Merge pull request #348 from tovrstra/drop-lit-error

Replace lit.error by raise LoadError

CONTRIBUTING.rst

@@ -153,9 +153,22 @@ If necessary, you can *push back* the line for later reading with ``lit.back(lin
             lit.back(line)
             break
 
-When you encounter a file format error while reading the file, call ``lit.error(msg)``,
-where ``msg`` is a short message describing the problem.
+When you encounter a file format error while reading the file, raise a ``LoadError`` exception:
+
+.. code-block:: python
+
+    from ..utils import LoadError
+
+    @document_load_one(...)
+    def load_one(lit: LineIterator) -> dict:
+        ...
+        if something_wrong:
+            raise LoadError("Describe the problem in a sentence.", lit)
+
 The error that appears in the terminal will automatically include the file name and line number.
+If your code has already read the full file and encounters an error when processing the data,
+you can use ``raise LoadError("Describe problem in a sentence.", lit.filename)`` instead.
+This way, no line number is included in the error message.
 
 
 ``dump_one`` functions: writing a single IOData object to a file

iodata/api.py

@@ -154,10 +154,10 @@ def load_one(filename: str, fmt: Optional[str] = None, **kwargs) -> IOData:
             iodata = IOData(**format_module.load_one(lit, **kwargs))
         except LoadError:
             raise
-        except StopIteration:
-            lit.error("File ended before all data was read.")
+        except StopIteration as exc:
+            raise LoadError("File ended before all data was read.", lit) from exc
         except Exception as exc:
-            raise LoadError(f"{filename}: Uncaught exception while loading file.") from exc
+            raise LoadError("Uncaught exception while loading file.", lit) from exc
     return iodata
 
 
@@ -194,7 +194,7 @@ def load_many(filename: str, fmt: Optional[str] = None, **kwargs) -> Iterator[IO
         except LoadError:
             raise
         except Exception as exc:
-            raise LoadError(f"{filename}: Uncaught exception while loading file.") from exc
+            raise LoadError("Uncaught exception while loading file.", lit) from exc
 
 
 def _check_required(iodata: IOData, dump_func: Callable):

iodata/formats/charmm.py

@@ -32,7 +32,7 @@
 import numpy as np
 
 from ..docstrings import document_load_one
-from ..utils import LineIterator, amu, angstrom
+from ..utils import LineIterator, LoadError, amu, angstrom
 
 __all__ = []
 
@@ -48,8 +48,10 @@ def load_one(lit: LineIterator) -> dict:
     while True:
         try:
             line = next(lit)
-        except StopIteration:
-            lit.error("Title section of CRD has no ending marker (missing bare *).")
+        except StopIteration as exc:
+            raise LoadError(
+                "Title section of CRD has no ending marker (missing bare *).", lit
+            ) from exc
         # Get title from crd file.
         if line.startswith("*"):
             text = line[1:]
@@ -84,7 +86,7 @@ def _helper_read_crd(lit: LineIterator) -> tuple:
     # Read the line for number of atoms.
     natom = next(lit)
     if natom is None or not natom.strip().isdigit():
-        lit.error("The number of atoms must be an integer.")
+        raise LoadError("The number of atoms must be an integer.", lit)
     natom = int(natom)
     # Read the atom lines
     resnums = []

iodata/formats/cp2klog.py

@@ -27,7 +27,7 @@
 from ..docstrings import document_load_one
 from ..orbitals import MolecularOrbitals
 from ..overlap import factorial2
-from ..utils import LineIterator
+from ..utils import LineIterator, LoadError
 
 __all__ = []
 
@@ -185,8 +185,7 @@ def _read_cp2k_obasis(lit: LineIterator) -> dict:
         " ********************* Uncontracted Gaussian Type Orbitals *********************\n"
     ):
         return _read_cp2k_uncontracted_obasis(lit)
-    lit.error("Could not find basis set in CP2K ATOM output.")
-    return None
+    raise LoadError("Could not find basis set in CP2K ATOM output.", lit)
 
 
 def _read_cp2k_occupations_energies(
@@ -436,13 +435,13 @@ def load_one(lit: LineIterator) -> dict:
         " Atomic orbital expansion coefficients [Alpha]\n",
         " Atomic orbital expansion coefficients []\n",
     ]:
-        lit.error("Could not find orbital coefficients in CP2K ATOM output.")
+        raise LoadError("Could not find orbital coefficients in CP2K ATOM output.", lit)
     coeffs_alpha = _read_cp2k_orbital_coeffs(lit, oe_alpha)
 
     if not restricted:
         line = next(lit)
         if line != " Atomic orbital expansion coefficients [Beta]\n":
-            lit.error("Could not find beta orbital coefficient in CP2K ATOM output.")
+            raise LoadError("Could not find beta orbital coefficient in CP2K ATOM output.", lit)
         coeffs_beta = _read_cp2k_orbital_coeffs(lit, oe_beta)
 
