Conversion of documents to and from SGML

There are many good reasons for adopting SGML as the method of marking up source reference information when the data is primarily of textual form. Once such a decision has been reached, then it is usually necessary to consider two data conversion issues: how to get existing documents (which may exist in several different formats) into a target SGML form (up-translation); how to produce `documents´ from the SGML source data (down-translation). `Documents´ is quoted, because in many cases the desired deliverable form of the information may actually be some type of electronic database rather than a traditional printed documented. The author points out some of the issues to be considered in planning and performing these two types of conversion functions; some methodologies for carrying them out; and mentions some of the software tools which can be used for developing specific conversion applications