[thirtytwobits]

Adds the count-value constructors to cetl::VariableLengthArray, matching std::vector::vector (3):

VariableLengthArray(count, value, alloc);
VariableLengthArray(count, value, max_size_max, alloc);

Both the general and the packed-bool versions:

• construct count copies of value, using exactly the storage needed (whole bytes for bool)
• throw std::length_error when count is greater than max_size(), and std::bad_alloc when allocation fails
• with exceptions disabled, truncate to max_size() or to the storage the allocator provides, which you can detect by comparing size() with count

Conflicts with the existing constructors (the issue asked for this check)

• Initializer list + max_size_max. With a plain size_type count parameter, VLA<int>({x}, 5, alloc) becomes ambiguous. The braced list prefers initializer_list, but the int literal 5 converts better to const int& than to size_type. I confirmed the ambiguity with both GCC and Clang. To avoid it, the count's type is deduced, accepting any type convertible to size_type. A braced argument can't be deduced, so that call still picks the initializer-list constructor.
• (n, alloc). This still sets max_size_max. There is no std::vector-style (count, alloc) overload, because it would collide with that constructor. The docs say so.
• Range constructors. These are unchanged. Iterators, including pointers for VLA<std::size_t> and VLA<int*>, still select them.
• Integer arguments. VLA<int>(3, 4, alloc) used to be ill-formed. It now constructs three 4s, as std::vector does. The assign suite's "non-iterators must not select the range constructor" assertions relied on the old behaviour, so they now use nullptr_t arguments, which are neither iterators nor counts.

Tests

• New test_variable_length_array_construct.cpp, typed over int and bool and checked against std::vector. It covers byte boundaries, max_size_max, counts larger than max_size(), different count types (including an unscoped enum), and overload resolution against the max-size, initializer-list (parenthesised and braced) and range constructors, with size_type, pointer and std::string elements.
• test_variable_length_array_exception_safety.cpp adds count-value cases for throwing element copies and allocator construction-hook failures, plus a new test for length_error and bad_alloc that checks nothing is constructed or allocated.
• test_variable_length_array_compat.cpp adds a std::string count-value case that runs against both VLA and std::vector.
• The fuzzer's CopyConstruct operation now also checks count-value construction against the model. The input format is unchanged.

Validation

• GCC 13, C++14: all 12 VLA suites pass in Debug and DebugEP (-fno-exceptions)
• GCC 7.5 (toolshed ts24.4.3), C++14: the construct, exception-safety, compat and assign suites pass with and without exceptions
• GCC 13 and Clang, C++14/17/20, with and without exceptions: those four suites compile cleanly with -Wall -Wextra -Werror -Wconversion -Wsign-conversion -pedantic
• Clang, C++20, ASan + UBSan: those four suites pass
• libFuzzer, default and embedded profiles: ctest passes, plus 60 s runs of about 650k inputs each with no findings
• clang-format 19.1.7 from toolshed:ts24.4.3 reports no changes for any line I touched; git diff --check is clean. Some untouched lines in the compat and exception-safety tests already differ from the format on main, and I left them alone.

Closes #63.

🤖 Generated with Claude Code

[thirtytwobits]

@codex review

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-19T00:12:49.897530Z	db639ba	Manual request

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: db639ba303

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".