Add serd_node_new_blob and serd_base64_decode for handling arbitrary binary

data via base64 encoding.


git-svn-id: http://svn.drobilla.net/serd/trunk@280 490d8e77-9747-427b-9fa3-0b8f29cee8a0

1 files changed, 46 insertions, 0 deletions

diff --git a/src/string.c b/src/string.c
index 968a20e1..a97c6f98 100644
--- a/src/string.c
+++ b/src/string.c
@@ -112,3 +112,49 @@ serd_strtod(const char* str, char** endptr)
 	*endptr = (char*)s;
 	return result * sign;
 }
+
+/**
+   Base64 decoding table.
+   This is indexed by encoded characters and returns the numeric value used
+   for decoding, shifted up by 47 to be in the range of printable ASCII.
+   A '$' is a placeholder for characters not in the base64 alphabet.
+*/
+static const char b64_unmap[255] =
+	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$m$$$ncdefghijkl$$$$$$"
+	"$/0123456789:;<=>?@ABCDEFGH$$$$$$IJKLMNOPQRSTUVWXYZ[\\]^_`ab$$$$"
+	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$"
+	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$";
+
+static inline uint8_t unmap(const uint8_t in) { return b64_unmap[in] - 47; }
+
+/**
+   Decode 4 base64 characters to 3 raw bytes.
+*/
+static inline size_t
+decode_chunk(const uint8_t in[4], uint8_t out[3])
+{
+	out[0] = (uint8_t)(((unmap(in[0]) << 2))        | unmap(in[1]) >> 4);
+	out[1] = (uint8_t)(((unmap(in[1]) << 4) & 0xF0) | unmap(in[2]) >> 2);
+	out[2] = (uint8_t)(((unmap(in[2]) << 6) & 0xC0) | unmap(in[3]));
+	return 1 + (in[2] != '=') + ((in[2] != '=') && (in[3] != '='));
+}
+
+SERD_API
+void*
+serd_base64_decode(const uint8_t* str, size_t len, size_t* size)
+{
+	void* buf = malloc((len * 3) / 4 + 2);
+	*size = 0;
+	for (size_t i = 0, j = 0; i < len; j += 3) {
+		uint8_t in[4] = "====";
+		size_t  n_in  = 0;
+		for (; i < len && n_in < 4; ++n_in) {
+			for (; i < len && !is_base64(str[i]); ++i) {}  // Skip junk
+			in[n_in] = str[i++];
+		}
+		if (n_in > 1) {
+			*size += decode_chunk(in, (uint8_t*)buf + j);
+		}
+	}
+	return buf;
+}