# Gathering and plotting email statistics
# Last edited on 2008-05-03 14:20:14 by stolfi

PURPOSE

  The goal of this project is to extract and plot several statistics
  from Unix-format mail folders.

GLOBAL SETTINGS AND VARIABLES

  Execute:

    set maildir = ${HOME}/mail
    set outdir = ./out
    
    mkdir -p ${outdir}/froms
    
    set types = ( test sent unsorted good trash )
    set typescm = ( `echo "${types}" | tr ' ' ','` )
    echo "${typescm}"
  
COUNTING FROM LINES BY MONTH

  Prepare manually the folder lists:
  
     emacs {${typescm}}-folders.dir
     
  Place in each file the bare names of the folders of the
  corresponding type. The lists need not be disjoint.
  
  Run:

    foreach type ( ${types} )
      echo " "
      echo "=== ${type} ======================="
      set folders = ( `cat folders-${type}.dir` )
      ( cd ${maildir} && egrep -a -e '^From ' ${folders} ) \
        > ${outdir}/froms/${type}.raw
      reformat-from-lines ${outdir}/froms/${type}.raw \
        > ${outdir}/froms/${type}.frd
      cat ${outdir}/froms/${type}.frd \
        | gawk \
            ' //{ \
                dt = $1; \
                sub(/-[0-9][0-9]$/, "", dt); \
                print dt; \
              } \
            ' \
        | sort -k1,1 | uniq -c \
        > ${outdir}/froms/${type}.cts
      cat ${outdir}/froms/${type}.cts
    end

PLOTTING TRASH VS. NON-TRASH

  Create the "*.mspm" files as above.  Then run:
  
    plot-mail-per-month \
      ${outdir} \
      "2004-08" "2008-04" \
      trash good sent
    
  or 
  
    plot-mail-per-month \
      ${outdir} \
      "2004-08" "2008-04" \
      trash good unsorted 
  
COUNTING "@DCC.UNICAMP.BR" MESSAGES

    foreach type ( test trash good )
      echo " "
      echo "=== ${type} ======================="
      printf "dcc: "
      egrep -i -e 'To: *stolfi@dcc' `cat folders-${type}.dir` | wc -l
      printf "ic:  "
      egrep -i -e 'To: *stolfi@ic' `cat folders-${type}.dir` | wc -l
    end