Added encoding and parsing for TeraChem optimization calculations.

.pre-commit-config.yaml

@@ -34,6 +34,10 @@ repos:
       - id: mypy
         additional_dependencies:
           [tokenize-rt==3.2.0, pydantic>=1.0.0, types-paramiko, types-toml, qcio>=0.11.8]
+  - repo: https://github.com/crate-ci/typos
+    rev: v1.24.5
+    hooks:
+      - id: typos
   - repo: local
     hooks:
       - id: tests

pyproject.toml

@@ -57,3 +57,9 @@ init_forbid_extra = true
 init_typed = true
 warn_required_dynamic_aliases = true
 warn_untyped_fields = true
+
+[tool.typos]
+# Exclude specific files or directories
+files.extend-exclude = [
+    "tests/data/terachem_opt/tc.out", # Single file
+]

qcparse/encoders/terachem.py

@@ -29,6 +29,11 @@ def encode(inp_obj: ProgramInput) -> NativeInput:
         calctype = "frequencies"
     elif inp_obj.calctype.value == CalcType.optimization:
         calctype = "minimize"
+        if not inp_obj.keywords.get("new_minimizer", "no") == "yes":
+            raise EncoderError(
+                "Only the new_minimizer is supported for optimizations. Add "
+                "'new_minimizer': 'yes' to the keywords."
+            )
     elif inp_obj.calctype.value == CalcType.transition_state:
         calctype = "ts"
     else:

qcparse/main.py

@@ -4,9 +4,9 @@
 import warnings
 from importlib import import_module
 from pathlib import Path
-from typing import List, Union
+from typing import List, Optional, Union
 
-from qcio import ProgramInput, SinglePointResults
+from qcio import CalcType, ProgramInput, SinglePointResults
 
 from .exceptions import EncoderError, MatchNotFoundError, ParserError
 from .models import NativeInput, ParserSpec, registry, single_point_results_namespace
@@ -20,6 +20,7 @@ def parse(
     data_or_path: Union[str, bytes, Path],
     program: str,
     filetype: str = "stdout",
+    calctype: Optional[CalcType] = None,
 ) -> SinglePointResults:
     """Parse a file using the parsers registered for the given program.
 
@@ -53,7 +54,7 @@ def parse(
 
     # Get the calctype if filetype is 'stdout'
     if filetype == "stdout":
-        calctype = parsers.parse_calctype(file_content)
+        calctype = calctype if calctype else parsers.parse_calctype(file_content)
 
     # Get all the parsers for the program, filetype, and calctype
     parser_specs: List[ParserSpec] = registry.get_parsers(program, filetype, calctype)

qcparse/parsers/terachem.py

@@ -183,13 +183,14 @@ def parse_optimization_dir(
     directory: Union[Path, str],
     stdout: str,
     *,
-    input_data: ProgramInput,
+    inp_obj: ProgramInput,
 ) -> OptimizationResults:
     """Parse the output directory of a TeraChem optimization calculation.
 
     Args:
         directory: Path to the directory containing the TeraChem output files.
         stdout: The contents of the TeraChem stdout file.
+        inp_obj: The input object used for the calculation.
 
     Returns:
         OptimizationResults object
@@ -200,7 +201,12 @@ def parse_optimization_dir(
     structures = Structure.open(directory / "optim.xyz")
     assert isinstance(structures, list), "Expected multiple structures in optim.xyz"
 
-    # Parse their gradients
+    # Parse Values
+    from qcparse import parse
+
+    # Parse all the values from the stdout file
+    spr = parse(stdout, "terachem", "stdout", CalcType.energy)
+
     gradients = parse_gradients(stdout)
     program_version = parse_version_string(stdout)
 
@@ -210,14 +216,17 @@ def parse_optimization_dir(
             input_data=ProgramInput(
                 calctype=CalcType.gradient,
                 structure=structure,
-                model=input_data.model,
-                keywords=input_data.keywords,
+                model=inp_obj.model,
+                keywords=inp_obj.keywords,
             ),
             results=SinglePointResults(
-                # TeraChem places the energy as the first comment in the xyz file
-                energy=structure.extras[Structure._xyz_comment_key][0],
-                # Will be coerced by Pydantic to np.ndarray
-                gradient=gradient,  # type: ignore
+                **{
+                    **spr.model_dump(),
+                    # TeraChem places the energy as the first comment in the xyz file
+                    "energy": structure.extras[Structure._xyz_comment_key][0],
+                    # # Will be coerced by Pydantic to np.ndarray
+                    "gradient": gradient,  # type: ignore
+                }
             ),
             success=True,
             provenance=Provenance(

tests/test_terachem.py

@@ -255,7 +255,7 @@ def test_parse_optimization_dir(test_data_dir, prog_inp):
     opt_inp = prog_inp("optimization")
     stdout = (test_data_dir / "terachem_opt" / "tc.out").read_text()
     opt_results = parse_optimization_dir(
-        test_data_dir / "terachem_opt", stdout, input_data=opt_inp
+        test_data_dir / "terachem_opt", stdout, inp_obj=opt_inp
     )
     for prog_output in opt_results.trajectory:
         assert prog_output.input_data.calctype == CalcType.gradient