Merge pull request #47 from rvdheij/master

Solution with REXX filter in CMS Pipelines

05-pipeline/tf-05.rexx

@@ -0,0 +1,26 @@
+/* WORDFREQ REXX     Exercises in Programming - CMS Pipelines style  */
+/*                   Author: Rob van der Heij, 26 Apr 2019           */
+
+/* Use:                                                              */
+/* PIPE (end \) < pridenp txt  | w: wordfreq | cons                  */
+/*            \ < stop_wor txt | w:                                  */
+
+
+alpha = 'A-I a-i J-R j-r S-Z s-z a-i a-i j-r j-r s-z s-z'
+
+'callpipe (end \ name WORDFREQ.REXX:6)',
+   '\ *: ',
+   '| xlate *-* 00-FF blank' alpha ,  /* Keep words in lower case    */
+   '| split ',                        /* .. on separate records      */
+   '| locate 2',                      /* At least 2 letters long     */
+   '| l: not lookup',                 /* Drop all stop words         */
+   '| sort count',                    /* Count the remaining words   */
+   '| sort 1.10 d ',                  /* Sort on number of hits      */
+   '| take 25',                       /* Take top-25                 */
+   '| spec 11-* 1 , - , nw 1.10 strip nw ',    /* Make pretty layout */
+   '| *:',
+   '\ *.input.1: ',                   /* Read stop words             */
+   '| split ,',                       /* .. as one word per line     */
+   '| l:'                             /* into lookup table           */
+
+return rc * ( rc <> 12 )