#
#
# patch "cmd_files.cc"
#  from [e51fa86b305fd51c79fac69f57e31e54810efc1e]
#    to [2edcbe2df9fb1890c646a386be837879315f0995]
# 
# patch "database.cc"
#  from [3a221934a73d25f35c980a3f857b1cadde18d70f]
#    to [f52aceca72edd2079f6203ae0ad016d5cf5891fb]
# 
# patch "database.hh"
#  from [62ca0ac4e28df39f316de9117fc10a171c73f1a5]
#    to [47695146fe0bf270c37e3e4646efd6d6b78abef0]
# 
# patch "database_check.cc"
#  from [beb0e60421b8bf7552071a9af31ef6630cfd3b51]
#    to [ae542ecb0c5c74e73bc900a9307de9ceaaffc5a0]
# 
# patch "migrate_ancestry.cc"
#  from [2b91880e4558527585bccbba0a85705f6fda6304]
#    to [351c11cd194241e30554fa81c2f4b7425e76daf2]
# 
# patch "migrate_schema.cc"
#  from [ce591f0adeb455c607c3b5a934aa7a7179c15dc6]
#    to [71649cdd46641a9c63c8b6dfec3a02af6a426541]
# 
# patch "schema.sql"
#  from [956c61eca12f587d85043622d852a4d85c72a80b]
#    to [e95592782a6420be6e3a0e20f8d053147de38d70]
# 
# patch "vocab.hh"
#  from [17ac4c8f007f5b46d0c34f5140942ac8c9f8ae75]
#    to [c5ab0a9b34f8c7a2f23a00b9782e25542ea90083]
#
============================================================
--- database.cc	3a221934a73d25f35c980a3f857b1cadde18d70f
+++ database.cc	f52aceca72edd2079f6203ae0ad016d5cf5891fb
@@ -113,9 +113,10 @@ namespace
 {
   struct query_param
   {
-    enum arg_type { text, blob };
+    enum arg_type { text, blob, int64 };
     arg_type type;
-    string data;
+    string string_data;
+    u64 int_data;
   };
 
   query_param
@@ -130,6 +131,7 @@ namespace
     query_param q = {
       query_param::text,
       txt,
+      0,
     };
     return q;
   }
@@ -140,10 +142,22 @@ namespace
     query_param q = {
       query_param::blob,
       blb,
+      0,
     };
     return q;
   }
 
+  query_param
+  int64(u64 const & num)
+  {
+    query_param q = {
+      query_param::int64,
+      "",
+      num,
+    };
+    return q;
+  }
+
   struct query
   {
     explicit query(string const & cmd)
@@ -336,12 +350,20 @@ private:
                    string const & data_table,
                    string const & delta_table);
 
+  void get_version_size(id const & ident,
+                        file_size & size,
+                        string const & size_table);
+
   void drop(id const & base,
             string const & table);
+
   void put_file_delta(file_id const & ident,
                       file_id const & base,
                       file_delta const & del);
 
+  void put_file_size(file_id const & ident,
+                     file_data const & data);
+
   void put_roster_delta(revision_id const & ident,
                         revision_id const & base,
                         roster_delta const & del);
@@ -617,7 +639,9 @@ database_impl::check_caches()
 {
   if (table_has_data("revisions"))
     {
-      E(table_has_data("rosters") && table_has_data("heights"),
+      E(table_has_data("rosters")
+        && table_has_data("heights")
+        && table_has_data("file_sizes"),
         origin::no_fault,
         F("database %s lacks some cached data\n"
           "run '%s db regenerate_caches' to restore use of this database")
@@ -995,6 +1019,7 @@ database::info(ostream & out, bool analy
   counts.push_back(imp->count("roster_deltas"));
   counts.push_back(imp->count("files"));
   counts.push_back(imp->count("file_deltas"));
+  counts.push_back(imp->count("file_sizes"));
   counts.push_back(imp->count("revisions"));
   counts.push_back(imp->count("revision_ancestry"));
   counts.push_back(imp->count("revision_certs"));
@@ -1032,6 +1057,8 @@ database::info(ostream & out, bool analy
     bytes.push_back(imp->space("files", "length(id) + length(data)", total));
     bytes.push_back(imp->space("file_deltas",
                           "length(id) + length(base) + length(delta)", total));
+    bytes.push_back(imp->space("file_sizes",
+                          "length(id) + length(size)", total));
     bytes.push_back(imp->space("revisions", "length(id) + length(data)", total));
     bytes.push_back(imp->space("revision_ancestry",
                           "length(parent) + length(child)", total));
@@ -1067,6 +1094,7 @@ database::info(ostream & out, bool analy
       "  roster deltas   : %s\n"
       "  full files      : %s\n"
       "  file deltas     : %s\n"
+      "  file sizes      : %s\n"
       "  revisions       : %s\n"
       "  ancestry edges  : %s\n"
       "  certs           : %s\n"
@@ -1076,6 +1104,7 @@ database::info(ostream & out, bool analy
       "  roster deltas   : %s\n"
       "  full files      : %s\n"
       "  file deltas     : %s\n"
+      "  file sizes      : %s\n"
       "  revisions       : %s\n"
       "  cached ancestry : %s\n"
       "  certs           : %s\n"
@@ -1444,13 +1473,21 @@ database_impl::fetch(results & res,
       if (global_sanity.debug_p())
         {
           string prefix;
-          string log(query.args[param-1].data);
+          string log;
 
