This article describes how the Unicode chars and the several Dicom charsets have been implemented in Imebra.

The support for Unicode and the Dicom charsets has been introduced in the version 0.0.24 of Imebra released on the 8th of april 2006. 

The Dicom standard supports several charsets, including the unicode (UTF8 encoded).

Imebra uses the following approach when dealing with tags that could have been encoded using one of the Dicom charsets:

• when the dataset is loaded, a special attribute in all the tags is updated with the charsets supported by the dataset.
• when a dataset is created, the application can decide the charsets that can be used in the dataset: all the new tags will use the charsets specified in the dataset
• when an application writes a value in a tag, it can change the tag's charsets: before a dataset is saved all the tag's charsets are analyzed and checked for inconsistence or invalid charsets combinations and the tag 0008,0005 is updated with the dataset's charsets.

The conversion from/to the dataset defined charsets happens when a dataHandler is created or destroyed: while the dataHandler is "alive", the strings are kept in unicode format and converted to ansi if a non-unicode function is called.

The conversion uses MultiByteToWideChars/WideCharsToMultiByte on Windows and iconv on Posix systems.

---

Bookmark this