode)
+void  ctdl_iconv_open(const char *tocode, const char *fromcode, void *pic)
 {
+#ifdef HAVE_ICONV
        iconv_t ic = (iconv_t)(-1) ;
        ic = iconv_open(tocode, fromcode);
        if (ic == (iconv_t)(-1) ) {
@@ -1501,13 +1564,13 @@ static iconv_t ctdl_iconv_open(const char *tocode, const char *fromcode)
                        ic = iconv_open(tocode, alias_fromcode);
                }
        }
-       return(ic);
+       *(iconv_t *)pic = ic;
+#endif
 }
 
 
-#ifdef HAVE_ICONV
 
-static inline char *FindNextEnd (StrBuf *Buf, char *bptr)
+static inline char *FindNextEnd (const StrBuf *Buf, char *bptr)
 {
        char * end;
        /* Find the next ?Q? */
@@ -1532,23 +1595,119 @@ static inline char *FindNextEnd (StrBuf *Buf, char *bptr)
 }
 
 
-/*
- * Handle subjects with RFC2047 encoding such as:
- * =?koi8-r?B?78bP0s3Mxc7JxSDXz9rE1dvO2c3JINvB0sHNySDP?=
- */
-void StrBuf_RFC822_to_Utf8(StrBuf **Buf, const StrBuf* DefaultCharset) {
-       StrBuf *TmpBuf, *ConvertBuf, *ConvertBuf2;
-       StrBuf StaticBuf;
-       char *start, *end, *next, *nextend, *ptr;
-       char charset[128];
-       char encoding[16];
-       iconv_t ic = (iconv_t)(-1) ;
+void StrBufConvert(StrBuf *ConvertBuf, StrBuf *TmpBuf, void *pic)
+{
+#ifdef HAVE_ICONV
+       int BufSize;
+       iconv_t ic;
        char *ibuf;                     /**< Buffer of characters to be converted */
        char *obuf;                     /**< Buffer for converted characters */
        size_t ibuflen;                 /**< Length of input buffer */
        size_t obuflen;                 /**< Length of output buffer */
-       char *isav;                     /**< Saved pointer to input buffer */
+
+
+       if (ConvertBuf->BufUsed > TmpBuf->BufSize)
+               IncreaseBuf(TmpBuf, 0, ConvertBuf->BufUsed);
+
+       ic = *(iconv_t*)pic;
+       ibuf = ConvertBuf->buf;
+       ibuflen = ConvertBuf->BufUsed;
+       obuf = TmpBuf->buf;
+       obuflen = TmpBuf->BufSize;
        
