luci-base: add urldecode() and urlencode() C implementations

The C implementations of urlencode and urldecode are considerably faster
than their current Lua counterparts.

On an AMD Geode system, the C variant is up to ten times faster when
decoding strings and up to four times faster when encoding them.

The functions are also designed to only allocate new strings when any
actual changes are required, otherwise they reuse the existing input
strings, reducing the overal memory usage somewhat.

Signed-off-by: Jo-Philipp Wich <jo@mein.io>

1 files changed, 123 insertions, 1 deletions

diff --git a/modules/luci-base/src/template_utils.c b/modules/luci-base/src/template_utils.c
index 3979487f12..eefdd17008 100644
--- a/modules/luci-base/src/template_utils.c
+++ b/modules/luci-base/src/template_utils.c
@@ -1,7 +1,7 @@
 /*
  * LuCI Template - Utility functions
  *
- *   Copyright (C) 2010 Jo-Philipp Wich <jow@openwrt.org>
+ *   Copyright (C) 2010-2018 Jo-Philipp Wich <jo@mein.io>
  *
  *  Licensed under the Apache License, Version 2.0 (the "License");
  *  you may not use this file except in compliance with the License.
@@ -428,6 +428,128 @@ char * striptags(const char *s, unsigned int l)
 	return buf_destroy(buf);
 }
 
+
+static inline bool is_urlencode_char(char c)
+{
+	return !((c >= 'a' && c <= 'z') ||
+	         (c >= 'A' && c <= 'Z') ||
+	         (c >= '0' && c <= '9') ||
+	         (c == '$') || (c == '_') ||
+	         (c == '-') || (c == '.') ||
+	         (c == '~'));
+}
+
+/*
+ * URL-encode all special characters in given string and return
+ * encoded copy.
+ *
+ * If no encoding was required, returns NULL. If an encoded_len
+ * pointer is passed, it is set to the length of the encoded string.
+ *
+ * Sets encoded_len and returns NULL if memory allocation failed.
+ */
+char *urlencode(const char *s, size_t *encoded_len)
+{
+	size_t i, enc_len;
+	char *enc, *ptr;
+
+	for (i = 0, enc_len = 0; s[i]; i++)
+		if (is_urlencode_char(s[i]))
+			enc_len += 3;
+		else
+			enc_len++;
+
+	if (i != enc_len)
+	{
+		if (encoded_len)
+			*encoded_len = enc_len;
+
+		enc = calloc(1, enc_len + 1);
+
+		if (!enc)
+			return NULL;
+
+		for (i = 0, ptr = enc; s[i]; i++)
+			if (is_urlencode_char(s[i]))
+				ptr += snprintf(ptr, 4, "%%%02x", (unsigned char)s[i]);
+			else
+				*ptr++ = s[i];
+
+		return enc;
+	}
+
+	return NULL;
+}
+
+/*
+ * URL-decode given string and return decoded copy.
+ *
+ * If no decoding was required, returns NULL. If an decoded_len
+ * pointer is passed, it is set to the length of the decoded string.
+ *
+ * When keep_plus is true, skip decoding of plus ("+") signs into
+ * space (0x20) characters.
+ *
+ * Sets decoded_len and returns NULL if memory allocation failed.
+ */
+
+#define hex(x) \
+	(((x) <= '9') ? ((x) - '0') : \
+		(((x) <= 'F') ? ((x) - 'A' + 10) : \
+			((x) - 'a' + 10)))
+
+char *urldecode(const char *s, size_t *decoded_len, bool keep_plus)
+{
+	bool changed = false;
+	size_t i, dec_len;
+	char *dec, *ptr;
+
+	for (i = 0, dec_len = 0; s[i]; i++, dec_len++)
+	{
+		if (s[i] == '%' && isxdigit(s[i+1]) && isxdigit(s[i+2]))
+		{
+			changed = true;
+			i += 2;
+		}
+		else if (!keep_plus && s[i] == '+')
+		{
+			changed = true;
+		}
+	}
+
+	if (changed)
+	{
+		if (decoded_len)
+			*decoded_len = dec_len;
+
+		dec = calloc(1, dec_len + 1);
+
+		if (!dec)
+			return NULL;
+
+		for (i = 0, ptr = dec; s[i]; i++)
+		{
+			if (s[i] == '%' && isxdigit(s[i+1]) && isxdigit(s[i+2]))
+			{
+				*ptr++ = (char)(16 * hex(s[i+1]) + hex(s[i+2]));
+				i += 2;
+			}
+			else if (!keep_plus && s[i] == '+')
+			{
+				*ptr++ = ' ';
+			}
+			else
+			{
+				*ptr++ = s[i];
+			}
+		}
+
+		return dec;
+	}
+
+	return NULL;
+}
+
 void luastr_escape(struct template_buffer *out, const char *s, unsigned int l,
 				   int escape_xml)
 {