I coded it in 'C' and the performance is about ~60X faster than the regular expression solution. And handles nested quotes. Send me a note if this is something worth finishing up and posting somewhere. ####### Test txt=""" "Bond, James", 13, 45 'That\'s "absolutely" possible.', 1999, 8, 30 """ buf=txt buf=buf*50000 print len(buf) t1=time.time() l=split(buf) print len(l) print "Time: %.2f"%(time.time()-t1) print l[4] print l[5] #### Output: 3400000 # 3.4meg input 100000 # Number of records Time: 1.91 # Runtime # Some output ['Bond', ' James', '13', '45'] ['That\'s "absolutely" possible.', '1999', '8', '30'] --Darrell ----- Original Message ----- From: Li Dongfeng <mavip5@inet.polyu.edu.hk> To: Max M <maxm@normik.dk> Cc: <Zope@zope.org> Sent: Thursday, November 04, 1999 11:55 PM Subject: Re: [Zope] Comma delimited file and the re python object
I have got a working solution for this. I have done this to read in user supplied datasets.
The solution is ugly, slow, but works for me.