+       iconv(ic, &ibuf, &ibuflen, &obuf, &obuflen);
+
+       /* little card game: wheres the red lady? */
+       ibuf = ConvertBuf->buf;
+       BufSize = ConvertBuf->BufSize;
+
+       ConvertBuf->buf = TmpBuf->buf;
+       ConvertBuf->BufSize = TmpBuf->BufSize;
+       ConvertBuf->BufUsed = TmpBuf->BufSize - obuflen;
+       ConvertBuf->buf[ConvertBuf->BufUsed] = '\0';
+       
+       TmpBuf->buf = ibuf;
+       TmpBuf->BufSize = BufSize;
+       TmpBuf->BufUsed = 0;
+       TmpBuf->buf[0] = '\0';
+#endif
+}
+
+
+
+
+inline static void DecodeSegment(StrBuf *Target, 
+                                const StrBuf *DecodeMe, 
+                                char *SegmentStart, 
+                                char *SegmentEnd, 
+                                StrBuf *ConvertBuf,
+                                StrBuf *ConvertBuf2, 
+                                StrBuf *FoundCharset)
+{
+       StrBuf StaticBuf;
+       char charset[128];
+       char encoding[16];
+       iconv_t ic = (iconv_t)(-1);
+
+       /* Now we handle foreign character sets properly encoded
+        * in RFC2047 format.
+        */
+       StaticBuf.buf = SegmentStart;
+       StaticBuf.BufUsed = SegmentEnd - SegmentStart;
+       StaticBuf.BufSize = DecodeMe->BufSize - (SegmentStart - DecodeMe->buf);
+       extract_token(charset, SegmentStart, 1, '?', sizeof charset);
+       if (FoundCharset != NULL) {
+               FlushStrBuf(FoundCharset);
+               StrBufAppendBufPlain(FoundCharset, charset, -1, 0);
+       }
+       extract_token(encoding, SegmentStart, 2, '?', sizeof encoding);
+       StrBufExtract_token(ConvertBuf, &StaticBuf, 3, '?');
+       
+       *encoding = toupper(*encoding);
+       if (*encoding == 'B') { /**< base64 */
+               ConvertBuf2->BufUsed = CtdlDecodeBase64(ConvertBuf2->buf, 
+                                                       ConvertBuf->buf, 
+                                                       ConvertBuf->BufUsed);
+       }
+       else if (*encoding == 'Q') {    /**< quoted-printable */
+               long pos;
+               
+               pos = 0;
+               while (pos < ConvertBuf->BufUsed)
+               {
+                       if (ConvertBuf->buf[pos] == '_') 
+                               ConvertBuf->buf[pos] = ' ';
+                       pos++;
+               }
+               
+               ConvertBuf2->BufUsed = CtdlDecodeQuotedPrintable(
+                       ConvertBuf2->buf, 
+                       ConvertBuf->buf,
+                       ConvertBuf->BufUsed);
+       }
+       else {
+               StrBufAppendBuf(ConvertBuf2, ConvertBuf, 0);
+       }
+
+       ctdl_iconv_open("UTF-8", charset, &ic);
+       if (ic != (iconv_t)(-1) ) {             
+               StrBufConvert(ConvertBuf2, ConvertBuf, &ic);
+               StrBufAppendBuf(Target, ConvertBuf2, 0);
+               iconv_close(ic);
+       }
+       else {
+               StrBufAppendBufPlain(Target, HKEY("(unreadable)"), 0);
+       }
+}
+/*
+ * Handle subjects with RFC2047 encoding such as:
+ * =?koi8-r?B?78bP0s3Mxc7JxSDXz9rE1dvO2c3JINvB0sHNySDP?=
+ */
+void StrBuf_RFC822_to_Utf8(StrBuf *Target, const StrBuf *DecodeMe, const StrBuf* DefaultCharset, StrBuf *FoundCharset)
+{
+       StrBuf *ConvertBuf, *ConvertBuf2;
+       char *start, *end, *next, *nextend, *ptr = NULL;
+       iconv_t ic = (iconv_t)(-1) ;
        const char *eptr;
        int passes = 0;
        int i, len, delta;
@@ -1560,57 +1719,70 @@ void StrBuf_RFC822_to_Utf8(StrBuf **Buf, const StrBuf* DefaultCharset) {
         *  handle it anyway by converting from a user-specified default
         *  charset to UTF-8 if we see any nonprintable characters.
         */
-       TmpBuf = NewStrBufPlain(NULL, StrLength(*Buf));
-
-       len = StrLength(*Buf);
-       for (i=0; i<(*Buf)->BufUsed; ++i) {
-               if (((*Buf)->buf[i] < 32) || ((*Buf)->buf[i] > 126)) {
+       
+       len = StrLength(DecodeMe);
+       for (i=0; i<DecodeMe->BufUsed; ++i) {
+               if ((DecodeMe->buf[i] < 32) || (DecodeMe->buf[i] > 126)) {
                        illegal_non_rfc2047_encoding = 1;
                        break;
                }
        }
 
-       if (illegal_non_rfc2047_encoding) {
-               if ( (strcasecmp(ChrPtr(DefaultCharset), "UTF-8")) && 
-                    (strcasecmp(ChrPtr(DefaultCharset), "us-ascii")) ) {
-                       ic = ctdl_iconv_open("UTF-8", ChrPtr(DefaultCharset));
-                       if (ic != (iconv_t)(-1) ) {
-                               ibuf = (*Buf)->buf;
-                               obuf = TmpBuf->buf;
-                               ibuflen = (*Buf)->BufUsed;
-                               obuflen = TmpBuf->BufSize;
-
-                               iconv(ic, &ibuf, &ibuflen, &obuf, &obuflen);
-                               TmpBuf->BufUsed = TmpBuf->BufSize - obuflen;
-                               TmpBuf->buf[TmpBuf->BufUsed] = '\0';
-
-                               FreeStrBuf(Buf);
-                               *Buf = TmpBuf;
-                               TmpBuf = NewStrBufPlain(NULL, StrLength(*Buf));
-
-                               iconv_close(ic);
-                       }
+       ConvertBuf = NewStrBufPlain(NULL, StrLength(DecodeMe));
+       if ((illegal_non_rfc2047_encoding) &&
+           (strcasecmp(ChrPtr(DefaultCharset), "UTF-8")) && 
+           (strcasecmp(ChrPtr(DefaultCharset), "us-ascii")) )
+       {
+               ctdl_iconv_open("UTF-8", ChrPtr(DefaultCharset), &ic);
+               if (ic != (iconv_t)(-1) ) {
+                       StrBufConvert((StrBuf*)DecodeMe, ConvertBuf, &ic);///TODO: don't void const?
+                       iconv_close(ic);
                }
        }
 
