Question

I have a need to import a CLOB and BLOB objects exported from other DBMS to PostgreSQL. Such LOBs are a separate files whose path references are placed in a CSV file. The structure of table is (id int PK, clobcol text).
Here is the example structure:

• data.csv

  1|some_clob.dat
  2|some_clob.dat
  

• configuration file: pgloader.conf

  [pgsql]
  host = localhost
  port = 5432
  base = test
  user = some_user
  pass = 
  
  log_file            = d:/tmp/pgloader.log
  log_min_messages    = DEBUG
  client_min_messages = WARNING
  lc_messages         = C
  
  client_encoding = 'utf-8'
  copy_every      = 5
  commit_every    = 5
  #copy_delimiter  = %
  
  null         = ""
  empty_string = "\ "
  
  [clob]
  table           = clob_table
  format          = text
  filename        = D:/pgloader-2.3.2/test/data.csv
  field_sep       = |
  quotechar       = "
  columns         = id:1, clobcol:2
  index           = id:1
  blob_columns    = clobcol:2:ifx_clob
  

For that import I’ve decided to use PGloader tool (Python script, I’m on Windows BTW). But when I execute:

pgloader.py -c pgloader.conf

I get the error in log file:

Warning: column 'clobcol' is not a valid blob reference: d:/pgloader-2.3.2/test/clob.out
need more than 1 value to unpack

COPY error on lines 1

and the record lands in database in form of: (id: 1, clobcol: ‘some_clob.dat’).
I have no idea what’s wrong and how to correct that. Can you give me any suggestions?
Thank you

Regards

Answer 1

Resolved: missing range of CLOB data to copy.
E.g:

    1|0,25,some_clob.dat