et

#!/bin/sh

html_entities_to_ascii () {
  echo -n "$1" | sed -E "s/&amp\;/\& /g"
}

get_abbreviations () {
  ABBRS_URL="https://www.abbreviations.com/$1"
  USERAGENT='Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:92.0) Gecko/20100101 Firefox/92.0'

  ABBREVIATIONS=$(
    echo -n "$(curl -s -A "$USERAGENT" $ABBRS_URL)" |\
    sed -E "/<table class=\"table tdata no-margin\">.+<\/table>/ !d" |\
    perl -pe "s/(<td class=\"tal dx fsl\">.+?<\/td>)/\n\$1\n/g" |\
    sed -E "/^<td class=\"tal dx fsl\">/ !d;\
            s/<td class=\"tal dx fsl\"><p class=\"desc\">(.*)<\/p><p class=\"path\">(.*)<\/p><\/td>/[\2] \1/;\
            s/<\/{0,1}a[^>]*>//g;\
            s/&raquo\;/»/g;\
            s/ -- and more...//g"
  )

  ABBREVIATIONS=$(html_entities_to_ascii "$ABBREVIATIONS")

  echo "$ABBREVIATIONS"
}

get_antonyms () {
  WORD="$1"

  spell $WORD 0

  ANTONYMS_URL="https://words.bighugelabs.com/$WORD"
  ANTONYMS_RES="$(curl -s $ANTONYMS_URL)"

  ANTONYMS=""
  declare -a ANT_TYPES=("noun" "verb" "adjective" "adverb")
  for ANT_TYPE in "${ANT_TYPES[@]}"; do
    ANTONYMS+=$(
cat << EOF

$(echo -n "$ANTONYMS_RES" |\
sed -E "/<h3>$ANT_TYPE<\/h3>/,/(<h3>)|(<\/div>)/ !d;\
/<h4>antonyms<\/h4>/,/<\/ul>/ !d;\
/<li>/ !d;\
s/\s*<li><a.*>(.*)<\/a><\/li>/($ANT_TYPE) \1/g"
)
EOF
)
  done

  ANTONYMS=$(echo -n "$ANTONYMS" | sed '1 d')

  echo "$ANTONYMS"
}

define () {
  spell $1 0

  WORDNIK_URL="https://www.wordnik.com/words/$WORD"
  USERAGENT='Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:92.0) Gecko/20100101 Firefox/92.0'

  DEFINITIONS=$(
    echo -n "$(curl -s -A "$USERAGENT" $WORDNIK_URL)" |\
    sed "/guts active/,/<\/ul>/!d;\
         /<li>/!d" |\
    sed -E "s/\s*<li><abbr title=\"partOfSpeech\">(.+)<\/abbr> <i>(.*)<\/i> (.*)<\/li>/(\1) [\2] \3/;\
            s/ \[\]//g;\
            s/\s{2,}/ /g;\
            s/<[^\;][^>]*>//g;"
  )

  DEFINITIONS=$(html_entities_to_ascii "$DEFINITIONS")

  echo "$DEFINITIONS"
}

pronounce () {
  WORD="$1"
  SOURCE="$2"

  spell $WORD 0

  SOUND_URL=""

  if [[ "$SOURCE" == "macmillan" ]]; then
    USERAGENT='Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:92.0) Gecko/20100101 Firefox/92.0'

    SOUND_URL=$(
      curl -s -A "$USERAGENT" "https://www.wordnik.com/fragments/sounds/$WORD" | \
      tac | \
      sed -n '/macmillandictionary/, /play_sound/ p' | \
      tail -n 1 | \
      sed -E 's/.*doc-fileUrl="([^"]+)".*/\1/; s/&amp\;/\&/g'
    )
  else
    SOUND_URL=$(
      curl -s "https://api.dictionaryapi.dev/api/v2/entries/en/$WORD" | \
      grep -oE '"audio":"[^"]+"' | \
      cut -d ':' -f 2- | \
      sort | \
      uniq | \
      sed -E 's/^"|"$//g' | \
      grep "$WORD-$SOURCE\.mp3"
    )

    if [[ -z "$SOUND_URL" ]]; then
      echo "Invalid source. See available ones with -lst|--list-sources."
      exit 1
    fi
  fi

  play -q -t mp3 "$SOUND_URL"
}

list_pronunciation_sources () {
  WORD="$1"

  spell $WORD 0

  DICTAPI_SOURCES=$(
    curl -s "https://api.dictionaryapi.dev/api/v2/entries/en/$WORD" | \
    grep -oE '"audio":"[^"]+"' | \
    cut -d ':' -f 2- | \
    sed -E "s/.*$WORD-(.+)\.mp3\"$/\1/" | \
    sort | \
    uniq
  )

  echo "- macmillan (default)"
  echo "$DICTAPI_SOURCES" | sed 's/^/- /'
}

spell () {
  agrep -iyB "$1" /usr/share/wordlists/et_wordlist.txt | grep -iw "$1" > /dev/null