     # Turn orbital data into a MolecularOrbitals object.

iodata/formats/fchk.py

@@ -29,7 +29,7 @@
 from ..docstrings import document_dump_one, document_load_many, document_load_one
 from ..iodata import IOData
 from ..orbitals import MolecularOrbitals
-from ..utils import DumpError, LineIterator, PrepareDumpError, amu
+from ..utils import DumpError, LineIterator, LoadError, PrepareDumpError, amu
 
 __all__ = []
 
@@ -219,9 +219,9 @@ def load_one(lit: LineIterator) -> dict:
     nalpha = fchk["Number of alpha electrons"]
     nbeta = fchk["Number of beta electrons"]
     if nalpha < 0 or nbeta < 0 or nalpha + nbeta <= 0:
-        lit.error("The number of electrons is not positive.")
+        raise LoadError("The number of electrons is not positive.", lit)
     if nalpha < nbeta:
-        lit.error(f"n_alpha={nalpha} < n_beta={nbeta} is not valid!")
+        raise LoadError(f"n_alpha={nalpha} < n_beta={nbeta} is invalid.", lit)
 
     norba = fchk["Alpha Orbital Energies"].shape[0]
     mo_coeffs = np.copy(fchk["Alpha MO coefficients"].reshape(norba, nbasis).T)
@@ -323,7 +323,7 @@ def load_many(lit: LineIterator) -> Iterator[dict]:
         prefix = "Opt point"
         nsteps = fchk["Optimization Number of geometries"]
     else:
-        lit.error("Could not find IRC or Optimization trajectory in FCHK file.")
+        raise LoadError("Cannot find IRC or Optimization trajectory in FCHK file.", lit)
 
     natom = fchk["Atomic numbers"].size
     for ipoint, nstep in enumerate(nsteps):
@@ -382,7 +382,7 @@ def _load_fchk_low(lit: LineIterator, label_patterns: Optional[list[str]] = None
     elif len(words) == 2:
         result["command"], result["lot"] = words
     else:
-        lit.error("The second line of the FCHK file should contain two or three words.")
+        raise LoadError("The second line of the FCHK file should contain two or three words.", lit)
 
     while True:
         try:
@@ -434,11 +434,11 @@ def _load_fchk_field(lit: LineIterator, label_patterns: list[str]) -> tuple[str,
         if len(words) == 2:
             try:
                 return label, datatype(words[1])
-            except ValueError:
-                lit.error(f"Could not interpret: {words[1]}")
+            except ValueError as exc:
+                raise LoadError(f"Could not interpret as {datatype}: {words[1]}", lit) from exc
         elif len(words) == 3:
             if words[1] != "N=":
-                lit.error("Expected N= not found.")
+                raise LoadError("Expected N= not found.", lit)
             length = int(words[2])
             value = np.zeros(length, datatype)
             counter = 0
@@ -449,8 +449,8 @@ def _load_fchk_field(lit: LineIterator, label_patterns: list[str]) -> tuple[str,
                 word = words.pop(0)
                 try:
                     value[counter] = datatype(word)
-                except (ValueError, OverflowError):
-                    lit.error(f"Could not interpret: {word}")
+                except (ValueError, OverflowError) as exc:
+                    raise LoadError(f"Could not interpret as {datatype}: {word}", lit) from exc
                 counter += 1
             return label, value
 

iodata/formats/fcidump.py

@@ -34,7 +34,7 @@
 
 from ..docstrings import document_dump_one, document_load_one
 from ..iodata import IOData
-from ..utils import LineIterator, set_four_index_element
+from ..utils import LineIterator, LoadError, set_four_index_element
 
 __all__ = []
 
@@ -50,7 +50,7 @@ def load_one(lit: LineIterator) -> dict:
     # check header
     line = next(lit)
     if not line.startswith(" &FCI NORB="):
-        lit.error("Incorrect file header")
+        raise LoadError(f"Incorrect file header: {line.strip()}", lit)
 