           if (query.args[param-1].type == query_param::blob)
             {
               prefix = "x";
-              log = encode_hexenc(log, origin::internal);
+              log = encode_hexenc(query.args[param-1].string_data, origin::internal);
             }
+          else if (query.args[param-1].type == query_param::int64)
+            {
+              log = lexical_cast<string>(query.args[param-1].int_data);
+            }
+          else
+            {
+              log = query.args[param-1].string_data;
+            }
 
           if (log.size() > constants::db_log_line_sz)
             log = log.substr(0, constants::db_log_line_sz - 2) + "..";
@@ -1462,17 +1499,23 @@ database_impl::fetch(results & res,
         {
         case query_param::text:
           sqlite3_bind_text(i->second.stmt(), param,
-                            idx(query.args, param - 1).data.c_str(), -1,
+                            idx(query.args, param - 1).string_data.c_str(), -1,
                             SQLITE_STATIC);
           break;
         case query_param::blob:
           {
-            string const & data = idx(query.args, param - 1).data;
+            string const & data = idx(query.args, param - 1).string_data;
             sqlite3_bind_blob(i->second.stmt(), param,
                               data.data(), data.size(),
                               SQLITE_STATIC);
           }
           break;
+        case query_param::int64:
+          {
+            u64 data = idx(query.args, param - 1).int_data;
+            sqlite3_bind_int64(i->second.stmt(), param, data);
+          }
+          break;
         default:
           I(false);
         }
@@ -1591,7 +1634,7 @@ database_impl::schedule_delayed_file(fil
 
 void
 database_impl::schedule_delayed_file(file_id const & an_id,
-                                      file_data const & dat)
+                                     file_data const & dat)
 {
   if (!have_delayed_file(an_id))
     {
@@ -1934,6 +1977,16 @@ void
 }
 
 void
+database_impl::put_file_size(file_id const & ident,
+                             file_data const & data)
+{
+  I(!null_id(ident));
+  file_size size = data.inner()().size();
+  query q("INSERT INTO file_sizes(id, size) VALUES (?, ?)");
+  execute(q % blob(ident.inner()()) % int64(size));
+}
+
+void
 database_impl::put_roster_delta(revision_id const & ident,
                                  revision_id const & base,
                                  roster_delta const & del)
@@ -2044,6 +2097,18 @@ database_impl::get_version(id const & id
     vcache.insert_clean(ident, dat);
 }
 
+void
+database_impl::get_version_size(id const & ident,
+                                file_size & size,
+                                std::string const & size_table)
+{
+  results res;
+  query q("SELECT size FROM " + size_table + " WHERE id = ?");
+  fetch(res, one_col, one_row, q % blob(ident()));
+  I(!res.empty());
+  size = lexical_cast<u64>(res[0][0]);
+}
+
 struct roster_reconstruction_graph : public reconstruction_graph
 {
   database_impl & imp;
@@ -2339,6 +2404,13 @@ void
 }
 
 void
+database::get_file_size(file_id const & id,
+                        file_size & size)
+{
+  imp->get_version_size(id.inner(), size, "file_sizes");
+}
+
+void
 database::get_manifest_version(manifest_id const & id,
                                manifest_data & dat)
 {
@@ -2352,9 +2424,12 @@ database::put_file(file_id const & id,
                    file_data const & dat)
 {
   if (file_version_exists(id))
-    L(FL("file version '%s' already exists in db") % id);
-  else
-    imp->schedule_delayed_file(id, dat);
+    {
+      L(FL("file version '%s' already exists in db") % id);
+      return;
+    }
+  imp->schedule_delayed_file(id, dat);
+  imp->put_file_size(id, dat);
 }
 