  FOUND=$?
  PROCEED_TO_DEFINE=$2
  if [ $FOUND -eq 0 ]; then
    if [ $PROCEED_TO_DEFINE -ne 0 ]; then
      echo "\"$1\" appears to be spelled correctly"
    fi
  else
    SUGGESTIONS=$(agrep -iyB "$1" /usr/share/wordlists/et_wordlist.txt | awk '{print length, $0}' | sort -n | cut -d " " -f2-)

    echo "$SUGGESTIONS"
    exit 1
  fi
}

get_synonyms () {
  WORD="$1"

  spell $WORD 0

  SYNONYMS_URL="https://words.bighugelabs.com/$WORD"
  SYNONYMS_RES="$(curl -s $SYNONYMS_URL)"

  SYNONYMS=""
  declare -a SYN_TYPES=("noun" "verb" "adjective" "adverb")
  for SYN_TYPE in "${SYN_TYPES[@]}"; do
    SYNONYMS+=$(
cat << EOF

$(echo -n "$SYNONYMS_RES" |\
sed -E "/<h3>$SYN_TYPE<\/h3>/,/<\/ul>/ !d;\
/<li>/ !d;\
s/\s*<li><a.*>(.*)<\/a><\/li>/($SYN_TYPE) \1/g"
)
EOF
)
  done

  SYNONYMS=$(echo -n "$SYNONYMS" | sed '1 d')

  echo "$SYNONYMS"
}

usage () {
  cat << EOF
Usage: et OPTION...

Options:
  -abr, --abbreviations <abr>
    Print what <abr> might stand for from abbreviations.com.

  -ant, --antonyms <word>
    Print antonyms for <word> from bighugelabs.com.

  -def, --define <word>
    Print definitions for <word> from wordnik.com if correctly spelled, otherwise print spell suggestions.

  -pro, --pronounce <word>
    Play pronunciation of <word>.

    Defaults to American English pronunciation (listed as macmillan in sources list) from wordnik.com (which in turn depends on macmillandictionary). Other pronunciation sources/accents (from dictionaryapi.dev) can be listed with -lst|--list-sources and specified with -src|--source (must come after -pro|--pronounce).

  -lst, --list-sources <word>
    List available pronunciation sources/accents for <word>.

  -src, --source <source>
    Specify the <source> to use for pronunciation. This option must come after -pro|--pronounce.

  -spl, --spell <word>
    Print spell suggestions for <word> from wordlist if not spelled correctly (exits with 1 as exit code), otherwise print a message indicating that <word> is spelled correctly.

  -syn, --synonyms <word>
    Print synonyms for <word> from bighugelabs.com.

  -h, --help
    Print this help message and exit
EOF
}

if [[ "$@" ]]; then
  OPTION=$(echo "$@" | cut -d " " -f 1)
  WORD=$(echo "$@" | cut -d " " -f 2)

  if [ "$OPTION" == "-h" ] || [ "$OPTION" == "--help" ]; then
    usage
  else
    SPACE_COUNT=$(echo "$@" | tr -cd " \t" | wc -c)

    if [ "$WORD" == "" ] || [ $SPACE_COUNT -lt 1 ]; then
      echo "You need to specifiy an option and a word"
      exit 1
    else
      WORD=$(echo "$WORD" | awk '{print tolower($0)}')

      if [ "$OPTION" == "-def" ] || [ "$OPTION" == "--define" ]; then
        define $WORD
      elif [ "$OPTION" == "-spl" ] || [ "$OPTION" == "--spell" ]; then
        spell $WORD 1
      elif [ "$OPTION" == "-abr" ] || [ "$OPTION" == "--abbreviations" ]; then
        get_abbreviations $WORD
      elif [ "$OPTION" == "-syn" ] || [ "$OPTION" == "--synonyms" ]; then
        get_synonyms $WORD
      elif [ "$OPTION" == "-ant" ] || [ "$OPTION" == "--antonyms" ]; then
        get_antonyms $WORD
      elif [ "$OPTION" == "-pro" ] || [ "$OPTION" == "--pronounce" ]; then
        SOURCE="macmillan"

        OPTION="$(echo "$@" | cut -d ' ' -f 3)"
        if [[ "$OPTION" == "-src" ]] || [[ "$OPTION" == "--source" ]]; then
          SOURCE="$(echo "$@" | cut -d ' ' -f 4 | awk '{print tolower($0)}')"
        fi

        pronounce "$WORD" "$SOURCE"
      elif [ "$OPTION" == "-lst" ] || [ "$OPTION" == "--list-sources" ]; then
        list_pronunciation_sources $WORD
      else
        echo "Unkown option \"$OPTION\""
        exit 1
      fi
    fi
  fi
else
  usage
fi