XMLmind   
    Home    Products    Services    Resources
Products
 
FO Converter
Spell Checker
    Description
    Why Choose XSC?
    Evaluate
    Buy
    Upgrade
    Documentation
        Changes
    Support
    Dictionary Builder
    User-Contributed Dictionaries
XML Editor

 

Site Map
Contact

1.3 (March 16, 2007)

Enhancements:

  • Previously German dictionary de.dar contained two possible spellings for a number of words. For example: it contained both "Fluß" (traditional spelling) and "Fluss" (spelling after the Rechtschreibreform).

    Now de.dar is supposed to only contain words compliant with the German spelling reform. Alternate German dictionary de-alt.dar has been created for those who still need to use the traditional spelling.

  • Localized all the dialog boxes to French, German and Spanish.
  • When the name of a dictionary follows ISO639 rules (e.g. "en", en-US, "fr", "fr-CH"), the English and localized display names of this dictionary are automatically inferred. For example: the display name for dictionary "fr-CH" is inferred to be "French (Switzerland)" in English and "Français (Suisse)" in French.
  • Added the -verbose and the -dump options to the Dictionary Builder.
  • JavaTM 1.3 is no longer supported. JavaTM 1.4.1 (or a more recent version: 1.4.2, 1.5, 1.6, etc) is required to compile and run XMLmind Spell Checker.

    Therefore XMLmind Spell Checker now natively uses interface java.lang.CharSequence and its implementations: java.lang.String and java.lang.StringBuffer. This makes the spell checker engine more efficient and easier to program.

    Unfortunately, this also has a slight impact on the APIs: all interfaces and classes used to emulate java.lang.CharSequence have been removed: com.xmlmind.spellcheck.util.CharSequence, StringCharSeq, StringBufferCharSeq.

    Important: Do not upgrade if you do not intend to update your source code! This is a simple, mechanical, change: merely changing something like:

    spellChecker.setInput(new StringCharSeq(text));

    to:

    spellChecker.setInput(text);

    On the other hand custom dictionaries that worked with 1.1 and 1.2 will work fine with 1.3.

Bug fixes:

  • A word like "Creutzfeld-Jakob" was considered to be a proper name which caused the spell checker
    • to accept "Creutzfeld-jakob", "CREUTZFELD-JAKOB", "Creutzfeld-Jakob" and "creutzfeld-Jakob",
    • and to suggest "Creutzfeld-jakob" for a mistake like "creutzfeld-jakob".

    Now this word is considered to be a mixed-case one and therefore only "Creutzfeld-Jakob" is accepted.

  • In certain cases, the -sub option of the Dictionary Builder failed to remove words from a compiled dictionary (.cdi file).
  • The "Replace All" sub-menu of the popup menu displayed by JSpellDialog.showAutoCheckMenu skipped words to be replaced if these words were found within the first 100 characters of the document.

1.2 (March 4, 2005)

This release is essentially motivated by non-technical reasons:

  • The names, prices and licences of certain products of the XMLmind Spell Checker product line have changed.
  • It is now possible to download evaluation versions of XMLmind Spell Checker SDK Edition and Web Edition (formerly Client/Server Edition). These evaluation versions are identical to the real products except that they will expire thirty days after the date of the download.
  • The distributions of all the products have been completely reorganized.
  • The web site of the XMLmind Spell Checker product line has been completely reorganized.
  • JavaTM package names have been changed from azcheck.* to com.xmlmind.spellcheck.*.

    Important: Do not upgrade if you do not intend to update your source code! This is a simple, mechanical, change. All 1.1 APIs are compatible with 1.2 APIs. Custom dictionaries that worked with 1.1 will work fine with 1.2.

Enhancements:

  • In ``on the fly'' spell-checking mode, erroneous words are now underlined using a nice red jagged line.
  • In ``on the fly'' spell-checking mode, right-clicking on an underlined erroneous word can now popup a menu which allows the user to fix the spelling error.

Bug fixes:

  • Fixed an ArrayIndexOutOfBoundsException that occurred when checking very long words (length > 100 characters).
  • The dictionary builder rejected UTF-8 files (hint files, word lists) which started with the UTF-8 BOM (Byte Order Mark: 0xEF 0xBB 0xBF). This kind of UTF-8 files is typically created using Windows 2000/XP notepad.
  • The dialog box of the Web Edition did not fix the ``duplicated words'' error by deleting the superfluous word.
  • Web Edition now works with Safari (1.2+), the standard web browser of MacOS X.

 


© 2003-2007 Pixware. Updated 2007-03-16 using qizx/open.

Java and all Java-based marks are trademarks or registered trademarks of Sun Microsystems, Inc. in the U.S. and other countries.
Acrobat and PostScript are trademarks of Adobe Systems Incorporated.