From 0e7ba4e401211f2a08dbf1bff007e6edb78a2cfc Mon Sep 17 00:00:00 2001
From: Mingye Wang <arthur200126@gmail.com>
Date: Sun, 16 Feb 2020 22:42:48 +0800
Subject: [PATCH] Faster base64 (#63)

Cool kids on the Internet use two tables to do base64. That looks cool.
We should get it.
---
 src/misc.cpp | 49 ++++++++++++++++++++++++++++++++++---------------
 src/misc.h   |  4 +++-
 2 files changed, 37 insertions(+), 16 deletions(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index f9ee102..b6588e1 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -243,23 +243,37 @@ std::string base64_encode(const std::string &string_to_encode)
 
 }
 
-std::string base64_decode(const std::string &encoded_string)
+std::string base64_decode(const std::string &encoded_string, bool accept_urlsafe)
 {
-    int in_len = encoded_string.size();
-    int i = 0;
-    int j = 0;
-    int in_ = 0;
+    string_size in_len = encoded_string.size();
+    string_size i = 0;
+    string_size in_ = 0;
     unsigned char char_array_4[4], char_array_3[3];
+    static unsigned char dtable[256], itable[256], table_ready = 0;
     std::string ret;
 
-    while (in_len-- && ( encoded_string[in_] != '=') && is_base64(encoded_string[in_]))
+    // Should not need thread_local with the flag...
+    if (!table_ready) {
+        // No memset needed for static/TLS
+        for (string_size k = 0; k < base64_chars.length(); k++) {
+            dtable[base64_chars[k]] = k;  // decode (find)
+            itable[base64_chars[k]] = 1;  // is_base64
+        }
+        // Add urlsafe table
+        dtable['-'] = dtable['+']; itable['-'] = 2;
+        dtable['_'] = dtable['/']; itable['_'] = 2;
+        table_ready = 1;
+    }
+
+    while (in_len-- && (encoded_string[in_] != '=') &&
+           (accept_urlsafe ? itable[encoded_string[in_]] : (itable[encoded_string[in_]] == 1)))
     {
         char_array_4[i++] = encoded_string[in_];
         in_++;
-        if (i ==4)
+        if (i == 4)
         {
-            for (i = 0; i <4; i++)
-                char_array_4[i] = base64_chars.find(char_array_4[i]);
+            for (string_size j = 0; j < 4; j++)
+                char_array_4[j] = dtable[char_array_4[j]];
 
             char_array_3[0] = (char_array_4[0] << 2) + ((char_array_4[1] & 0x30) >> 4);
             char_array_3[1] = ((char_array_4[1] & 0xf) << 4) + ((char_array_4[2] & 0x3c) >> 2);
@@ -273,17 +287,17 @@ std::string base64_decode(const std::string &encoded_string)
 
     if (i)
     {
-        for (j = i; j <4; j++)
+        for (string_size j = i; j <4; j++)
             char_array_4[j] = 0;
 
-        for (j = 0; j <4; j++)
-            char_array_4[j] = base64_chars.find(char_array_4[j]);
+        for (string_size j = 0; j <4; j++)
+            char_array_4[j] = dtable[char_array_4[j]];
 
         char_array_3[0] = (char_array_4[0] << 2) + ((char_array_4[1] & 0x30) >> 4);
         char_array_3[1] = ((char_array_4[1] & 0xf) << 4) + ((char_array_4[2] & 0x3c) >> 2);
         char_array_3[2] = ((char_array_4[2] & 0x3) << 6) + char_array_4[3];
 
-        for (j = 0; (j < i - 1); j++)
+        for (string_size j = 0; (j < i - 1); j++)
             ret += char_array_3[j];
     }
 
@@ -621,14 +635,19 @@ std::string urlsafe_base64_reverse(const std::string &encoded_string)
     return replace_all_distinct(replace_all_distinct(encoded_string, "-", "+"), "_", "/");
 }
 
+std::string urlsafe_base64(const std::string &encoded_string)
+{
+    return replace_all_distinct(replace_all_distinct(replace_all_distinct(encoded_string, "+", "-"), "/", "_"), "=", "");
+}
+
 std::string urlsafe_base64_decode(const std::string &encoded_string)
 {
-    return base64_decode(urlsafe_base64_reverse(encoded_string));
+    return base64_decode(encoded_string, true);
 }
 
 std::string urlsafe_base64_encode(const std::string &string_to_encode)
 {
-    return replace_all_distinct(replace_all_distinct(replace_all_distinct(base64_encode(string_to_encode), "+", "-"), "/", "_"), "=", "");
+    return urlsafe_base64(base64_encode(string_to_encode));
 }
 
 std::string getMD5(const std::string &data)
diff --git a/src/misc.h b/src/misc.h
index 39f7685..a70a348 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -13,6 +13,7 @@
 #define PATH_SLASH "//"
 #endif // _WIN32
 
+typedef std::string::size_type string_size;
 typedef std::vector<std::string> string_array;
 typedef const std::string &refCnstStr;
 
@@ -23,12 +24,13 @@ static const std::string base64_chars =
 
 std::string UrlEncode(const std::string& str);
 std::string UrlDecode(const std::string& str);
-std::string base64_decode(const std::string &encoded_string);
+std::string base64_decode(const std::string &encoded_string, bool accept_urlsafe = false);
 std::string base64_encode(const std::string &string_to_encode);
 
 std::vector<std::string> split(const std::string &s, const std::string &seperator);
 std::string getUrlArg(const std::string &url, const std::string &request);
 std::string replace_all_distinct(std::string str, const std::string &old_value, const std::string &new_value);
+std::string urlsafe_base64(const std::string &encoded_string);
 std::string urlsafe_base64_reverse(const std::string &encoded_string);
 std::string urlsafe_base64_decode(const std::string &encoded_string);
 std::string urlsafe_base64_encode(const std::string &string_to_encode);