     # read info from header
     words = line[5:].split(",")
@@ -77,7 +77,9 @@ def load_one(lit: LineIterator) -> dict:
     for line in lit:
         words = line.split()
         if len(words) != 5:
-            lit.error("Expecting 5 fields on each data line in FCIDUMP")
+            raise LoadError(
+                f"Expecting 5 fields on each data line in FCIDUMP, got {len(words)}.", lit
+            )
         value = float(words[0])
         if words[3] != "0":
             ii = int(words[1]) - 1

iodata/formats/gamess.py

@@ -22,7 +22,7 @@
 from numpy.typing import NDArray
 
 from ..docstrings import document_load_one
-from ..utils import LineIterator, angstrom
+from ..utils import LineIterator, LoadError, angstrom
 
 __all__ = []
 
@@ -37,7 +37,7 @@ def _read_data(lit: LineIterator) -> tuple[str, str, list[str]]:
     # The dat file only contains symmetry-unique atoms, so we would be incapable of
     # supporting non-C1 symmetry without significant additional coding.
     if symmetry != "C1":
-        lit.error(f"Only C1 symmetry is supported. Got {symmetry}")
+        raise LoadError(f"Only C1 symmetry is supported, got {symmetry}.", lit)
     symbols = []
     line = True
     while line != " $END      \n":
@@ -86,7 +86,9 @@ def _read_hessian(lit: LineIterator, result: dict[str]) -> NDArray[float]:
     """Extract ``hessian`` from the punch file."""
     # check that $HESS is not already parsed
     if "athessian" in result:
-        lit.error("Cannot parse $HESS twice! Make sure approximate hessian is not being parsed!")
+        raise LoadError(
+            "Cannot parse $HESS twice. Make sure approximate hessian is not being parsed."
+        )
     next(lit)
     natom = len(result["symbols"])
     hessian = np.zeros((3 * natom, 3 * natom), float)

iodata/formats/gaussianinput.py

@@ -22,7 +22,7 @@
 
 from ..docstrings import document_load_one
 from ..periodic import sym2num
-from ..utils import LineIterator, angstrom
+from ..utils import LineIterator, LoadError, angstrom
 
 __all__ = []
 
@@ -68,7 +68,7 @@ def load_one(lit: LineIterator):
         if not contents:
             break
         if len(contents) != 4:
-            lit.error("No Cartesian Structure is detected")
+            raise LoadError("No Cartesian Structure is detected.", lit)
         numbers.append(sym2num[contents[0]])
         coor = list(map(float, contents[1:]))
         coordinates.append(coor)

iodata/formats/gromacs.py

@@ -41,27 +41,24 @@
 @document_load_one("GRO", ["atcoords", "atffparams", "cellvecs", "extra", "title"])
 def load_one(lit: LineIterator) -> dict:
     """Do not edit this docstring. It will be overwritten."""
-    while True:
-        data = _helper_read_frame(lit)
-        title = data[0]
-        time = data[1]
-        resnums = np.array(data[2])
-        resnames = np.array(data[3])
-        attypes = np.array(data[4])
-        atcoords = data[5]
-        velocities = data[6]
-        cellvecs = data[7]
-        atffparams = {"attypes": attypes, "resnames": resnames, "resnums": resnums}
-        extra = {"time": time, "velocities": velocities}
-        return {
-            "atcoords": atcoords,
-            "atffparams": atffparams,
-            "cellvecs": cellvecs,
-            "extra": extra,
-            "title": title,
-        }
-    lit.error("Gromacs gro file could not be read.")
-    return None
+    data = _helper_read_frame(lit)
+    title = data[0]
+    time = data[1]
+    resnums = np.array(data[2])
+    resnames = np.array(data[3])
+    attypes = np.array(data[4])
+    atcoords = data[5]
+    velocities = data[6]
+    cellvecs = data[7]
+    atffparams = {"attypes": attypes, "resnames": resnames, "resnums": resnums}
+    extra = {"time": time, "velocities": velocities}
+    return {
+        "atcoords": atcoords,
+        "atffparams": atffparams,
+        "cellvecs": cellvecs,
+        "extra": extra,
+        "title": title,
+    }
 
 
 @document_load_many("GRO", ["atcoords", "atffparams", "cellvecs", "extra", "title"])