Add matid capability to the sesame2spiner library by buechlerm · Pull Request #659 · lanl/singularity-eos

buechlerm · 2026-09-23T13:42:03Z

Add API for a vector of matid's in the sesame2spiner library.

PR Summary

Adds a saveAllMaterials interface to the sesame2spiner library so host codes can generate an sp5 file directly from a list of sesame matids — with or without per-material Params overrides — instead of having to write out input files first, plus an hid_t overload that lets callers build a file up one material at a time. All overloads, including the existing file-based one, now funnel into a single core implementation, so duplicate matid and name detection queries the HDF5 file itself (H5Lexists) rather than in-memory sets and stays correct across incremental calls. Also fixes three latent bugs surfaced by the refactor: a matid absent from the sesame file previously yielded all-zero metadata bounds that reached log() and produced NaN grids with no diagnostic (now reported and skipped, other materials still save), eosGetMetadata no longer ignores the caller's verbosity in favor of a hardcoded Verbosity::Debug, and appending to an existing file now validates its log_type root attribute against the current build.

Adds a test for any bugs fixed. Adds tests for new features.
Format your changes by using the make format command after configuring with cmake.
Document any new features, update documentation for changes made.
Make sure the copyright notice on any files you modified is up to date.
After creating a pull request, note it in the CHANGELOG.md file.
LANL employees: make sure tests pass both on the github CI and on the Darwin CI
If ML was used, make sure to add a disclaimer at the top of a file indicating ML was used to assist in generating the file.
If Agentic AI was used, have the AI generate a "proposed changes" markdown file and store it in the plan_histories folder, with a filename the same as the MR number.

If preparing for a new release, in addition please check the following:

Update the version in cmake.
Move the changes in the CHANGELOG.md file under a new header for the new release, and reset the categories.
Maintainers: ensure spackages are up to date:
- LANL-internal team, update XCAP spackages
- Current maintainer of upstream spackages, submit MR to spack

adamdempsey90 · 2026-09-25T20:31:41Z

+    if (H5Lexists(file, name.c_str(), H5P_DEFAULT) > 0) {
+      std::string new_name;
+      int suffix = 2;
+      do {
+        new_name = name + "_" + std::to_string(suffix++);
+      } while (H5Lexists(file, new_name.c_str(), H5P_DEFAULT) > 0);


i haven't wrapped my head around what this is for yet...

if multiple materials are named the same thing but have different matids, this lets you add duplicates of them. It has to walk through the file and increment the "index" of the material, so you might have steel, steel_2, steel_3, etc... I'm not sure how important this is, but now that the file is queried, keeping the old functionality requires this do while loop.

Although actually, would a cleaner implementation be while(H5Lexists) and drop the if () { do {} while()} construction?

Yurlungur · 2026-10-01T20:38:53Z

I think this new section in the README should also be added to the singularity-eos sphinx-rst. These readmes should maybe be deprecated as I want people to not have to hunt through multiple places for documentation.

Yurlungur · 2026-10-01T20:40:12Z


 ### Added (new features/APIs/variables/...)
 - [[PR658]](https://github.com/lanl/singularity-eos/pull/XXX) Add `MinimumInternalEnergy`/`MaximumInternalEnergy` to the EOS introspection API, so energy bounds are reachable through modifiers and the `singularity::EOS` variant
+- Added `sesame2spiner::saveAllMaterials` overloads taking a list of matids instead of a list of input files, with optional per-material `Params` overrides, so host codes can generate an sp5 file without writing input decks to disk. Also added an overload taking an already-open `hid_t`, plus `writeSP5RootAttributes`, so a single sp5 file can be built up one material at a time.


This change should be pinned to an MR link.

Yurlungur · 2026-10-01T20:40:22Z

+- `sesame2spiner` now honors the requested verbosity when reading material metadata, rather than always using `Verbosity::Debug`. Default command line output is correspondingly quieter.
+- `sesame2spiner::getMatBounds` no longer takes a leading index argument, which was unused.


These changes should be pinned to an MR link.

Yurlungur · 2026-10-01T20:44:47Z

+    if (H5Lexists(file, name.c_str(), H5P_DEFAULT) > 0) {
+      std::string new_name;
+      int suffix = 2;
+      do {
+        new_name = name + "_" + std::to_string(suffix++);
+      } while (H5Lexists(file, new_name.c_str(), H5P_DEFAULT) > 0);


if multiple materials are named the same thing but have different matids, this lets you add duplicates of them. It has to walk through the file and increment the "index" of the material, so you might have steel, steel_2, steel_3, etc... I'm not sure how important this is, but now that the file is queried, keeping the old functionality requires this do while loop.

Yurlungur · 2026-10-01T20:45:33Z

+    if (H5Lexists(file, name.c_str(), H5P_DEFAULT) > 0) {
+      std::string new_name;
+      int suffix = 2;
+      do {
+        new_name = name + "_" + std::to_string(suffix++);
+      } while (H5Lexists(file, new_name.c_str(), H5P_DEFAULT) > 0);


Although actually, would a cleaner implementation be while(H5Lexists) and drop the if () { do {} while()} construction?

Yurlungur · 2026-10-01T20:47:33Z

+    // Track status per material so that one failure does not get reported against
+    // every material that follows it.
+    const herr_t mat_status = saveMaterial(file, metadata, lRhoBounds, lTBounds, leBounds,
+                                           name, add_subtables, eospacWarn);
+    if (mat_status != H5_SUCCESS) {
+      std::cerr << "ERROR [" << matid << "]: problem with HDF5 while saving material."
+                << std::endl;
+      num_failed += 1;
    }
  }

+  if (num_failed > 0) {
+    std::cerr << "WARNING: " << num_failed << " of " << matids.size()
+              << " materials could not be saved." << std::endl;
+  }
+  // Count failures rather than summing herr_t values, which can cancel out and
+  // report success.
+  return (num_failed == 0) ? H5_SUCCESS : -1;


perhaps we should track which matids failed and summarize them at the end. I know it's in the error messages above, but I think that might still be nice to have a summary at the bottom.

buechlerm added 3 commits September 23, 2026 08:35

Add matid capability to the sesame2spiner library

224be90

Update plan and changelog with MR number

67b7282

Remove redudant plan hisotry

7ee7597

buechlerm force-pushed the buechler/writeMaterials branch from 583d870 to 7ee7597 Compare September 23, 2026 14:38

buechlerm requested review from Yurlungur and adamdempsey90 September 23, 2026 22:35

adamdempsey90 approved these changes Sep 25, 2026

View reviewed changes

Yurlungur approved these changes Oct 1, 2026

View reviewed changes

Navigation Menu

Add matid capability to the sesame2spiner library - #659

Add matid capability to the sesame2spiner library#659
buechlerm wants to merge 3 commits into
mainfrom
buechler/writeMaterials

buechlerm commented Sep 23, 2026 •

edited

Loading

adamdempsey90 Sep 25, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

Yurlungur Oct 1, 2026

This branch has not been deployed

Labels

3 participants

		- `sesame2spiner` now honors the requested verbosity when reading material metadata, rather than always using `Verbosity::Debug`. Default command line output is correspondingly quieter.
		- `sesame2spiner::getMatBounds` no longer takes a leading index argument, which was unused.

Conversation

buechlerm commented Sep 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Summary

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This branch has not been deployed

Labels

3 participants

buechlerm commented Sep 23, 2026 •

edited

Loading