        /* pre evaluate the first pair */
        nextend = end = NULL;
-       len = StrLength(*Buf);
-       start = strstr((*Buf)->buf, "=?");
-       eptr = (*Buf)->buf + (*Buf)->BufUsed;
+       len = StrLength(DecodeMe);
+       start = strstr(DecodeMe->buf, "=?");
+       eptr = DecodeMe->buf + DecodeMe->BufUsed;
        if (start != NULL) 
-               end = FindNextEnd (*Buf, start);
+               end = FindNextEnd (DecodeMe, start);
+       else {
+               StrBufAppendBuf(Target, DecodeMe, 0);
+               FreeStrBuf(&ConvertBuf);
+               return;
+       }
+
+       ConvertBuf2 = NewStrBufPlain(NULL, StrLength(DecodeMe));
 
+       if (start != DecodeMe->buf)
+               StrBufAppendBufPlain(Target, DecodeMe->buf, start - DecodeMe->buf, 0);
+       /*
+        * Since spammers will go to all sorts of absurd lengths to get their
+        * messages through, there are LOTS of corrupt headers out there.
+        * So, prevent a really badly formed RFC2047 header from throwing
+        * this function into an infinite loop.
+        */
        while ((start != NULL) && 
               (end != NULL) && 
               (start < eptr) && 
-              (end < eptr))
+              (end < eptr) && 
+              (passes < 20))
        {
+               passes++;
+               DecodeSegment(Target, 
+                             DecodeMe, 
+                             start, 
+                             end, 
+                             ConvertBuf,
+                             ConvertBuf2,
+                             FoundCharset);
+               
                next = strstr(end, "=?");
                nextend = NULL;
                if ((next != NULL) && 
                    (next < eptr))
-                       nextend = FindNextEnd(*Buf, next);
+                       nextend = FindNextEnd(DecodeMe, next);
                if (nextend == NULL)
                        next = NULL;
 
@@ -1633,9 +1805,9 @@ void StrBuf_RFC822_to_Utf8(StrBuf **Buf, const StrBuf* DefaultCharset) {
                                         len - (next - start));
                                
                                /* now terminate the gab at the end */
-                               delta = (next - end) - 2;
-                               (*Buf)->BufUsed -= delta;
-                               (*Buf)->buf[(*Buf)->BufUsed] = '\0';
+                               delta = (next - end) - 2; ////TODO: const! 
+                               ((StrBuf*)DecodeMe)->BufUsed -= delta;
+                               ((StrBuf*)DecodeMe)->buf[DecodeMe->BufUsed] = '\0';
 
                                /* move next to its new location. */
                                next -= delta;
@@ -1643,92 +1815,86 @@ void StrBuf_RFC822_to_Utf8(StrBuf **Buf, const StrBuf* DefaultCharset) {
                        }
                }
                /* our next-pair is our new first pair now. */
+               ptr = end + 2;
                start = next;
                end = nextend;
        }
+       end = ptr;
+       nextend = DecodeMe->buf + DecodeMe->BufUsed;
+       if ((end != NULL) && (end < nextend)) {
+               ptr = end;
+               while ( (ptr < nextend) &&
+                       (isspace(*ptr) ||
+                        (*ptr == '\r') ||
+                        (*ptr == '\n') || 
+                        (*ptr == '\t')))
+                       ptr ++;
+               if (ptr < nextend)
+                       StrBufAppendBufPlain(Target, end, nextend - end, 0);
+       }
+       FreeStrBuf(&ConvertBuf);
+       FreeStrBuf(&ConvertBuf2);
+}
 
-       ConvertBuf = NewStrBufPlain(NULL, StrLength(*Buf));
-       ConvertBuf2 = NewStrBufPlain(NULL, StrLength(*Buf));
-       /* Now we handle foreign character sets properly encoded
-        * in RFC2047 format.
-        */
-       while (start=strstr((*Buf)->buf, "=?"), 
-              end=FindNextEnd((*Buf), ((start != NULL)? start : (*Buf)->buf)),
-                              ((start != NULL) && 
-                               (end != NULL) && 
-                               (end > start)) )
-       {
-               StaticBuf.buf = start;
-               StaticBuf.BufUsed = (*Buf)->BufUsed - ((*Buf)->buf - start);
-               StaticBuf.BufSize = (*Buf)->BufSize - ((*Buf)->buf - start);
-               extract_token(charset, start, 1, '?', sizeof charset);
-               extract_token(encoding, start, 2, '?', sizeof encoding);
-               StrBufExtract_token(ConvertBuf, &StaticBuf, 3, '?');
-
-               if (!strcasecmp(encoding, "B")) {       /**< base64 */
-                       ConvertBuf2->BufUsed = CtdlDecodeBase64(ConvertBuf2->buf, 
-                                                               ConvertBuf->buf, 
-                                                               ConvertBuf->BufUsed);
-               }
-               else if (!strcasecmp(encoding, "Q")) {  /**< quoted-printable */
-                       long pos;
-                       
-                       pos = 0;
-                       while (pos < ConvertBuf->BufUsed)
-                       {
-                               if (ConvertBuf->buf[pos] == '_') 
-                                       ConvertBuf->buf[pos] = ' ';
-                               pos++;
-                       }
 
