#!/bin/bash
# save_page_as -- Save a webpage using your browser.
# https://github.com/abiyani/automate-save-page-as
# Last Modified By: Dan Jacobson
# http://jidanni.org/
# Last Modified On: Thu Oct 18 00:17:04 2018
# Update Count    : 27

set -e
set -u
set -o pipefail

# Assert existence of xdotool to begin with
if ! xdotool --help &>/dev/null; then
    printf "ERROR: 'xdotool' is not present (or not in the PATH).
Please visit http://www.semicomplete.com/projects/xdotool/ to download it for your platform.\n" >&2
    exit 1
fi

load_wait_time=4
save_wait_time=8
dialog_wait_time=3
scriptname=$(basename $0)
destination=.
browser=${BROWSER?}
suffix=""
url=""

function print_usage() { #Not too wide. Must fit on all screens.
    cat  1>&2 <<EOF
$scriptname: Open the given url in a browser tab/window,
perform 'Save As' operation and close the tab/window.

USAGE:
   $scriptname URL [OPTIONS]

URL                      The url of the web page to be saved.

options:
  -d, --destination      Destination path. If a directory, then file is saved
        with default name inside the directory, else assumed to be full path
        of target file. Default = '$destination'
  -s, --suffix           An optional suffix string for the target file name
        Supply any dots yourself: e.g., -s .txt will add .txt
        (ignored if --destination arg is a full path).
  -b, --browser          Browser executable to be used (must be one listed in
        $scriptname source file. Default = \$BROWSER ($BROWSER)
  --load-wait-time       Number of seconds to wait for the page to be loaded
        (i.e., seconds to sleep before Ctrl+S is 'pressed'). Default = $load_wait_time
  --save-wait-time       Number of seconds to wait for the page to be saved
        (i.e., seconds to sleep before Ctrl+F4 is 'pressed'). Default = $save_wait_time
  --dialog-wait-time       Number of seconds to wait for the dialog to appear
        (i.e., seconds to sleep before ALT+n is 'pressed'). Default = $dialog_wait_time
  -h, --help             Display this help message and exit.

Note: Save to "Webpage Complete" or "HTML Only" etc.: Many browsers
remember the choice made last time you saved something.
So you might need to "prep" your browser by saving something by hand first,
if you don't like its current choice. (Maybe this program might be enhanced
one faraway day to ensure saving to the right choice...)
EOF

}


while [ $# -gt 0 ]
do
    case "$1" in
        -d | --destination)
            shift;
            destination="$1"
            shift
            ;;
        -s | --suffix)
            shift;
            suffix="$1"
            shift;
            ;;
        -b | --browser)
            shift;
            browser="$1"
            shift
            ;;

        --load-wait-time)
            shift;
            load_wait_time="$1"
            shift
            ;;
        --save-wait-time)
            shift;
            save_wait_time="$1"
            shift
            ;;
        --dialog-wait-time)
            shift;
            dialog_wait_time="$1"
            shift
            ;;
        -h | --help)
            print_usage
            exit 0
            ;;
        -*)
            printf "ERROR: Unknown option: %s\n" "${1}">&2
            print_usage
            exit 1
            ;;
        *)  if [ ! -z "$url" ]; then
                printf "ERROR: Expected exactly one positional argument (URL) to be present,
 but encountered a second one ('%s').\n\n" "${1}" >&2
                print_usage
                exit 1
            fi
            url="$1"
            shift;
            ;;
    esac
done

# Returns 1 if input param contains any non-printable or non-ascii character, else returns 0
# (Inspiration: http://stackoverflow.com/a/13596664/1857518)
function has_non_printable_or_non_ascii() {
    LANG=C
    if printf "%s" "${1}" | grep '[^ -~]\+' &>/dev/null; then
        printf 1
    else
        printf 0
    fi
}

function validate_input() {
    if [[ -z "${url}" ]]; then
        printf "ERROR: URL must be specified.\n" >&2
        print_usage
        exit 1
    fi

    if [[ -d "${destination}" ]]; then
        printf "INFO: The specified destination ('%s') is a directory path,
        will save file inside it with the default name.\n" "${destination}">&2
    else
        local basedir="$(dirname "${destination}")"
        if [[ ! -d "${basedir}" ]]; then
            printf "ERROR: Directory '%s' does not exist - Will NOT continue.\n" "${basedir}" >&2
            exit 1
        fi
    fi
    destination="$(readlink -f "$destination")"  # Ensure absolute path

    case $browser in
	google-chrome|chromium|firefox|epiphany);;
	*) echo $0: ERROR: Browser $browser is not supported. 1>&2; exit 1;;
    esac

    if ! command -v "${browser}" &>/dev/null; then
        printf "ERROR: Command '${browser}' not found. Make sure it is installed, and in path.\n" >&2
        exit 1
    fi

    local num_regexp='^.[0-9]+$|^[0-9]+$|^[0-9]+.[0-9]+$'  # Matches a valid number (in decimal notation)
    if [[ ! "${load_wait_time}" =~ $num_regexp || ! "${save_wait_time}" =~ $num_regexp ||
	   ! "${dialog_wait_time}" =~ $num_regexp ]] ; then
        printf "ERROR: --load-wait-time (='%s'),
and --save_wait_time(='%s')
and --dialog-wait-time(='%s')
must be valid numbers.\n" "${load_wait_time}" "${load_wait_time}" "${dialog_wait_time}" >&2
        exit 1
    fi

    if [[ $(has_non_printable_or_non_ascii "${destination}") -eq 1 ||
	      $(has_non_printable_or_non_ascii "${suffix}") -eq 1 ]]; then
        printf "ERROR: Either --destination ('%s') or --suffix ('%s') contains a non ascii
or non-printable ascii character(s).\n" "${destination}" "${suffix}" >&2
        printf "'xdotool' does not mingle well with non-ascii characters
(https://code.google.com/p/semicomplete/issues/detail?id=14).\n\n" >&2
        printf '!!!! Will NOT proceed !!!!\n' >&2
        exit 1
    fi
}
validate_input
##############

