1. 

   If you are doing a full database import, just pre-create the tablespaces like the soourceDB.
   do a full import. It will work.
   personally i dont do this.
   if The number of schema/users to be ported are less, do the import schema-by-schema.
   just recreate the users in targetDB with minumum roles.
   the grants can be given during the import.
   or
   if u want to generate the script from dump file you can also do it.
   
   -- how to generate script from dump file
   -- this will not DO the import. but just goes through the whole process of reading the
   -- dumpfile and just gives some readable text.
   imp dba/password full=y file=somedumpfile.dmp show=y log=somelogfile.log

   now if u open the somelogfile.log, you can see all the details you want.
   -- typical entries in somelogfile.log would be like

 "CREATE TABLESPACE "USERS" BLOCKSIZE 8192 DATAFILE  '/u01/app/oracle/oradata"
 "/mutation/users01.dbf' SIZE 279183360       AUTOEXTEND ON NEXT 1310720  MAX"
 "SIZE 32767M, '/u01/app/oracle/oradata/mutation/users02.dbf' SIZE 10485760  "
 "     EXTENT MANAGEMENT LOCAL  AUTOALLOCATE  ONLINE PERMANENT  NOLOGGING SEG"
 "MENT SPACE MANAGEMENT AUTO"

"CREATE USER "TEST" IDENTIFIED BY VALUES '7A0F2B316C212D67' DEFAULT TABLESPA"
 "CE "USERS" TEMPORARY TABLESPACE "TEMP""

   -- how to create the users in new targetDB with same password like in sourceDB
   -- (if you dont know the passcode in source).
   --  do the import with show=y and log the output.
   --  If you are createing the user TEST, copy the corresponding entry and run 
   --  it in targetDB.
"CREATE USER "TEST" IDENTIFIED BY VALUES '7A0F2B316C212D67' DEFAULT TABLESPACE "USERS" TEMPORARY TABLESPACE "TEMP""

   
2.
  if there is too much volume of data,
  it is recomended to load the data first and then rebuild the indexes manually.
  so during import 
  -- to load just table and data and NO INDEXES
  imp dba/password full=y file=somedumpfile.dmp indexes=N
  -- how to recreate the indexes in targetDB?
  -- method 1.
  -- get the DDL from the dump file as shown above.
  -- run ddl against the targetDB.
  -- method 2.
  -- works only in 9i, using dbms_metadata.
  -- run the script ddl.sql againts the sourcedb and spool the output. 
  -- you can get the ddl of indexes. now run the same spooled output against the targetDB.
	

dbadmin@republic_lawp1 > @ddl.sql

  CREATE INDEX "DBADMIN"."LAWPSTATS" ON "DBADMIN"."LAWPSTATS" ("STATID", "TYPE",
 "C5", "C1", "C2", "C3", "C4", "VERSION")
  PCTFREE 10 INITRANS 2 MAXTRANS 255
  STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645
  PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1 BUFFER_POOL DEFAULT)
  TABLESPACE "SYSTEM"
 ;

  CREATE UNIQUE INDEX "DBADMIN"."SYS_IL0000028549C00003$$" ON "DBADMIN"."XMLTABL
E" (
  PCTFREE 10 INITRANS 2 MAXTRANS 255
  STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645
  PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1 BUFFER_POOL DEFAULT)
  TABLESPACE "SYSTEM"
  PARALLEL (DEGREE 0 INSTANCES 0)
 ;

--
-- this the sourcecode for ddl.sql
-- 
dbadmin@republic_lawp1 > get ddl.sql
  1  set long 500000000000000
  2  SET HEAD off
  3  SELECT DBMS_METADATA.GET_DDL('INDEX',D.INDEX_NAME)||';' FROM user_indexes D;