Shallow duplicate in r_vector(const r_vector& rhs) constructor after all
#387
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The latter goes from `matrix -> SEXP -> r_vector` in a way that always avoids a copy, regardless of read-only or writable. This new way ensures that we go `r_vector -> r_vector`, and the correct `r_vector` constructor can get called. If its a writable input, it will be correctly copied now.
Otherwise treat it like other attributes that just get shallow copied over
…ructor after all The fact that `reserve_data()` drops `dim` and `dimnames` unconditionally is problematic here, because this copy constructor gets called from inside the `matrix` class. Really we should restrict usage of `reserve_data()` to JUST `reserve()`. This should be fine because it will only get called if we `push_back()`, and you can't `push_back()` on a `matrix`.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Reverts part of #383, specifically the note here https://github.com/r-lib/cpp11/pull/383/files#r1723749948
We really do need to call
Rf_shallow_duplicate()because we want ALL of the attributes on therhsto get copied over, evendimanddimnames.reserve_data()is meant to be used when we KNOW we have a vector thatpush_back()has been called on, which should hopefully never be amatrix(there is nomatrixAPI for this). If that holds, then it is ok that thedimanddimnamesattributes get dropped on growth (i.e. during apush_back()) and truncation (i.e. in theSEXPoperator).The downside of this comes up when you do:
push_back(), then use the copy constructor, then return that value. The copy constructor retains all the extra capacity (meaning the allocation is larger than it maybe needs to be), and then when we return the value we also have to truncate. But this combination is probably somewhat rare, so its fine.Without this, literanger breaks. We have now added tests to ensure we can't break this again accidentally.