Change the code to use Spookyhash.

* This change removes the external dependency on OpenSSL, which should
  allow the code to compile on Mac OS X without specifying the
  MAC_OS_X_VERSION_MIN_REQUIRED non-sense.
* Spookyhash should be faster than SHA-1's Merkle-Damgård construction.
* Spookyhash is only 128-bits as opposed to SHA-1's 160-bits so memory
  consumption should be less. I doubt it will lead to too many
  collisions.
* The docs are updated to reflect the fact that OpenSSL is not a
  dependency.

README

@@ -23,8 +23,6 @@ Lnav requires the following software packages:
   readline  - The readline line editing library.
   zlib      - The zlib compression library.
   bz2       - The bzip2 compression library.
-  openssl   - The OpenSSL cryptographic library.  (Only the hash
-              functions from this library are used.)
 
 
 INSTALLATION

docs/source/intro.rst

@@ -20,7 +20,6 @@ When compiling from source, the following dependencies are required:
 * `PCRE <http://www.pcre.org>`_ -- Versions greater than 8.20 give better
   performance since the PCRE JIT will be leveraged.
 * `SQLite <http://www.sqlite.org>`_
-* `OpenSSL <http://www.openssl.org>`_
 * `ZLib <http://wwww.zlib.net>`_
 * `Bzip2 <http://www.bzip.org>`_
 * `Readline <http://www.gnu.org/s/readline>`_

src/data_parser.hh

@@ -32,7 +32,7 @@
 
 #include <stdio.h>
 
-#include <openssl/sha.h>
+#include "spookyhash/SpookyV2.h"
 
 #include <list>
 #include <vector>
@@ -173,7 +173,7 @@ public:
 
     static FILE *TRACE_FILE;
 
-    typedef byte_array<SHA_DIGEST_LENGTH> schema_id_t;
+    typedef byte_array<16> schema_id_t;
 
     struct element;
     /* typedef std::list<element> element_list_t; */
@@ -421,7 +421,7 @@ private:
         element_list_t ELEMENT_LIST_T(el_stack), ELEMENT_LIST_T(free_row),
         ELEMENT_LIST_T(key_comps), ELEMENT_LIST_T(value),
         ELEMENT_LIST_T(prefix);
-        SHA_CTX context;
+        SpookyHash context;
 
         POINT_TRACE("pairup_start");
 
@@ -558,7 +558,7 @@ private:
 
         POINT_TRACE("pairup_stack");
 
-        SHA_Init(&context);
+        context.Init(0, 0);
         while (!el_stack.empty()) {
             element_list_t::iterator kv_iter = el_stack.begin();
             if (kv_iter->e_token == DNT_VALUE) {
@@ -595,7 +595,7 @@ private:
 
 
             if (schema != NULL) {
-                SHA_Update(&context, key_val.c_str(), key_val.length());
+                context.Update(key_val.c_str(), key_val.length());
             }
 
             while (!free_row.empty()) {
@@ -658,7 +658,7 @@ private:
                                 value.e_sub_elements->begin(),
                                 value.e_sub_elements->end());
                 pairs_out.clear();
-                SHA_Init(&context);
+                context.Init(0, 0);
             }
         }
 
@@ -700,7 +700,7 @@ private:
                     // columns is significant.  I don't think we want to
                     // use the token ID since some columns values might vary
                     // between rows.
-                    SHA_Update(&context, " ", 1);
+                    context.Update(" ", 1);
                 }
                 break;
 
@@ -711,7 +711,7 @@ private:
                     std::string key_val = this->get_element_string(
                         free_row.front());
 
-                    SHA_Update(&context, key_val.c_str(), key_val.length());
+                    context.Update(key_val.c_str(), key_val.length());
                 }
                 break;
                 }
@@ -733,7 +733,8 @@ private:
         }
 
         if (schema != NULL) {
-            SHA_Final(this->dp_schema_id.ba_data, &context);
+            context.Final((uint64 *)(this->dp_schema_id.ba_data),
+                          (uint64 *)((uint64 *)(this->dp_schema_id.ba_data)+1));
         }
     };
 

src/lnav_util.cc

@@ -44,12 +44,12 @@
 
 std::string hash_string(const std::string &str)
 {
-    byte_array<SHA_DIGEST_LENGTH> hash;
-    SHA_CTX context;
+    byte_array<16> hash;
+    SpookyHash context;
 
-    SHA_Init(&context);
-    SHA_Update(&context, str.c_str(), str.length());
-    SHA_Final(hash.out(), &context);
+    context.Init(0, 0);
+    context.Update(str.c_str(), str.length());
+    context.Final((uint64 *)hash.out(), (uint64 *)((uint64 *)(hash.out())+1));
 
     return hash.to_string();
 }

src/lnav_util.hh

@@ -37,7 +37,7 @@
 #include <time.h>
 #include <sys/types.h>
 
-#include <openssl/sha.h>
+#include "spookyhash/SpookyV2.h"
 
