Bugzilla – Full Text Bug Listing
| Summary: | Handling of unsigned types by Java generator | ||
|---|---|---|---|
| Product: | [Modeling] eTrice | Reporter: | Klaus Birken <klaus.birken> |
| Component: | Generators | Assignee: | Project Inbox <etrice.generators-inbox> |
| Status: | CLOSED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | hrr, ts |
| Version: | 0.1.0 | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Windows XP | ||
| Whiteboard: | 0.2.0M1 | ||
For ROOM primitive types that can't be mapped to Java types an error message is printed and the code generation is terminated. Done with 8ea6a8c215986e859eb47e202c083f13c8d830b7 It still needs to be discussed whether something like the proposed "mapUnsignedTypes" option should be implemented. (In reply to comment #1) > > It still needs to be discussed whether something like the proposed > "mapUnsignedTypes" option should be implemented. I've opened bug 366798 to track this. tagged as version 0.1.0 |
Java doesn't support unsigned types, but the ROOM language supports them. Currently the dummy "undefined_type" is generated by eTrice's Java generator. It would be better to issue a generator error if the Java generator encounters these types. It would also be a nice extension for the generator to support the option "mapUnsignedTypes", which does a mapping: uint8 => Java int uint16 => Java int uint32 => Java long uint64 => BigInteger (currently uint64 is not supported by eTrice's ROOM language). Without this option, a user would be stuck if porting an existing ROOM model to a Java environment. BTW: There is a nice page explaining the details: http://www.javamex.com/java_equivalents/unsigned.shtml