 void
@@ -2415,6 +2490,7 @@ database::put_file_version(file_id const
       if (!file_or_manifest_base_exists(new_id, "files"))
         {
           imp->schedule_delayed_file(new_id, new_data);
+          imp->put_file_size(new_id, new_data);
         }
       if (!imp->delta_exists(old_id, new_id, "file_deltas"))
         {
@@ -2423,6 +2499,10 @@ database::put_file_version(file_id const
     }
   if (make_forward_deltas)
     {
+      if (!file_or_manifest_base_exists(new_id, "files"))
+        {
+          imp->put_file_size(new_id, new_data);
+        }
       if (!imp->delta_exists(new_id, old_id, "file_deltas"))
         {
           put_file_delta(new_id, old_id, del);
@@ -3039,6 +3119,12 @@ database::delete_existing_branch_leaves(
   imp->execute(query("DELETE FROM branch_leaves"));
 }
 
+void
+database::delete_existing_file_sizes()
+{
+  imp->execute(query("DELETE FROM file_sizes"));
+}
+
 /// Deletes one revision from the local database.
 /// @see kill_rev_locally
 void
============================================================
--- database.hh	62ca0ac4e28df39f316de9117fc10a171c73f1a5
+++ database.hh	47695146fe0bf270c37e3e4646efd6d6b78abef0
@@ -129,6 +129,10 @@ public:
   void get_file_version(file_id const & ident,
                         file_data & dat);
 
+  // gets the (cached) size of the file if it exists
+  void get_file_size(file_id const & ident,
+                     file_size & size);
+
   // put file w/o predecessor into db
   void put_file(file_id const & new_id,
                 file_data const & dat);
@@ -328,6 +332,7 @@ public:
   void compute_branch_leaves(cert_value const & branch_name, std::set<revision_id> & revs);
   void recalc_branch_leaves(cert_value const & branch_name);
   void delete_existing_branch_leaves();
+  void delete_existing_file_sizes();
 
   // Used through project.cc
   outdated_indicator get_revision_certs(revision_id const & ident,
============================================================
--- schema.sql	956c61eca12f587d85043622d852a4d85c72a80b
+++ schema.sql	e95592782a6420be6e3a0e20f8d053147de38d70
@@ -37,6 +37,12 @@ CREATE TABLE file_deltas
 	unique(id, base)
 	);
 
+CREATE TABLE file_sizes
+	(
+	id primary key,     -- joins with files.id or file_deltas.id
+	size not null       -- the size of the file in byte
+	);
+
 CREATE TABLE revisions
 	(
 	id primary key,      -- SHA1(text of revision)
============================================================
--- vocab.hh	17ac4c8f007f5b46d0c34f5140942ac8c9f8ae75
+++ vocab.hh	c5ab0a9b34f8c7a2f23a00b9782e25542ea90083
@@ -90,6 +90,8 @@ typedef     file< delta >      file_delt
 typedef manifest< delta >  manifest_delta;
 typedef     file< delta >      file_delta;
 
+typedef u64 file_size;
+
 // diff type; this is here and not diff_patch.hh, because options_list.hh
 // needs to refer to it
 enum diff_type
============================================================
--- migrate_schema.cc	ce591f0adeb455c607c3b5a934aa7a7179c15dc6
+++ migrate_schema.cc	71649cdd46641a9c63c8b6dfec3a02af6a426541
@@ -676,7 +676,7 @@ char const migrate_add_heights[] =
   ;
 
 // this is a function because it has to refer to the numeric constant
-// defined in schema_migration.hh.
+// defined in migration.hh.
 static void
 migrate_add_ccode(sqlite3 * db, key_store &)
 {
@@ -716,7 +716,7 @@ char const migrate_to_binary_hashes[] =
   "DROP TABLE tmp;"
   "CREATE INDEX revision_certs__id ON revision_certs (id);"
 
-  // We altered a comment on this table, thus we need to recreated it.
+  // We altered a comment on this table, thus we need to recreate it.
   // Additionally, this is the only schema change, so that we get another
   // schema hash to upgrade to.
   "ALTER TABLE branch_epochs RENAME TO tmp;"
@@ -863,7 +863,14 @@ migrate_add_branch_leaf_cache(sqlite3 * 
     }
 }
 
+char const migrate_add_file_sizes[] =
+    "CREATE TABLE file_sizes\n"
+    "        (\n"
+    "        id primary key,     -- joins with files.id or file_deltas.id\n"
+    "        size not null       -- the size of the file in byte\n"
+    "        );";
 
