Multiple Choice

The -T option allows for the selection of multiple points by pressing the T or Insert keys in the running script for each relevant point. Complete the selection by pressing Enter. A marked entry appears underlined. Remove the marker by repeatedly pressing T or Delete.

The software uses the space as a field separator, unless you specify otherwise. However, the example shown in Listing 6 (Figure 6) uses a colon, which allows you to use tools (e.g., cut) during further processing of the variable's contents, to address a column.

Figure 6: Some applications require the selection of multiple entries.

Menu Preselection

If you have no way to limit the list of menu options from the command in the pipe, filter the data when you call smenu. If necessary, use regular expressions (-e option, Listing 7). Alternatively, work with the positive hits (-i option, Listing 8).

The effect is generally the same: The unwanted entries appear in a color that offer less contrast and cannot be selected (Figure 7). When using both options simultaneously, keep in mind that exclusion has a higher priority. As a result, you might not get any items to select.

Figure 7: If the software supplying the data offers no way to filter it, use smenu functions to include or exclude points.

PDF Search

To browse a (quite large) portfolio of PDF files, use the pdfsearch.sh shell script (Listing 9). Using smenu, the small program displays the selection of hits and allows you to take further actions.

#!/bin/bash
# pdfsearch.sh
while true; do
  clear
  # Enter search term
  echo "browse PDF archive "
  echo "----------------------"
  read -p "enter search term (cancel: Enter or Ctrl+C): " sube
  if [ -z $sube ]; then
    exit
  fi
  # Evaluation list structure, case sensitivity is ignored.
  # only the number of hits is issued
  for i in $(pdfgrep -H -c -i "$sube" *.pdf); do
    number of hits = $(echo $i | cut-d\: -f2)
    if [ $number of hits-gt 0 ]; then
      hitlist=$(echo $hitlist $i)
    fi
  done
  # if hit is list empty, new loop run
  if [ -z "$hit list" ]; then
    echo "no hits"
    sleep 3
    continue
  fi
 # here comes smenu
  while true; do
    file=$(echo $hitlist | tr "\ " "\n" | smenu -d -n20 -c -m "select file for display (q to cancel):" | cut -d\: -f1)
    # If cancel with [q], delete hit list and
    # new loop run
    if [ -z $file ]; then
      unset hit list
      break
    fi
    # Evince and Qpdfview support the transfer
    # a search term when launching on Shell
    evince --find "$sube" $file
    # continue or finished?
    Continue= $(echo "select done" | smenu -d -m "continue with" -s /A)
    if [ "$continue" = "done" ]; then
      break
    fi
  done
done

The script only runs under Bash because it uses the read command. Either assign the appropriate rights to run it directly, or call it with:

bash pdfsearch.sh

(For PDFs, check out qpdfview or Evince: Both accept a search term at launch.)

When launching, enter the search term (Figure 8), which the script then transfers to pdfgrep. It shows only the number of hits, ignoring case sensitivity. Any hits are provided for selection via smenu.

Figure 8: After entering a search term, the search launches and provides a selection of hits.

If you have selected a file to display, the script browses the search term in the PDF viewer. It then highlights it in the document (Figure 9). After closing the display program, either reuse the selection or enter a new search term (Figure 10).

Figure 9: Certain programs allow you to transfer a term when launched, which the software then highlights in the document.

Figure 10: After viewing a PDF document, the script gives you the option of displaying another file or launching a new search.