XMP Specification September 2005 103 Embedding XMP Metadata in Application Files PostScript and EPS 5 PostScript and EPS XMP metadata can be placed in PostScript® or EPS files, for use in either PostScript or PDF workflows. This section describes how to place XMP into PostScript or EPS for both the outer document level (main XMP) and for internal objects such as an image (object XMP). It also specifically discusses issues involving Acrobat Distiller, since workflows often use Distiller to produce PDF from PostScript and EPS. NOTE: This does not imply that use of Distiller is necessary, or that other application issues do not exist. There are three important “flavors” of PostScript files that can affect how XMP is written, found, and used. They are: ● DSC PostScript (or just “PostScript”): PostScript conforming to the DSC conventions defined in Appendix G of the PostScript Language Reference. ● Raw PostScript: PostScript following no particular structural conventions. The use of raw PostScript is discouraged. As mentioned in “Ordering of Content” on page 104, a special DSC comment is required to support fast and reliable location of the main XMP. ● EPS: PostScript conforming to the EPS conventions defined in Appendix H of the PostScript Language Reference. EPS is a subset of DSC PostScript. Because of common usage issues, document-level XMP should be written differently for PostScript and EPS. Object-level XMP is written identically for PostScript and EPS. The XMP in a PostScript/EPS file must be encoded as UTF-8. Document-Level Metadata As with any file format, locating contained XMP in PostScript or EPS is most reliably done by fully processing the file format. For PostScript, this means executing the PostScript interpreter. Packet scanning is not reliable whenever a file contains multiple XMP packets, or object XMP without main XMP. It is often worthwhile to find the main XMP and ignore (at least temporarily) object XMP. Interpretation of the entire PostScript file to locate the main XMP can be very expensive. A hint and careful ordering are used to allow a combination of XMP packet scanning and PostScript comment scanning to reliably find the main XMP. To write document-level metadata in PostScript, an application must: ● Write the %ADO_ContainsXMP comment as described under “Ordering of Content” on page 104. ● Write the XMP packet as described under “Document-Level XMP in PostScript” on page 105. To write document-level metadata in EPS an application must: