Separate base64 implementation

5 files changed, 179 insertions, 82 deletions

diff --git a/src/base64.c b/src/base64.c
new file mode 100644
index 00000000..49cae6bc
--- /dev/null
+++ b/src/base64.c
@@ -0,0 +1,124 @@
+/*
+  Copyright 2011-2018 David Robillard <http://drobilla.net>
+
+  Permission to use, copy, modify, and/or distribute this software for any
+  purpose with or without fee is hereby granted, provided that the above
+  copyright notice and this permission notice appear in all copies.
+
+  THIS SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+  WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+  MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+  ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+  WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+  ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+  OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+*/
+
+#include "base64.h"
+
+#include "serd_internal.h"
+#include "string_utils.h"
+
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+
+/**
+   Base64 encoding table.
+
+   @see <a href="http://tools.ietf.org/html/rfc3548#section-3">RFC3986 S3</a>.
+*/
+static const uint8_t b64_map[] =
+	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
+
+/**
+   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[] =
+	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$m$$$ncdefghijkl$$$$$$"
+	"$/0123456789:;<=>?@ABCDEFGH$$$$$$IJKLMNOPQRSTUVWXYZ[\\]^_`ab$$$$"
+	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$"
+	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$";
+
+/** Encode 3 raw bytes to 4 base64 characters. */
+static inline void
+encode_chunk(uint8_t out[4], const uint8_t in[3], size_t n_in)
+{
+	out[0] = b64_map[in[0] >> 2];
+	out[1] = b64_map[((in[0] & 0x03) << 4) | ((in[1] & 0xF0) >> 4)];
+	out[2] = ((n_in > 1)
+	          ? (b64_map[((in[1] & 0x0F) << 2) | ((in[2] & 0xC0) >> 6)])
+	          : (uint8_t)'=');
+	out[3] = ((n_in > 2) ? b64_map[in[2] & 0x3F] : (uint8_t)'=');
+}
+
+size_t
+serd_base64_get_length(const size_t size, const bool wrap_lines)
+{
+	return ((size + 2) / 3) * 4 + (wrap_lines ? (size / 57) : 0);
+}
+
+bool
+serd_base64_encode(char* const       str,
+                   const void* const buf,
+                   const size_t      size,
+                   const bool        wrap_lines)
+{
+	uint8_t* const out         = (uint8_t*)str;
+	bool           has_newline = false;
+	for (size_t i = 0, j = 0; i < size; i += 3, j += 4) {
+		uint8_t in[4] = { 0, 0, 0, 0 };
+		size_t  n_in  = MIN(3, size - i);
+		memcpy(in, (const uint8_t*)buf + i, n_in);
+
+		if (wrap_lines && i > 0 && (i % 57) == 0) {
+			out[j++]    = '\n';
+			has_newline = true;
+		}
+
+		encode_chunk(out + j, in, n_in);
+	}
+
+	return has_newline;
+}
+
+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] != '='));
+}
+
+void*
+serd_base64_decode(const char* str, size_t len, size_t* size)
+{
+	const uint8_t* ustr = (const uint8_t*)str;
+
+	void* buf = malloc((len * 3) / 4 + 2);
+	*size     = 0;
+	for (size_t i = 0, j = 0; i < len; j += 3) {
+		uint8_t in[] = "====";
+		size_t  n_in = 0;
+		for (; i < len && n_in < 4; ++n_in) {
+			for (; i < len && !is_base64(ustr[i]); ++i) {} // Skip junk
+			in[n_in] = ustr[i++];
+		}
+		if (n_in > 1) {
+			*size += decode_chunk(in, (uint8_t*)buf + j);
+		}
+	}
+	return buf;
+}
diff --git a/src/base64.h b/src/base64.h
new file mode 100644
index 00000000..b5e04373
--- /dev/null
+++ b/src/base64.h
@@ -0,0 +1,47 @@
+/*
+  Copyright 2011-2018 David Robillard <http://drobilla.net>
+
+  Permission to use, copy, modify, and/or distribute this software for any
+  purpose with or without fee is hereby granted, provided that the above
+  copyright notice and this permission notice appear in all copies.
+
+  THIS SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+  WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+  MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+  ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+  WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+  ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+  OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+*/
+
+#ifndef SERD_BASE64_H
+#define SERD_BASE64_H
+
+#include "serd/serd.h"
+
+#include <stdbool.h>
+#include <stddef.h>
+
+/**
+   Return the number of bytes required to encode `size` bytes in base64.
+
+   @param size The number of input (binary) bytes to encode.
+   @param wrap_lines Whether to wrap lines at a fixed length.
+   @return The length of the base64 encoding, exclusing null terminator.
+*/
+size_t
+serd_base64_get_length(size_t size, bool wrap_lines);
+
+/**
+   Encode `size` bytes of `buf` into `str`, which must be large enough.
+
+   @param str Output string buffer.
+   @param buf Input binary data.
+   @param size Number of bytes to encode from `buf`.
+   @param wrap_lines Whether to wrap lines at a fixed length.
+   @return True iff `str` contains newlines.
+*/
+bool
+serd_base64_encode(char* str, const void* buf, size_t size, bool wrap_lines);
+
+#endif  // SERD_BASE64_H
diff --git a/src/node.c b/src/node.c
index 7f124f13..36c2c8ca 100644
--- a/src/node.c
+++ b/src/node.c
@@ -16,6 +16,7 @@
 
