If your HTML, XML and JavaDoc or Doxygen is to some extend German (and not just English) Eclipse’s spell checker gets more bothering than a help. Here the developer gets a comprehensive German dictionary containing more than 24 800 word forms including OO, Java and process control.

Get and install

Download: weinert-automation.de/software/eclipse/de_DE.dic.
Store the file where you like, e.g. at: C:\util\eclipse\dropins

Install in Eclipse as “User defined directory” at: Window>Preferences>General>Editors>Texteditors>spelling.

Hint: Eclipse probably requires write access for its users to the file C:\util\eclipse\dropins\de_DE.dic.
On Windows this is done in the Explorer or by cacls in the shell. Under Linux, you get there by something like

    sudo cp '/where/it/is/de_DE.dic' '/usr/lib/eclipse/dropins'
    sudo chmod 777 /usr/lib/eclipse/dropins/de_DE.dic

Hint 2: Of all Eclipse editors the “Text Editor” has the best spell checking integration. So it should be used from time to time if your .java, .xml, .php, .html or whatever contains a considerable amount of readable text.

Structure and content

The most simple form of a .dic file for Eclipse is used here

  • one word respectively word form per line
  • no repetitions, not even for just first capital letter
    If “wandern” is contained, Eclipse will accept “Wandern” anyway. Hence, the Frame4J tool UCopy (option -eclipDic) will drop the upper case doublet.
  • no (abbreviation) dots at the end (Eclipse accepts “etc.” in the text by a .dic entry “etc”)
  • no hyphenated words
    Eclipse checks “Baden-Württemberg” against the two words “baden” and “Württemberg”. Baden is also a verb (to bathe), hence the lower case variant, only.
  • not sorted (Eclipse seems to sort on load anyway. Other information?)
     

Modify or make it yourself

Frame4J‘s tool UCopy will bring a text file to this form by

    java UCopy source.txt -eclipDic -v destination.dic
  • doing those transformations
  • separate word in single lines
  • delete a trailing dot
  • delete single letters and doublets (also those distinguished by first capital only)
  • delete words containing digits or upper case characters only.

This tool greatly simplifies the making and adding to such .dic files. Without bothering about line structure or word forms already contained one just adds correctly (!) spelled words and lets UCopy with -eclipdic option crunch it.