X-Git-Url: http://git.tuebingen.mpg.de/?p=paraslash.git;a=blobdiff_plain;f=base64.c;h=122465f8feea79cf06d09bfd759d2446ee0f3f12;hp=e9892244d62a0e00a320003bb3fe740e5e89f866;hb=e7f035b3ad7ecc64ec87d9ce890f89365dec1411;hpb=3ea5399aeec092dcdb9f039288f1e146aadbb2f7

diff --git a/base64.c b/base64.c
index e9892244..122465f8 100644
--- a/base64.c
+++ b/base64.c
@@ -13,117 +13,146 @@
 #include "base64.h"
 #include "string.h"
 
-static const char Base64[] =
-	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
-static const char Pad64 = '=';
-
+static const unsigned char base64_tab[256] = {
+	255, 255, 255, 255, 255, 255, 255, 255, /* 00-07 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 08-0f */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 10-17 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 18-1f */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 20-2f */
+	255, 255, 255,  62, 255, 255, 255,  63, /* 28-2f */
+	52 ,  53,  54,  55,  56,  57,  58,  59, /* 30-37 */
+	60 ,  61, 255, 255, 255, 255, 255, 255, /* 38-3f */
+	255,   0,   1,   2,   3,   4,   5,   6, /* 40-47 */
+	7  ,   8,   9,  10,  11,  12,  13,  14, /* 48-4f */
+	15 ,  16,  17,  18,  19,  20,  21,  22, /* 50-57 */
+	23 ,  24,  25, 255, 255, 255, 255, 255, /* 58-5f */
+	255,  26,  27,  28,  29,  30,  31,  32, /* 60-6f */
+	33 ,  34,  35,  36,  37,  38,  39,  40, /* 68-6f */
+	41 ,  42,  43,  44,  45,  46,  47,  48, /* 70-77 */
+	49 ,  50,  51, 255, 255, 255, 255, 255, /* 78-7f */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 80-87 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 88-8f */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 90-97 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* 98-9f */
+	255, 255, 255, 255, 255, 255, 255, 255, /* a0-a7 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* a8-af */
+	255, 255, 255, 255, 255, 255, 255, 255, /* b0-b7 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* b8-bf */
+	255, 255, 255, 255, 255, 255, 255, 255, /* c0-c7 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* c8-cf */
+	255, 255, 255, 255, 255, 255, 255, 255, /* d0-d7 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* d8-df */
+	255, 255, 255, 255, 255, 255, 255, 255, /* e0-e7 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* e8-ef */
+	255, 255, 255, 255, 255, 255, 255, 255, /* f0-f7 */
+	255, 255, 255, 255, 255, 255, 255, 255, /* f8-ff */
+};
+
+/** Maximal possible size of the decoded data. */
+#define BASE64_MAX_DECODED_SIZE(_encoded_size) ((_encoded_size) / 4 * 3)
+
+#define PAD64 '='
 /**
  * base64-decode a buffer.
  *
  * \param src The buffer to decode.
- * \param target Result is stored here.
- * \param targsize Number of bytes of \a target.
+ * \param encoded_size The special value -1 means: look for terminating zero byte.
+ * \param result Points to dynamically allocated target buffer on success.
+ * \param decoded_size Number of bytes written to \a result.
  *
  * Skips all whitespace anywhere. Converts characters, four at a time, starting
  * at (or after) src from base - 64 numbers into three 8 bit bytes in the
  * target area.
  *
- * \return The number of data bytes stored at the target, -E_BASE64 on errors.
+ * It is OK to pass a \p NULL pointer as \a decoded_size. The result is
+ * terminated with a zero byte.
+ *
+ * \return Standard. The contents of result \a and \a decoded_size are
+ * undefined on failure.
  */
-int base64_decode(char const *src, unsigned char *target, size_t targsize)
+int base64_decode(char const *src, size_t encoded_size, char **result,
+		size_t *decoded_size)
 {
-	unsigned int tarindex, state;
-	int ch;
-	char *pos;
-
-	state = 0;
-	tarindex = 0;
-
-	while ((ch = *src++) != '\0') {
-		if (para_isspace(ch)) /* Skip whitespace anywhere. */
+	size_t i, j, state; /* source/target indices */
+	const char *end = src + encoded_size, *p;
+	unsigned char *target, uch;
+
+	if (encoded_size == (size_t)-1)
+		encoded_size = strlen(src);
+	target = para_malloc(BASE64_MAX_DECODED_SIZE(encoded_size) + 1);
+
+	for (
+		i = 0, j = 0, state = 0;
+		i < encoded_size && (uch = src[i]) != '\0';
+		i++
+	) {
+		if (para_isspace(uch)) /* Skip whitespace anywhere. */
 			continue;
-
-		if (ch == Pad64)
+		if (uch == PAD64)
 			break;
-
-		pos = strchr(Base64, ch);
-		if (pos == NULL) /* A non-base64 character. */
-			return -E_BASE64;
-
+		if (base64_tab[uch] == 255) /* A non-base64 character. */
+			goto fail;
+		uch = base64_tab[uch];
 		switch (state) {
 		case 0:
-			if (tarindex >= targsize)
-				return -E_BASE64;
-			target[tarindex] = (pos - Base64) << 2;
-			state = 1;
+			target[j] = uch << 2;
 			break;
 		case 1:
-			if (tarindex + 1 >= targsize)
-				return -E_BASE64;
-			target[tarindex] |= (pos - Base64) >> 4;
-			target[tarindex + 1] = ((pos - Base64) & 0x0f) << 4;
-			tarindex++;
-			state = 2;
+			target[j] |= uch >> 4;
+			j++;
+			target[j] = (uch & 0x0f) << 4;
 			break;
 		case 2:
-			if (tarindex + 1 >= targsize)
-				return -E_BASE64;
-			target[tarindex] |= (pos - Base64) >> 2;
-			target[tarindex + 1] = ((pos - Base64) & 0x03) << 6;
-			tarindex++;
-			state = 3;
+			target[j] |= uch >> 2;
+			j++;
+			target[j] = (uch & 0x03) << 6;
 			break;
 		case 3:
-			if (tarindex >= targsize)
-				return -E_BASE64;
-			target[tarindex] |= pos - Base64;
-			tarindex++;
-			state = 0;
+			target[j] |= uch;
+			j++;
 			break;
 		}
+		state = (state + 1) % 4;
 	}
-
+	p = (i < encoded_size)? src + i : NULL;
 	/*
 	 * We are done decoding Base-64 chars.  Let's see if we ended
 	 * on a byte boundary, and/or with erroneous trailing characters.
 	 */
-
-	if (ch == Pad64) {		/* We got a pad char. */
-		ch = *src++;		/* Skip it, get next. */
+	if (p && *p == PAD64) { /* We got a pad char. Skip it, get next. */
+		p++;
 		switch (state) {
-		case 0:		/* Invalid = in first position */
-		case 1:		/* Invalid = in second position */
-			return -E_BASE64;
+		case 0: /* Invalid = in first position */
+		case 1: /* Invalid = in second position */
+			goto fail;
 
