The feature is particularly useful when working with complex codebases or when trying to achieve specific outcomes in the generated code. By providing these instructions, users can ensure that the AI model produces code that meets their exact requirements and specifications.
Not yet available to public users.
Inline instruction is a special mechanism that allows users to provide specific guidance to AI models during the process of source code related tasks.
One key benefit of using inline instruction is that it allows users to fine-tune the output of the AI model without having to manually edit the code afterwards. This can save time and effort, as users can quickly and easily make adjustments to the generated code without having to start from scratch.
Users are able to give instructions on how certain parts of the code should be modified or optimized by using <ai>...</ai> tags in comments within the code.
For example, the instruction tells to CodePorting AI engine to rename crc_check function to verifyCrc:
# <ai>Rename function crc_check to verifyCrc</ai>
def crc_check(data, div):
...
return crc