Commandline tool to compace LevelDB databases.

Summary:
A simple CLI which calles DB->CompactRange()
Can take String key's as range.

Test Plan:
Inserted data into a table.
Waited for a minute, used compact tool on it. File modification time's
changed so Compact did something on the files.

Existing unit tests work.

Reviewers: heyongqiang, dhruba

Reviewed By: dhruba

Differential Revision: https://reviews.facebook.net/D5697

tools/ldb.cc

@@ -7,9 +7,11 @@
 #include <sstream>
 
 #include "leveldb/db.h"
+#include "leveldb/options.h"
 #include "leveldb/iterator.h"
+#include "leveldb/slice.h"
 
-std::string HexToString(std::string str) {
+std::string HexToString(const std::string& str) {
   std::string parsed;
   for (int i = 0; i < str.length(); ) {
     int c;
@@ -20,73 +22,58 @@ std::string HexToString(std::string str) {
   return parsed;
 }
 
-static void print_help() {
+
+static void print_usage() {
   fprintf(stderr,
-      "db_dump "
+    "ldb [compact|dump] "
-      "--start=[START_KEY] "
+    "--db-path=database_path "
-      "--end=[END_KEY] "
+    "[--from=START KEY] "
-      "--max_keys=[NUM] "
+    "[--to=END KEY ] "
-      "--hex "
+    "[--max_keys=[NUM] (only for dump)] "
-      "--count_only "
+    "[--hex ] "
-      "--stats "
+    "[--count_only (only for dump)] "
-      "[PATH]\n");
+    "[--stats (only for dump) ] \n");
 }
 
-int main(int argc, char** argv) {
-  std::string db_path;
-  std::string start;
-  std::string end;
-  uint64_t max_keys = -1;
-  bool print_stats = false;
-  bool count_only = false;
-  uint64_t count = 0;
-  bool hex = false;
 
-  // Parse command line args
+static void safe_open_db(const std::string& dbname, leveldb::DB** db) {
-  char junk;
+  leveldb::Options options;
-  uint64_t n;
+  options.create_if_missing = false;
-  for (int i = 1; i < argc; i++) {
+  leveldb::Status status = leveldb::DB::Open(options, dbname, db);
-    if (strncmp(argv[i], "--start=", 8) == 0) {
+
-      start= argv[i] + 8;
+  if(!status.ok()) {
-    } else if (strncmp(argv[i], "--end=", 6) == 0) {
+    fprintf(
-      end = argv[i] + 6;
+      stderr,
-    } else if (sscanf(argv[i], "--max_keys=%ld%c", &n, &junk) == 1) {
+      "Could not open db at %s\nERROR: %s",
-      max_keys = n;
+      dbname.data(),
-    } else if (strncmp(argv[i], "--stats", 7) == 0) {
+      status.ToString().data()
-      print_stats = true;
+    );
-    } else if (strncmp(argv[i], "--count_only", 12) == 0) {
+    exit(1);
-      count_only = true;
-    } else if (strncmp(argv[i], "--hex", 5) == 0) {
-      hex = true;
-    } else if (i == (argc - 1)) {
-      db_path = argv[i];
-    } else {
-      print_help();
-      exit(1);
-    }
   }
+}
+
+
+static void dump_db(
+  const std::string& db_path,
+  std::string& start,
+  std::string& end,
+  int64_t max_keys,
+  const bool hex,
+  const bool print_stats,
+  const bool count_only
+) {
+    // Parse command line args
+  uint64_t count = 0;
+
 
   if (hex) {
     start = HexToString(start);
     end = HexToString(end);
   }
 
-  if (db_path.empty()) {
-    print_help();
-    exit(1);
-  }
-
-  // Open DB
-  leveldb::Options options;
   leveldb::DB *db;
-  leveldb::Status status = leveldb::DB::Open(options, db_path, &db);
+  safe_open_db(db_path, &db);
 
-  if (!status.ok()) {
-    fprintf(stderr, "%s\n", status.ToString().c_str());
-    exit(1);
-  }
-
-  // Print DB stats if desired
   if (print_stats) {
     std::string stats;
     if (db->GetProperty("leveldb.stats", &stats)) {
@@ -96,7 +83,7 @@ int main(int argc, char** argv) {
 
   // Setup key iterator
   leveldb::Iterator* iter = db->NewIterator(leveldb::ReadOptions());
-  status = iter->status();
+  leveldb::Status status = iter->status();
   if (!status.ok()) {
     fprintf(stderr, "%s\n", status.ToString().c_str());
     delete db;
@@ -140,7 +127,108 @@ int main(int argc, char** argv) {
   // Clean up
   delete iter;
   delete db;
+}
 
-  return 0;
+static void compact(
+  const std::string dbname,
+  std::string from,
+  std::string to,
+  const bool hex
+) {
+
+  leveldb::DB* db;
+  safe_open_db(dbname, &db);
+
+  if(hex) {
+    from = HexToString(from);
+    to = HexToString(to);
+  }
+
+  leveldb::Slice* begin = from.empty() ? NULL : new leveldb::Slice(from);
+  leveldb::Slice* end = to.empty() ? NULL : new leveldb::Slice(to);
+  db->CompactRange(begin, end);
+  delete db;
 }
 
+int main(int argc, char** argv) {
+
+  enum {
+    DUMP, COMPACT
+  } command;
+
+  if (argc < 2) {
+    print_usage();
+    exit(1);
+  }
+
+  size_t n;
+
+  const std::string dbnameKey = "--db-path=";
+  const std::string toKey = "--to=";
+  const std::string fromKey = "--from=";
+  std::string dbname;
+  bool dbnameFound = false;
+  std::string from;
+  std::string to;
+  int64_t temp;
+  int64_t max_keys = -1;
+
+  bool print_stats = false;
+  bool count_only = false;
+  bool hex = false;
+  char junk;
+
+
+  std::string commandString = argv[1];
+  if (commandString == "dump") {
+    command = DUMP;
+  } else if (commandString == "compact") {
+    command = COMPACT;
+  } else {
+    print_usage();
+    exit(1);
+  }
+
+  for (int i = 2; i < argc; i++) {
+    std::string param(argv[i]);
+    if ((n = param.find(dbnameKey)) != std::string::npos) {
+      dbname = param.substr(dbnameKey.size());
+      dbnameFound = true;
+    } else if ((n = param.find(fromKey)) != std::string::npos) {
+      from = param.substr(fromKey.size());
+    } else if ((n = param.find(toKey)) != std::string::npos) {
+      to = param.substr(toKey.size());
+    } else if (sscanf(argv[i], "--max_keys=%ld%c", &temp, &junk) == 1) {
+      max_keys = temp;
+    } else if (strncmp(argv[i], "--stats", 7) == 0) {
+      print_stats = true;
+    } else if (strncmp(argv[i], "--count_only", 12) == 0) {
+      count_only = true;
+    } else if (strncmp(argv[i], "--hex", 5) == 0) {
+      hex = true;
+    } else {
+      print_usage();
+      exit(1);
+    }
+  }
+
+  if (!dbnameFound || dbname.empty()) {
+    fprintf(stderr, "DB path required. See help\n");
+    print_usage();
+    exit(1);
+  }
+
+  switch(command) {
+    case DUMP:
+      dump_db(dbname, from, to, max_keys, hex, print_stats, count_only);
+      break;
+    case COMPACT:
+      compact(dbname, from, to, hex);
+      break;
+    default:
+      print_usage();
+      exit(1);
+  }
+
+  return 0;
+}