IWordBreaker
The IWordBreaker interface is a language-specific language resource component. The word breaker parses text and identifies individual words and phrases. The word breaker is used in background processes and must be optimized for both throughput and minimal use of resources.
When to Implement
Implement this interface to create a custom word breaker for a language. Indexing Service calls the methods of this interface when it builds content indexes and runs queries.
Methods in Vtable Order
| IWordBreaker method | Description |
|---|
| Init | Initializes the IWordBreaker implementation and indicates the mode in which the component operates. |
| BreakText | Breaks text to identify words and phrases and provides the results to the WordSink and PhraseSink objects. |
| ComposePhrase | Not supported. |
| GetLicenseToUse | Gets the license information for this IWordBreaker implementation. |
Windows NT or later: Requires Windows NT 4.0 with the Windows NT 4.0 Option Pack or Windows 2000 or later.
Header: Declared in indexsrv.h.
Import Library: User defined.
Remarks
Word breaker components for Indexing Service run in the Local Security context. They should be written to manage buffers and to stack correctly. All string copies must have explicit checks to guard against buffer overruns. You should always verify the allocated size of the buffer and test the size of the data against the size of the buffer.
See Also
Implementing a Word Breaker, Language Resource Samples, PhraseSink, Secure Code Practices, WordSink