+
 // these must be listed in order so that ones listed earlier override ones
 // listed later
 enum upgrade_regime
@@ -950,9 +957,13 @@ const migration_event migration_events[]
 
   { "9c8d5a9ea8e29c69be6459300982a68321b0ec12",
     0, migrate_add_branch_leaf_cache, upgrade_none },
+
+  { "0c956abae3e52522e4e0b7c5cbe7868f5047153e",
+    migrate_add_file_sizes, 0, upgrade_regen_caches },
+
   // The last entry in this table should always be the current
   // schema ID, with 0 for the migrators.
-  { "0c956abae3e52522e4e0b7c5cbe7868f5047153e", 0, 0, upgrade_none }
+  { "1f60cec1b0f6c8c095dc6d0ffeff2bd0af971ce1", 0, 0, upgrade_none }
 };
 const size_t n_migration_events = (sizeof migration_events
                                    / sizeof migration_events[0]);
@@ -1029,6 +1040,7 @@ calculate_schema_id(sqlite3 * db, string
   id tid;
   calculate_ident(data(schema, origin::database), tid);
   ident = encode_hexenc(tid(), tid.made_from);
+  L(FL("calculated schema id %s") % ident);
 }
 
 // Look through the migration_events table and return a pointer to the entry
============================================================
--- database_check.cc	beb0e60421b8bf7552071a9af31ef6630cfd3b51
+++ database_check.cc	ae542ecb0c5c74e73bc900a9307de9ceaaffc5a0
@@ -68,9 +68,10 @@ struct checked_file {
 
 struct checked_file {
   bool found;           // found in db, retrieved and verified sha1 hash
+  bool size_ok;         // recorded file size is correct
   size_t roster_refs; // number of roster references to this file
 
-  checked_file(): found(false), roster_refs(0) {}
+  checked_file(): found(false), size_ok(false), roster_refs(0) {}
 };
 
 struct checked_roster {
@@ -170,6 +171,11 @@ check_files(database & db, map<file_id, 
       file_data data;
       db.get_file_version(*i, data);
       checked_files[*i].found = true;
+
+      file_size stored_size, calculated_size;
+      calculated_size = data.inner()().size();
+      db.get_file_size(*i, stored_size);
+      checked_files[*i].size_ok = stored_size == calculated_size;
       ++ticks;
     }
 
============================================================
--- cmd_files.cc	e51fa86b305fd51c79fac69f57e31e54810efc1e
+++ cmd_files.cc	2edcbe2df9fb1890c646a386be837879315f0995
@@ -24,11 +24,13 @@
 #include "database.hh"
 #include "work.hh"
 #include "roster.hh"
+#include "lexical_cast.hh"
 
 using std::cout;
 using std::ostream_iterator;
 using std::string;
 using std::vector;
+using boost::lexical_cast;
 
 // fload, fmerge, and fdiff are simple commands for debugging the line
 // merger.
@@ -338,6 +340,32 @@ CMD_AUTOMATE(get_file, N_("FILEID"),
   dump_file(db, output, ident);
 }
 
+// Name: get_file_size
+// Arguments:
+//   1: a file id
+// Added in: 13.0
+// Purpose: Prints the recorded size of the file in bytes
+//
+// Output format: A integer > 0
+//
+// Error conditions: If the file id specified is unknown or invalid prints
+// an error message to stderr and exits with status 1.
+CMD_AUTOMATE(get_file_size, N_("FILEID"),
+             N_("Prints the size of a file (given an identifier)"),
+             "",
+             options::opts::none)
+{
+  E(args.size() == 1, origin::user,
+    F("wrong argument count"));
+
+  database db(app);
+  hexenc<id> hident(idx(args, 0)(), origin::user);
+  file_id ident(decode_hexenc_as<file_id>(hident(), hident.made_from));
+  file_size size;
+  db.get_file_size(ident, size);
+  output << lexical_cast<string>(size) << "\n";
+}
+
 // Name: get_file_of
 // Arguments:
 //   1: a filename
============================================================
--- migrate_ancestry.cc	2b91880e4558527585bccbba0a85705f6fda6304
+++ migrate_ancestry.cc	351c11cd194241e30554fa81c2f4b7425e76daf2
@@ -993,6 +993,7 @@ regenerate_caches(database & db)
 
     db.delete_existing_rosters();
     db.delete_existing_heights();
+    db.delete_existing_file_sizes();
 
     vector<revision_id> sorted_ids;
     allrevs_toposorted(db, sorted_ids);