Files and directories
STEPcode (SC) includes the class libraries, some of the most widely used EXPRESS schemas, some tools to work with EXPRESS, and support libraries for those tools. Two of the tools can create schema-specific libraries that are used with the class libraries. There are also some test files and programs.
The schema-specific libraries handle specialized data structures described in the schema, allowing bi-directional translation between an in-memory representation of the data and one suitable for serialization. The in-memory representation stores the data in SDAI objects (below).
The class libraries (under src/, directories clutils, cldai, cleditor, clstepcore) are runtime libraries - that is, they are used by an application that wants to read or write STEP data.
- cldai implements many of the classes for SDAI (ISO 10303-22), the Standard Data Access Interface. Any classes not implemented in cldai are implemented in clstepcore (below).
- cleditor implements read and write for ASCII files that conform to ISO 10303-21, commonly referred to as Part 21 files or Exchange files.
- clstepcore contains a large number of classes used by cldai, cleditor, and the schema-specific libraries.
- clutils contains various generic utility functions.
The tools are fedex_plus, fedex_python, check_express, and exppp.
- check_express uses libexpress (below) to check an EXPRESS schema for errors. Formerly named express, it is the simplest possible use of libexpress.
- fedex_plus parses an EXPRESS schema using libexpress, analyzes the in-memory representation of the EXPRESS, and writes schema-specific C++ source code that is the compiled into one of the aforementioned schema-specific libraries.
- fedex_python is a work in progress. It is similar to fedex_plus but writes Python code instead.
- exppp, the EXPRESS Pretty Printer, formats EXPRESS in a way that is easy to read. The executable is currently not built because it is rarely used, but the library is still built and is used by fedex_plus and fedex_python.
The following libraries are used prior to runtime, in the process of generating source code for the schema-specific libraries. base is also used at runtime
, along with a few hashing functions in libexpress.
- base currently only contains memory management functions - implementations of malloc, calloc, realloc, free, new, and delete that warn of improper use and record the file and line number.
- libexppp formats EXPRESS so that it is easy to read. It is used to format some things that fedex_plus and fedex_python do not currently translate into C++ or Python, such as FUNCTIONs and RULEs.
- libexpress parses an EXPRESS schema in a file and builds a model of the schema in memory.
Standard (and draft standard) schemas are located in subdirectories under data/. Part 21 files that are known to conform (if any) are located in those subdirectories. See the README in that directory for details.
- p21read is the most commonly used test program. When the executable is built, it is linked to a schema-specific library. When invoked, it reads a part 21 file specified on the command line. Once the data is loaded in memory, it is written to another file. Currently, we only check that p21read reports success. It would be best to compare the input and output data, but this is complicated by the fact that variations are allowed in the form of whitespace and comments, and entities may be reorderd.
- src/test contains the p21read source as well as some other example programs that are not currently utilized.
- When CMake is configured with testing enabled, any Part 21 files found in the same directory as a schema that is to be built are automatically used for tests with p21read.
- test/ contains all other test files and scripts