Trim element symbol in atomic_data abort message

[djkees]

Summary

Fixes a stray space in the "Symbol not found" abort message in get_atom (source/atomic_data.f90) when the element symbol is a single letter.

Changes

symbol is a fixed 2-character field (element_name_len = 2), so for single-letter symbols like H it's stored as "H " with a trailing space. The abort message embedded symbol directly instead of trimming it, producing Symbol H not found in the element list. (double space) instead of Symbol H not found in the element list.. Every other abort() call in the codebase that embeds a variable name trims it first (see mixture.f90, input.f90, equilibrium.f90) — this was the one outlier. Added trim() around symbol in the one abort() call at source/atomic_data.f90:155.

Testing

• Built and ran the full test suite (core-c preset, GNU/gfortran on Windows): all 14 tests pass.
• Manually reproduced the abort path with a crafted input (reac name Test Q 1 wt%=100, an unrecognized single-letter element in an exploded formula, forcing molecular_weight_from_formula -> get_atom_weight -> get_atom). Confirmed the message changed from a double space to a single space: CRITICAL: Symbol Q not found in the element list.

Compatibility / Numerical behavior

• No expected changes to numerical results

This only changes the text of an error message on an abort path; it does not touch any numerical computation.

---

Drafted with Claude's assistance

• Confirmed the bug and the fix by reading source/atomic_data.f90 directly (the element_name_len = 2 field and the single abort() call missing trim()), and by grepping every other abort() call site to verify the trim() pattern is otherwise universal.
• Built the project (core-c preset) and ran ctest — all 14 existing tests passed before and after the change.
• Reproduced the bug end-to-end with a real cea.exe invocation before and after the fix to confirm the message actually changes as described, rather than trusting the diff alone.