This can be done by using a task template in the import. It is possible to

• protect tags in segments, that contain a valid xml string
• protect HTML5-tags in segments, even if the segment is no valid HTML5 or xml.
• protect arbitary strings via PHP-PCRE-regular expressions (it is chooseable, if the regex is executed before tag-protection, after tag-protection or both). The user can define as many regex as he needs to.

The order of protection in CSV-files is as follows:

1. Protect MQM tags within the CSVs to ensure that they are imported and destroyed by one of the next steps
2. Protect everything per REGEX, that is marked in the task-template as to be protected by REGEX BEFORE tag-parsing
3. if tag protection in the task template is enabled
   • If a segment is valid XML, protect all Tags
   • If segment is not valid XML: Protect all known HTML5 tags
4. Protect everything per REGEX, that is marked in the task-template as to be protected by REGEX AFTER tag-parsing
5. Protect Whitespace (linebreaks and multiple spaces
6. Protect all XML-special-chars that are left by encoding them as entities

Please see task templates for more information.