python - Strip carriage returns / line feeds from database records before they are written to csv -

i have following code reads lines database , writes matching results csv file. problem running there carriage returns / line feeds in of fields in different rows cause csv file unusable because there bogus rows.

for example, here sample of happens when there carriage returns / line feeds in sql data , affect has on file. ... sample content of messed file:

field1|field2|field3|field4|field5 value 1|value 2|value 3|value 4|value 5 value 1|value 2|value 3|value 4|value 5 value 1|value 2|val ue 3|value 4|value 5 value 1|value 2|value 3|va lue 4|value 5 

here code writes results of sql query output file. trying strip results have carriage returns / line feeds.

    '''     while loop read each row.  if compares row[2] (updated) against last record processed     '''     latest = params #declare 'latest' variable consumption while loop     while row:         if row[2] > latest:             latest = row[2]         logger.debug("[%s] - writing row %s", correlationid, row)         writer.writerow(row)         row = cursor.fetchone()      logger.info("[%s] - last letter date %s " % (correlationid, lastprocessed))     lastprocessedlog = open(last_processed_logfile , 'wt')     lastprocessedstring = str(latest)     lastprocessedstring = lastprocessedstring[0:19]     lastprocessedlog.write(lastprocessedstring)     lastprocessedlog.close()      conn.close()     ofile.close()     logger.info("[%s] - copying %s root loadbackflow %s", correlationid, writefile, outfile)     shutil.copyfile(writefile, outfile)     logger.info("[%s] - moving %s completion folder %s", correlationid, writefile, completionfolder)     shutil.move(writefile, completionfolder) 

i have tried changing writer.write(row) line include replace error. similarly, errors when trying use replace row = row.replace("\r\n", "") ... have pasted attempts , corresponding errors below.

any insights on how can strip carriage returns / line feeds @ time being read sql query results data file appreciated.

thanks in advance! :)

# attempt1: writer.writerow(row).replace("\r\n", "") # error: unexpected error: 'nonetype' object has no attribute 'replace'  # attempt2: row = row.replace("\r\n", "") #error: unexpected error: 'tuple' object has no attribute 'replace'  #attempt3: row = row.replace("\r", "") row = row.replace("\n", "") #error: unexpected error: 'tuple' object has no attribute 'replace' 

programming permutation known antipattern... db.cursor.fetchone() returns tuple, - error message tells - has no replace() method (what semantic of (1, 2, 3).replace("a string", "another string") ?), , csv.writer.writerow() return none (and since have written row, point of trying modify afterward anyway ?).

so, make long story short, have 2 options:

1. modify the strings in row before passing writerow()
2. use csv format allow escape newlines

i don't know how plan on using generated csv, solution #2 still best approach if possible - these newlines might here reason - , requires passing right arguments when instanciating csv.writer.

if that's definitly not option, solution #1 needs bit more work, tuples , strings immutable:

def preprocess_item(item):     if isinstance(item, str):         return item.replace("\n", " ").replace("\r", " ")         return item  def preprocess_row(row):     return tuple(preprocess_item(item) item in row)  def yourfunction(whatever):     ### code here     row = cursor.fetchone()     row = preprocess_row(row)     writer.writerow(row)