 #include "node.h"
 
+#include "base64.h"
 #include "serd_internal.h"
 #include "string_utils.h"
 
@@ -470,45 +471,16 @@ serd_node_new_integer(int64_t i)
 	return node;
 }
 
-/**
-   Base64 encoding table.
-   @see <a href="http://tools.ietf.org/html/rfc3548#section-3">RFC3986 S3</a>.
-*/
-static const uint8_t b64_map[] =
-	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
-
-/**
-   Encode 3 raw bytes to 4 base64 characters.
-*/
-static inline void
-encode_chunk(uint8_t out[4], const uint8_t in[3], size_t n_in)
-{
-	out[0] = b64_map[in[0] >> 2];
-	out[1] = b64_map[((in[0] & 0x03) << 4) | ((in[1] & 0xF0) >> 4)];
-	out[2] = ((n_in > 1)
-	          ? (b64_map[((in[1] & 0x0F) << 2) | ((in[2] & 0xC0) >> 6)])
-	          : (uint8_t)'=');
-	out[3] = ((n_in > 2) ? b64_map[in[2] & 0x3F] : (uint8_t)'=');
-}
-
 SerdNode*
 serd_node_new_blob(const void* buf, size_t size, bool wrap_lines)
 {
-	const size_t len  = ((size + 2) / 3) * 4 + (wrap_lines ? (size / 57) : 0);
-	SerdNode*    node = serd_node_malloc(len + 1, 0, SERD_LITERAL);
-	uint8_t*     str  = (uint8_t*)serd_node_buffer(node);
-	for (size_t i = 0, j = 0; i < size; i += 3, j += 4) {
-		uint8_t in[4] = { 0, 0, 0, 0 };
-		size_t  n_in  = MIN(3, size - i);
-		memcpy(in, (const uint8_t*)buf + i, n_in);
-
-		if (wrap_lines && i > 0 && (i % 57) == 0) {
-			str[j++] = '\n';
-			node->flags |= SERD_HAS_NEWLINE;
-		}
+	const size_t    len  = serd_base64_get_length(size, wrap_lines);
+	SerdNode* const node = serd_node_malloc(len + 1, 0, SERD_LITERAL);
 
-		encode_chunk(str + j, in, n_in);
+	if (serd_base64_encode(serd_node_buffer(node), buf, size, wrap_lines)) {
+		node->flags |= SERD_HAS_NEWLINE;
 	}
+
 	node->n_bytes = len + 1;
 	return node;
 }
diff --git a/src/string.c b/src/string.c
index 0ab6cb4a..1755d48a 100644
--- a/src/string.c
+++ b/src/string.c
@@ -107,50 +107,3 @@ serd_strtod(const char* str, size_t* end)
 
 	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[] =
-	"$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$m$$$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] != '='));
-}
-
-void*
-serd_base64_decode(const char* str, size_t len, size_t* size)
-{
-	const uint8_t* ustr = (const uint8_t*)str;
-
-	void* buf = malloc((len * 3) / 4 + 2);
-	*size = 0;
-	for (size_t i = 0, j = 0; i < len; j += 3) {
-		uint8_t in[] = "====";
-		size_t  n_in = 0;
-		for (; i < len && n_in < 4; ++n_in) {
-			for (; i < len && !is_base64(ustr[i]); ++i) {}  // Skip junk
-			in[n_in] = ustr[i++];
-		}
-		if (n_in > 1) {
-			*size += decode_chunk(in, (uint8_t*)buf + j);
-		}
-	}
-	return buf;
-}
diff --git a/wscript b/wscript
index 47b2c10e..ea5d2ba7 100644
--- a/wscript
+++ b/wscript
@@ -75,7 +75,8 @@ def configure(conf):
     autowaf.display_msg(conf, 'Unit tests', bool(conf.env.BUILD_TESTS))
     print('')
 
-lib_source = ['src/byte_source.c',
+lib_source = ['src/base64.c',
+              'src/byte_source.c',
               'src/env.c',
               'src/n3.c',
               'src/node.c',