Context-Sensitive Spelling Checker (csSpell) is a spelling checker that examines words and their context to detect potential misspellings instead of using a spelling dictionary (the cs in csSpell stands for context sensitive). csSpell can scan plain text files; files that contain non-text elements such as SGML, HTML, or XML tags; and even files that contain program code.

How does it work?

csSpell scans a group of files, analyzes each word in each file, and reports potential inconsistencies. Instead of comparing each word to a list of known words (as dictionary-based spelling checkers do), csSpell compares each word to all the other words in the group of files being scanned. It can therefore identify likely misspellings that dictionary-based spell checkers cannot, such as

• misspellings of words not listed in any spelling dictionary (for example, product names, product acronyms, and command keywords)
• misspellings that are correct spellings of a different, unintended word (for example, mange instead of manage)

There are certain types of misspellings that csSpell cannot identify. For example, if a word is consistently misspelled, csSpell will not know to flag it. For this reason, csSpell should be used in addition to, not instead of, a dictionary-based spelling checker.

Instead of scanning a group of files, with the new version 1.2 you can run csSpell directly against a DB2 database. This will examine all textual columns of your database and indicate the potential misspellings, together with the table and column where each potential misspelling was found.

Andreas Arning works at IBM Germany, doing software development. He joined IBM in 1986, holds a Ph.D. and a Master's degree in Computer Science, and has filed 25 patents.