If your text always starts,ends with a ", then you can probably use below transformations: text.map(lambda x:(1,x)).reduceByKey(lambda x,y:' '.join([x,y])).map(lambda x:x[1][1:-2]).flatMap(lambda x:x.split('" "')).collect() where text represents an object that reads below lines "The csv file is about to be loaded into Phoenix" "another line to parse" like: ['"The csv','file is about','to be loaded into','Phoenix",'"another line','to parse"'] While loading lines are split on a \n. This reduces them once again to a single line and splits on " ", so you get a list with portions between successive ".
... View more
