- when parsing cloze question types, hector now emits a warning when he cannot find a solution for the given cloze-field
- rendered output cells for ipynb submissions are now truncated to 50 lines at most (for printing to stdout). The full output can be found in the original notebooks, still.
- hector will no longer attempt to parse empty submissions as ipynb notebook
This is the first release where the links to the built binaries are automatically added to the release notes. Simply click to
This release represents a complete rewrite of rusty-hektor focused on maintainability and stability.
- Additionally to doing an empty test, a compile test for c code can be executed
- The output format of the mapping file has been changed to
.json, which is much easier to parse in the frontend
- The Hektor output format has been slightly changed. At the top level exists a
datacontaining all the exam data (formerly this was the root object) and
metacontaining the version of the file
Note that due to the rewrite and the lack of a specification for the ILIAS output format, this new version might fail to parse ILIAS exports that have an unexpected format.
.ipynbnotebook json can be interactively edited in case the json is malformed
- Submission Type description are preprocessed by default to enable latex rendering
display_codeis now named
- the original
codefield contains the displayed code, the
source_codefield contains original source code if it differs from the existing source code
- the original
Change 'text' to 'plaintext' in order to in line with the other options which are the same strings as used by highlight.js, the highlighting library used in the Grady frontend.
Changed field 'submission_type' of submissions to 'type'
Module information can now be interactively annotated. Also solution, max score and programming language can be added to submission types.
No more dependant on OpenSSL.
First Release of Rusty Hektor!