 #include <string>
 
@@ -90,15 +90,15 @@ std::string time_ago(time_t last_time);
 #error "off_t has unhandled size..."
 #endif
 
-struct sha_updater {
-    sha_updater(SHA_CTX *context) : su_context(context) { };
+struct hash_updater {
+    hash_updater(SpookyHash *context) : su_context(context) { };
 
     void operator()(const std::string &str)
     {
-        SHA_Update(this->su_context, str.c_str(), str.length());
+        this->su_context->Update(str.c_str(), str.length());
     }
 
-    SHA_CTX *su_context;
+    SpookyHash *su_context;
 };
 
 std::string hash_string(const std::string &str);

src/log_format.hh

@@ -241,7 +241,7 @@ public:
      *
      * @param ba The SHA-1 hash of the constant parts of this log line.
      */
-    void set_schema(const byte_array<20> &ba)
+    void set_schema(const byte_array<16> &ba)
     {
         memcpy(this->ll_schema, ba.in(), sizeof(this->ll_schema));
     };
@@ -255,7 +255,7 @@ public:
      * @return    True if the first four bytes of the given schema match the
      *   schema stored in this log line.
      */
-    bool match_schema(const byte_array<20> &ba) const
+    bool match_schema(const byte_array<16> &ba) const
     {
         return memcmp(this->ll_schema, ba.in(), sizeof(this->ll_schema)) == 0;
     }

src/sequence_matcher.cc

@@ -29,7 +29,7 @@
 
 #include "config.h"
 
-#include <openssl/sha.h>
+#include "spookyhash/SpookyV2.h"
 
 #include "sequence_matcher.hh"
 
@@ -65,18 +65,18 @@ sequence_matcher::sequence_matcher(field_col_t &example)
 void sequence_matcher::identity(const std::vector<string> &values,
                                 id_t &id_out)
 {
-    SHA_CTX context;
+    SpookyHash context;
     int     lpc = 0;
 
-    SHA_Init(&context);
+    context.Init(0, 0);
     for (std::list<field>::iterator iter = sm_fields.begin();
          iter != sm_fields.end();
          ++iter, lpc++) {
         if (iter->sf_type == FT_VARIABLE) {
-            SHA_Update(&context,
-                       values[lpc].c_str(),
-                       values[lpc].length() + 1);
+            context.Update( values[lpc].c_str(),
+                            values[lpc].length() + 1);
         }
     }
-    SHA_Final(id_out.ba_data, &context);
+    context.Final((uint64 *)id_out.ba_data,
+                  (uint64 *)((uint64 *)(id_out.ba_data)+1));
 }

src/sequence_matcher.hh

@@ -41,7 +41,7 @@ public:
     typedef std::vector<std::string> field_row_t;
     typedef std::list<field_row_t>   field_col_t;
 
-    typedef byte_array<20>           id_t;
+    typedef byte_array<16>           id_t;
 
     enum field_type_t {
         FT_VARIABLE,

src/session_data.cc

@@ -37,7 +37,7 @@
 #include <sys/types.h>
 #include <dirent.h>
 
-#include <openssl/sha.h>
+#include "spookyhash/SpookyV2.h"
 
 #include <algorithm>
 
@@ -243,17 +243,17 @@ static void cleanup_session_data(void)
 
 void init_session(void)
 {
-    byte_array<SHA_DIGEST_LENGTH> hash;
-    SHA_CTX context;
+    byte_array<16> hash;
+    SpookyHash context;
 
     lnav_data.ld_session_time = time(NULL);
 
-    SHA_Init(&context);
-    sha_updater updater(&context);
+    context.Init(0, 0);
+    hash_updater updater(&context);
     for_each(lnav_data.ld_file_names.begin(),
              lnav_data.ld_file_names.end(),
              object_field(updater, &pair<string, int>::first));
-    SHA_Final(hash.out(), &context);
+    context.Final((uint64 *)hash.out(), (uint64 *)((uint64 *)(hash.out())+1));
 
     lnav_data.ld_session_id = hash.to_string();
 }

src/spookyhash/SpookyV2.h

@@ -26,6 +26,8 @@
 // slower than MD5.
 //
 
+#ifndef _SPOOKYHASH_V2_H
+#define _SPOOKYHASH_V2_H
 #include <stddef.h>
 
 #ifdef _MSC_VER
@@ -296,4 +298,4 @@ private:
 };
 
 
-
+#endif

test/drive_sequencer.cc

@@ -38,8 +38,6 @@
 #include <unistd.h>
 #include <string.h>
 
-#include <openssl/sha.h>
-
 #include <map>
 #include <list>
 #include <vector>