Autosar-data

Released 2024-10-28

Fixes

- Fix a regression in the attribute parsing code, introduced in 0.15

Released 2024-10-24

Api

- Allow specifying a max depth for elements_dfs - Thanks shahn
- Add additional info the the errors `ElementInsertionConflict` and `ElementNotFound`

Fixes

- when loading two or more files, the model merging code could hang in an endless loop
- XML allows attributes in single quotes, but the parser did not accept this

Released 2024-08-16

API

- The usability of `Element::set_character_data` has been improved: it is no longer required to manually wrap the value in a CharacterData object.
- `CharacterData::decode_interger` has been renamed to `CharacterData::parse_integer`
- `CharacterData::Double` renamed to `CharacterData::Float`
- added `CharacterData::parse_float`
- The errors `ItemNameRequired`, `IncorrectContentType`, `InvalidSubElement`, and `DuplicateItemName` contain additional information
- implement PartialOrd / Ord for elements

Enhancements

- sorting of elements is now more complete, and takes the CharacterData and sub-elements into account
- criterion has been added for benchmarking

Fixes

- Add a fast path that dramatically speeds up a common case in element insertion

Released 2024-05-19.

API

- add `AutosarModel::duplicate()`
This creates a full copy of the model. It is useful because `clone()` is overridden to only create a new reference to the existing model.
- add `Element::named_parent()`
Unlike `parent()` which only returns the direct parent of the element, `named_parent()` performs several steps until a named (identifiable) parent is found.
- add `CharacterData::decode_integer()`
Many numbers in an arxml file are declared as strings, which allows them to be represented in hexadecimal (0x...) or binary (0b...) form in the serialized arxml file. This function parses these strings into integer values.
- add `Element::get_or_create_sub_element()` and `Element::get_or_create_named_sub_element()`
Often it is only required that some sub element should exist, but it doesn't matter if an existing element is used, or if a new one is ceated. These functions simplify that use case.
- make `Element::min_version()` public
It returns the `AutosarVersion` of file containing the element. The `Element` might be part of multiple arxml files in the merged model, and if this is the case, then the lowest version of all contining files is returned.
- modified `AutoarModel::identifiable_elements()` to return an iterator instead of a list.
If a list is desired, then the user of the function can `collect()` the iterator.
- add `ElementType::verify_reference_dest()`
It checks if the value of the DEST attribute in a reference is valid acoording to the specification.

Fixes

- fix a parsing failure if a comment contained an embedded `>` character. Additionally, `Element::set_comment()` now checks if the new string contains `--` and substitiutes with `__` if it does: The XML spec prohibits `--` in XML comments.
- fix `AutosarModel::check_references()`, which reported some valid references as invalid.
- fix `ArxmlFile::elements_dfs()`, which would panic if it was called on a deleted file or model.

Other

- mark several functions as `[must_use]`
- Fix the formatting of many doc comments. For example the section "Possible errors" was renamed to "Errors", and many code items are now enclosed in backticks.
- Spelling fixes in verious comments - Thanks b4skyx

---

Released 2024-02-01.

Fixes

- fix `check_file()` and `check_buffer()`, which were broken if the arxml header contained a comment.

---

Released 2023-12-13.

Features

- Support Autosar version R23-11
- Support comments in the model
Comments are no longer discarded during parsing. One comment can be stored with each element.
- The data model of autosar-data-specification was improved. It now provides more information.
- autosar-data-specification has a create feature `docstrings`. If this is set, the create can provide the original doc strings from the Autosar meta model for each element type.

API

- Comments can be modified with `Element::set_comment()` and retrieved with `Element::comment()`
- added `ElementType::std_restriction()`, which informs if an element is restricted to `AdaptivePlatform` or `ClassicPlatform`
- mark `AutosarDataError` and several other enums as `[non_exhaustive]`

Fixes

- The information returned by `ElementType::is_ordered()` and `ElementType::is_splittable()` was previously inaccurate in some cases. This was fixed by improving the representation of this information in the data tables.

---