allow arbitrary vector allocators for restarts

[PhilipDeegan]

samrai restart putVector style functions only support vectors with the default standard allocator

we could ask samrai folks to allow for other allocators, but that might never even happen

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 0cf1dbcb-db0a-4b11-8492-97ff81293a61

📥 Commits

Reviewing files that changed from the base of the PR and between 1b388e3 and dbf1d35.

📒 Files selected for processing (6)

• src/amr/data/field/field_data.hpp
• src/amr/data/particles/particles_data.hpp
• src/amr/data/tensorfield/tensor_field_data.hpp
• src/amr/samrai.cpp
• src/amr/samrai.hpp
• src/core/utilities/types.hpp

✅ Files skipped from review due to trivial changes (3)

• src/core/utilities/types.hpp
• src/amr/data/field/field_data.hpp
• src/amr/samrai.hpp

🚧 Files skipped from review as they are similar to previous changes (3)

• src/amr/data/tensorfield/tensor_field_data.hpp
• src/amr/data/particles/particles_data.hpp
• src/amr/samrai.cpp

---

📝 Walkthrough

Walkthrough

Refactors restart I/O to use new helper templates, adds those helpers in amr/samrai.hpp, updates SAMRAI lifecycle initialization/teardown in SamraiLifeCycle, and adds a SFINAE-friendly dependent_false_v utility. No public API signatures changed. (49 words)

Changes

Cohort / File(s)	Summary
SAMRAI helpers src/amr/samrai.hpp	Adds getVectorFromRestart and putVectorToRestart template helpers (supporting double and int) and adjusts includes; centralizes vector restart I/O.
AMR data restart I/O src/amr/data/field/field_data.hpp, src/amr/data/particles/particles_data.hpp, src/amr/data/tensorfield/tensor_field_data.hpp	Replaces direct restart_db->getDoubleArray/getVector and putVector calls with the new helper functions; minor const-correctness and include tweaks.
SAMRAI lifecycle src/amr/samrai.cpp	Explicit SAMRAI/MPI initialization and shutdown moved into SamraiLifeCycle (init/startup/shutdown/finalize); adds logger warning appender and conditional PHLOP timing setup; modifies restart-manager reset behavior.
Utilities src/core/utilities/types.hpp	Adds template<typename> inline constexpr bool dependent_false_v = false; for use in SFINAE/compile-time checks.

Sequence Diagram(s)

mermaid
sequenceDiagram
participant LifeCycle as SamraiLifeCycle
participant MPI as SAMRAI_MPI
participant Manager as SAMRAIManager
participant Restart as RestartManager
participant PHLOP as phlop::ScopeTimerMan

LifeCycle->>MPI: init()
LifeCycle->>Manager: initialize()
LifeCycle->>Manager: startup()
LifeCycle->>Restart: getRestartManager() (used later)
alt during reset
    LifeCycle->>Manager: shutdown()
    LifeCycle->>Restart: clearRestartItems()
    LifeCycle->>Manager: startup()
end
LifeCycle->>Manager: shutdown() (destructor)
LifeCycle->>Manager: finalize()
LifeCycle->>MPI: finalize()
alt PHLOP enabled
    LifeCycle->>PHLOP: ScopeTimerMan::reset()
end

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• python module per simulator template permutation #1047: Overlapping SAMRAI lifecycle and restart I/O refactors; likely closely related to the new helper and lifecycle changes.

Suggested labels

experimental

Suggested reviewers

• UCaromel
• nicolasaunai

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 22.22% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main objective of the changeset: enabling support for arbitrary vector allocators in restart operations.
Description check	✅ Passed	The description relates to the changeset by explaining the motivation: SAMRAI's restart functions only support default allocators, and this PR works around that limitation.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

src/amr/samrai.hpp (1)

72-83: Consider making empty vectors a no-op in the write helper.

Adding an early vec.empty() return here would centralize behavior and avoid repeating zero-size guards at call sites.

♻️ Proposed refactor

 template<typename T, typename A>
 void putVectorToRestart(auto& db, auto const& path, std::vector<T, A> const& vec)
 {
+    if (vec.empty())
+        return;
+
     if constexpr (std::is_same_v<T, double>)
         db.putDoubleArray(path, vec.data(), vec.size());

     else if constexpr (std::is_same_v<T, int>)
         db.putIntegerArray(path, vec.data(), vec.size());

As per coding guidelines, `**/*.hpp`: “Review the C++ code, point out issues relative to principles of clean code, expressiveness, and performance.”

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/amr/samrai.hpp` around lines 72 - 83, The helper putVectorToRestart
should early-return on empty vectors to centralize behavior and avoid callers
duplicating zero-size checks; modify putVectorToRestart to check if vec.empty()
and return immediately before calling db.putDoubleArray or db.putIntegerArray,
preserving the existing type-specific branches (std::is_same_v<T,double> and
std::is_same_v<T,int>) and keep the static_assert for unsupported types.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@src/amr/samrai.hpp`:
- Around line 72-83: The helper putVectorToRestart should early-return on empty
vectors to centralize behavior and avoid callers duplicating zero-size checks;
modify putVectorToRestart to check if vec.empty() and return immediately before
calling db.putDoubleArray or db.putIntegerArray, preserving the existing
type-specific branches (std::is_same_v<T,double> and std::is_same_v<T,int>) and
keep the static_assert for unsupported types.

---

ℹ️ Review info

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between d7d962d and 1b388e3.

📒 Files selected for processing (6)

• src/amr/data/field/field_data.hpp
• src/amr/data/particles/particles_data.hpp
• src/amr/data/tensorfield/tensor_field_data.hpp
• src/amr/samrai.cpp
• src/amr/samrai.hpp
• src/core/utilities/types.hpp