-		case 2:		/* Valid, means one byte of info */
+		case 2: /* Valid, means one byte of info */
 			/* Skip any number of spaces. */
-			for (; ch != '\0'; ch = *src++)
-				if (!isspace(ch))
+			for (; p < end && *p != '\0'; p++)
+				if (!para_isspace(*p))
 					break;
 			/* Make sure there is another trailing = sign. */
-			if (ch != Pad64)
-				return -E_BASE64;
-			ch = *src++;		/* Skip the = */
+			if (*p != PAD64)
+				goto fail;
 			/* Fall through to "single trailing =" case. */
-			/* FALLTHROUGH */
+			p++;
 
-		case 3:		/* Valid, means two bytes of info */
+		case 3: /* Valid, means two bytes of info */
 			/*
 			 * We know this char is an =.  Is there anything but
 			 * whitespace after it?
 			 */
-			for (; ch != '\0'; ch = *src++)
-				if (!isspace(ch))
-					return -E_BASE64;
-
+			for (; p < end && *p != '\0'; p++)
+				if (!para_isspace(*p))
+					goto fail;
 			/*
 			 * Now make sure for cases 2 and 3 that the "extra"
 			 * bits that slopped past the last full byte were
 			 * zeros.  If we don't check them, they become a
 			 * subliminal channel.
 			 */
-			if (target[tarindex] != 0)
-				return -E_BASE64;
+			if (target[j] != 0)
+				goto fail;
 		}
 	} else {
 		/*
@@ -131,37 +160,44 @@ int base64_decode(char const *src, unsigned char *target, size_t targsize)
 		 * have no partial bytes lying around.
 		 */
 		if (state != 0)
-			return -E_BASE64;
+			goto fail;
 	}
-	return tarindex;
+	/* success */
+	target[j] = '\0'; /* just to be sure */
+	if (decoded_size)
+		*decoded_size = j;
+	*result = (char *)target;
+	return 1;
+fail:
+	free(target);
+	return -E_BASE64;
 }
 
 /**
- * uudecode a buffer.
+ * Decode a buffer using the uuencode Base64 algorithm.
  *
  * \param src The buffer to decode.
- * \param target Result buffer.
- * \param targsize The length of \a target in bytes.
+ * \param encoded_size Number of input bytes in the source buffer.
+ * \param result Contains the decoded data on success.
+ * \param decoded_size Number of output bytes on success.
+ *
+ * This is just a simple wrapper for \ref base64_decode() which strips
+ * whitespace.
  *
- * This is just a simple wrapper for base64_decode() which strips whitespace.
+ * \return The return value of the underlying call to \ref base64_decode().
  *
- * \return The return value of the underlying call to base64_decode().
+ * \sa uuencode(1), uudecode(1).
  */
-int uudecode(const char *src, unsigned char *target, size_t targsize)
+int uudecode(char const *src, size_t encoded_size, char **result,
+		size_t *decoded_size)
 {
-	int len;
-	char *encoded, *p;
+	const char *end = src + encoded_size, *p;
 
-	/* copy the 'readonly' source */
-	encoded = para_strdup(src);
 	/* skip whitespace and data */
-	for (p = encoded; *p == ' ' || *p == '\t'; p++)
+	for (p = src; p < end && (*p == ' ' || *p == '\t'); p++)
 		;
-	for (; *p != '\0' && *p != ' ' && *p != '\t'; p++)
+	for (; p < end && *p != '\0' && *p != ' ' && *p != '\t'; p++)
 		;
 	/* and remove trailing whitespace because base64_decode needs this */
-	*p = '\0';
-	len = base64_decode(encoded, target, targsize);
-	free(encoded);
-	return len;
+	return base64_decode(src, p - src, result, decoded_size);
 }