# Launch ${browser}, and wait for the page to load
"${browser}" "${url}" &>/dev/null &
sleep ${load_wait_time}

# Find the id for the ${browser} window
browser_wid="$(xdotool search --sync --onlyvisible --class "${browser}" | head -n 1)"
wid_re='^[0-9]+$'  # window-id must be a valid integer
if [[ ! "${browser_wid}" =~ ${wid_re} ]]; then
    printf "ERROR: Unable to find X-server window id for browser.\n" >&2
    exit 1
fi

# Activate the ${browser} window, and "press" ctrl+s
xdotool windowactivate "${browser_wid}" key --clearmodifiers "ctrl+s"

sleep ${dialog_wait_time} # Give 'Save as' dialog box time to show up

# Resolve the expected title name for save file dialog box (chrome & firefox differ in this regard)
if [[ "${browser}" == "epiphany" ]]; then
    savefile_dialog_title="Save"
elif [[ "${browser}" == "firefox" ]]; then
    savefile_dialog_title="Save as"
else
    savefile_dialog_title="Save file"
fi
# Find window id for the "Save file" dialog box
savefile_wid="$(xdotool search --name "$savefile_dialog_title" | head -n 1)"
if [[ ! "${savefile_wid}" =~ ${wid_re}  ]]; then
    printf "ERROR: Unable to find window id for 'Save File' Dialog.\n" >&2
    exit 1
fi

# Fix for Issue #1: Explicitly focus on the "name" field (works on both: gnome, and kde)
xdotool windowactivate "${savefile_wid}" key --delay 20 --clearmodifier "Alt+n"

# Check if we are using kde
is_kde=0
# Don't feel bad if DESKTOP_SESSION env variable is not present
set +u
if [[ "${DESKTOP_SESSION}" =~ ^kde-? ]]; then
    is_kde=1
fi
set -u

if [[ ! -z "${suffix}" ]]; then
    ###########################
    # Make sure that we are at correct position before typing the "suffix"
    #
    # If the user is using 'kde-plasma', then the full name of the file including the extension is highlighted
    # in the name field, so simply pressing a Right key and adding suffix leads to incorrect result.
    # Hence as a special case for 'kde-*' we move back 5 characters Left from the end before adding the suffix.
    # Now this strategy is certainly not full proof and assumes that file extension is always 4 characters long ('html'),
    # but this is the only fix I can think for this special case right now. Of course it's easy to tweak the number of
    # Left key moves you need if you know your file types in advance.
    if [[ "${is_kde}" -eq 1 ]]; then
        printf "INFO: Desktop session is found to be '${DESKTOP_SESSION}',
hence the full file name will be highlighted.\n" >&2
        printf "Assuming extension .html to move back 5 character left
before adding suffix (change accordingly if you need to).\n" >&2
        xdotool windowactivate "${savefile_wid}" key --delay 40 --clearmodifier End Left Left Left Left Left
    else
        xdotool windowactivate "${savefile_wid}" key --delay 20 --clearmodifiers Right
    fi
    set -u
    ###########################

    extraarg=""
    if [[ "${suffix::1}}" == "-" ]]; then
        extraarg="-"
    fi
    xdotool type --delay 10 --clearmodifiers "${extraarg}" "${suffix}"
fi

# Activate the 'Save File' dialog and type in the appropriate filename
#(depending on ${destination} value: 1) directory, 2) full path, 3) empty)
if [[ ! -z "${destination}" ]]; then
    if [[ -d "${destination}" ]]; then
        # Case 1: --destination was a directory.
        xdotool windowactivate "${savefile_wid}" key --delay 20 --clearmodifiers Home
        xdotool type --delay 10 --clearmodifiers "${destination}/"
    else
        # Case 2: --destination was full path.
        xdotool windowactivate "${savefile_wid}" key --delay 20 --clearmodifiers "ctrl+a" "BackSpace"
        xdotool type --delay 10 --clearmodifiers "${destination}"
    fi
fi
xdotool windowactivate "${savefile_wid}" key --delay 20 --clearmodifiers Return

printf "INFO: Saving web page ...\n" >&2

# Wait for the file to be completely saved
sleep ${save_wait_time}

case $browser in epiphany) is_kde=1;; esac

# Close the browser tab/window (Ctrl+w for KDE, Ctrl+F4 otherwise)
if [[ "${is_kde}" -eq 1 ]]; then
    xdotool windowactivate "${browser_wid}" key --clearmodifiers "ctrl+w"
else
    xdotool windowactivate "${browser_wid}" key --clearmodifiers "ctrl+F4"
fi
printf "INFO: Done!\n">&2
#Maybe also say:
# echo See: #or Proof:
# ls -l $destination