Today, I had to identify all rows on a given block. I used DBMS_ROWID, but initially I wasn’t using it correctly. I figured out what I should have been doing (below)…<\/p>\n
\r\nselect work_id from xwc.frbr_holding_display partition(FRBR_HOLDING_DISPLAY_P12) where dbms_rowid.ROWID_BLOCK_NUMBER(rowid,'BIGFILE') > = 8566791\r\n<\/pre>\n…but in the meantime, I also ended up dumping the blocks and writing the following python script. This assumes the first column is a primary key or some other kind of unique identifier for the rows…<\/p>\n
\r\nalter system dump datafile &1 block &2;\r\ngrep \"col 0\" trace_file > tmp.txt\r\npython blockdump.py tmp.txt\r\n<\/pre>\n…where blockdump.py is below…<\/p>\n
\r\n#!\/usr\/bin\/python\r\n#---------------------------------------------------------------------------\r\n#Author: Steve Howard\r\n#Date: March 23, 2009\r\n#Organization: AppCrawler\r\n#Purpose: Simple script to print integer equivalents of block dump values.\r\n#---------------------------------------------------------------------------\r\n\r\nimport fileinput\r\nimport string\r\nimport sys\r\n\r\nfor line in fileinput.input([sys.argv[1:][0]]):\r\n dep=int(string.replace(string.split(string.split(line,\"]\")[1])[0],\"c\",\"\")) - 1\r\n i=0\r\n tot=0\r\n exp=dep\r\n for col in string.split(string.split(line,\"]\")[1]):\r\n if i > 0: \r\n tot = tot + ((int(col, 16) - 1) * (100**exp))\r\n exp = exp - 1\r\n i = i + 1\r\n print line.rstrip(\"\\n\") + \" converts to \" + str(tot)\r\n<\/pre>\nYou can then validate the output by using the dump function in the database, converted to base 16…<\/p>\n
\r\nselect dump(&one_of_the_values_returned_above,16) from dual;\r\n<\/pre>\n","protected":false},"excerpt":{"rendered":"Today, I had to identify all rows on a given block. I used DBMS_ROWID, but initially I wasn’t using it correctly. I figured out what I should have been doing (below)… select work_id from xwc.frbr_holding_display partition(FRBR_HOLDING_DISPLAY_P12) where dbms_rowid.ROWID_BLOCK_NUMBER(rowid,’BIGFILE’) > =…<\/p>\n