-                       ConvertBuf2->BufUsed = CtdlDecodeQuotedPrintable(
-                               ConvertBuf2->buf, 
-                               ConvertBuf->buf,
-                               ConvertBuf->BufUsed);
-               }
-               else {
-                       StrBufAppendBuf(ConvertBuf2, ConvertBuf, 0);
-               }
 
-               ic = ctdl_iconv_open("UTF-8", charset);
-               if (ic != (iconv_t)(-1) ) {
-                       ibuf = ConvertBuf2->buf;
-                       obuf = ConvertBuf->buf;
-                       ibuf = ConvertBuf2->buf;
-                       obuflen = ConvertBuf->BufSize;
-                       ibuflen = ConvertBuf2->BufUsed;
+long StrBuf_Utf8StrLen(StrBuf *Buf)
+{
+       return Ctdl_Utf8StrLen(Buf->buf);
+}
 
-                       iconv(ic, &ibuf, &ibuflen, &obuf, &obuflen);
-                       ConvertBuf->BufUsed = ConvertBuf->BufSize - obuflen;
-                       ConvertBuf->buf[ConvertBuf->BufUsed] = '\0';
+long StrBuf_Utf8StrCut(StrBuf *Buf, int maxlen)
+{
+       char *CutAt;
 
-                       StrBufAppendBuf(TmpBuf, ConvertBuf, 0);
-                       iconv_close(ic);
-               }
-               else {
+       CutAt = Ctdl_Utf8StrCut(Buf->buf, maxlen);
+       if (CutAt != NULL) {
+               Buf->BufUsed = CutAt - Buf->buf;
+               Buf->buf[Buf->BufUsed] = '\0';
+       }
+       return Buf->BufUsed;    
+}
 
-                       StrBufAppendBufPlain(TmpBuf, HKEY("(unreadable)"), 0);
-               }
 
-               free(isav);
-
-               /*
-                * Since spammers will go to all sorts of absurd lengths to get their
-                * messages through, there are LOTS of corrupt headers out there.
-                * So, prevent a really badly formed RFC2047 header from throwing
-                * this function into an infinite loop.
-                */
-               ++passes;
-               if (passes > 20)  { 
-                       FreeStrBuf(Buf);
-                       *Buf = TmpBuf;
-                       return;
+
+int StrBufSipLine(StrBuf *LineBuf, StrBuf *Buf, const char **Ptr)
+{
+       const char *aptr, *ptr, *eptr;
+       char *optr, *xptr;
+
+       if (Buf == NULL)
+               return 0;
+
+       if (*Ptr==NULL)
+               ptr = aptr = Buf->buf;
+       else
+               ptr = aptr = *Ptr;
+
+       optr = LineBuf->buf;
+       eptr = Buf->buf + Buf->BufUsed;
+       xptr = LineBuf->buf + LineBuf->BufSize;
+
+       while ((*ptr != '\n') &&
+              (*ptr != '\r') &&
+              (ptr < eptr))
+       {
+               *optr = *ptr;
+               optr++; ptr++;
+               if (optr == xptr) {
+                       LineBuf->BufUsed = optr - LineBuf->buf;
+                       IncreaseBuf(LineBuf,  1, LineBuf->BufUsed + 1);
+                       optr = LineBuf->buf + LineBuf->BufUsed;
+                       xptr = LineBuf->buf + LineBuf->BufSize;
                }
        }
-       FreeStrBuf(Buf);
-       *Buf = TmpBuf;
-}
-#else
-void StrBuf_RFC822_to_Utf8(StrBuf **Buf, const StrBuf* DefaultCharset) {};
+       LineBuf->BufUsed = optr - LineBuf->buf;
+       *optr = '\0';       
+       if (*ptr == '\r')
+               ptr ++;
+       if (*ptr == '\n')
+               ptr ++;
 
-#endif
+       *Ptr = ptr;
+
+       return Buf->BufUsed - (ptr - Buf->buf);
+}