Convert Babylon dictionary to QuickDic. It is unclear under which the Babylon "free" dictionaries are. They appear to be crowd-sourced and there is no aparent license.

babylon2quickdic.sh

sudo apt-get update
sudo apt-get install -y git python-tk tix default-jre
git clone https://github.com/ilius/pyglossary.git
cd pyglossary/

# Dictionary Details
# Created by: ADO
# Submitted to Babylon's Dictionary, Translation and Information
# Platform under the title: ADO's SPANISCH-DEUTSCH
# Number of definitions found in this dictionary: 65693
# Source Language: Spanish
# Target Language: German

# http://www.babylon.com/free-dictionaries/reference/dictionaries-thesauri/ADO%27s-SPANISCH-DEUTSCH/47378.html

wget http://dl.babylon.com/info/glossaries/B912/ADO_s_SPANISCH_DEUTSCH.BGL

# Convert to TXT tabfile
python pyglossary.pyw ADO_s_SPANISCH_DEUTSCH.BGL ADO_s_SPANISCH_DEUTSCH.txt

# Create "::" separated chemnitz format wordlist
python cleanup.py ADO_s_SPANISCH_DEUTSCH.txt > ADO_s_SPANISCH_DEUTSCH.chemnitz

# Cleanup
sed -i -e 's|;|; |g' ADO_s_SPANISCH_DEUTSCH.chemnitz

# Get stoplists from http://members.unine.ch/jacques.savoy/clef/
wget http://members.unine.ch/jacques.savoy/clef/germanST.txt
wget http://members.unine.ch/jacques.savoy/clef/spanishST.txt

wget http://dictionarypc.quickdic-dictionary.googlecode.com/git/custom_dictionary/DictionaryBuilder.jar

# Convert
java -Xmx512m -jar DictionaryBuilder.jar \
  --dictOut=ADO_s_SPANISCH_DEUTSCH.quickdic \
  --lang1=ES \
  --lang2=DE \
  --lang1Stoplist=spanishST.txt \
  --lang2Stoplist=germanST.txt \
  --dictInfo="" \
  --input1=ADO_s_SPANISCH_DEUTSCH.chemnitz \
  --input1Name="" \
  --input1Charset=UTF8 \
  --input1Format=chemnitz \
  --input1FlipColumns=false 

#####

# Similarly for French:

# http://www.babylon.com/free-dictionaries/reference/dictionaries-thesauri/ADO%27s-FRENCH-GERMAN/45238.html

wget http://members.unine.ch/jacques.savoy/clef/frenchST.txt
wget http://dl.babylon.com/info/glossaries/B0B6/ADO_s_FRENCH_GERMAN.BGL

java -Xmx512m -jar DictionaryBuilder.jar \
--dictOut=ADO_s_FRENCH_GERMAN.quickdic \
--lang1=FR \
--lang2=DE \
--lang1Stoplist=frenchST.txt \
--lang2Stoplist=germanST.txt \
--dictInfo="" \
--input1=ADO_s_FRENCH_GERMAN.chemnitz \
--input1Name="" \
--input1Charset=UTF8 \
--input1Format=chemnitz \
--input1FlipColumns=false 

cleanup.py

#!/usr/bin/python
# -*- coding: utf-8 -*-

# This is specifically for DE-EN and EN-DE so far; needs to be extended manually for other languages

import argparse
import re

parser = argparse.ArgumentParser()
parser.add_argument("filename")
args = parser.parse_args()

with open(args.filename) as f:
    notags = f.read()

notags = notags.replace("\t", " :: ")
notags = notags.replace(" <br>\\n", ", ")
notags = notags.replace("<br>\\n", " ")
notags = notags.replace("|", ", ")

# Remove all tags
notags = re.sub(r'(<.*>)', r'', notags)

# Remove all brackets
notags = re.sub(r'(\[.*\])', r'', notags)
notags = re.sub(r'(\(.*\))', r'', notags)

# Mark all persons for deletion
# Remove all brackets
notags = re.sub(r'\d\d\d\d-\d\d\d\d\)', r'XXXPERSONXXX', notags)
notags = re.sub(r'\(geboren \d\d\d\d\)', r'XXXPERSONXXX', notags)
notags = re.sub(r'\(geb. \d\d\d\d\)', r'XXXPERSONXXX', notags)
notags = re.sub(r'\\d\d\d\d geboren\)', r'XXXPERSONXXX', notags)

notags = notags.replace("  ", " ")
notags = notags.replace(" , ", ", ")

for line in notags.split("\n"):
    if("::" in line):
        # We do not want Babylon comments
        if "##" in line:
            continue
        # We do not want surnames
        if "Nachname" in line or 'XXXPERSONXXX' in line or 'geboren' in line or "family name" in line:
            continue
        # We do not want first names
        if "Vorname" in line or "first name" in line:
            continue
        # We do not want brand names
        if "Markenname" in line or "®" in line or "Inc." in line or "Ltd." in line or " Corporation" in line or " Limited" in line or "sches Unternehmen" in line:
            continue

        # We do not want long-winded explanations
        # if line.split("::")[1].count (',') > 2:
        #     print "shortened -->" 
        #     line = "".join(line.split("::")[0]) + " :: " + line.split("::")[1].split(',')[0] + ", " + line.split("::")[1].split(',')[1]

        # There must be as may closing as opening
        if line.count ('(') > line.count (')'):
            print "added ) -->" 
            line = line + ")" 

        print line

lingoes2quickdic.sh

# http://www.lingoes.net/en/dictionary/dict_down.php?id=79ACB1ECC9C3D044BFAE9961C2B9E1B5

# Extract lingoes format
git clone https://github.com/librehat/kdictionary-lingoes.git
sudo apt-get install cmake qt5-default g++
cmake .
make
echo "y" | ./kdictionary-lingoes -i ../Downloads/ADO*ld2 -o output.txt 

# Cleanup
sed -i -e 's| = | :: |g' ../Downloads/output.chemnitz
sed -i -e 's|;|; |g' ../Downloads/output.chemnitz

# Convert
java -Xmx512m -jar DictionaryBuilder.jar \
  --dictOut=ADO_s_ITALIAN_GERMAN.quickdic \
  --lang1=IT \
  --lang2=DE \
  --lang1Stoplist=italianST.txt \
  --lang2Stoplist=germanST.txt \
  --dictInfo="" \
  --input1=output.chemnitz \
  --input1Name="" \
  --input1Charset=UTF8 \
  --input1Format=chemnitz \
  --input1FlipColumns=false