Add from gaussian gjf method (#452)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

Author: njzjz

dpdata/gaussian/gjf.py

@@ -4,6 +4,7 @@
 """Generate Gaussian input file."""
 
 import itertools
+import re
 import uuid
 import warnings
 from typing import List, Optional, Tuple, Union
@@ -272,3 +273,68 @@ def make_gaussian_input(
         )
     buff.append("\n")
     return "\n".join(buff)
+
+
+def read_gaussian_input(inp: str):
+    """Read Gaussian input.
+
+    Parameters
+    ----------
+    inp : str
+        Gaussian input str
+
+    Returns
+    -------
+    dict
+        system data
+    """
+    flag = 0
+    coords = []
+    elements = []
+    cells = []
+    for line in inp.split("\n"):
+        if not line.strip():
+            # empty line
+            flag += 1
+        elif flag == 0:
+            # keywords
+            if line.startswith("#"):
+                # setting
+                keywords = line.split()
+            elif line.startswith("%"):
+                pass
+        elif flag == 1:
+            # title
+            pass
+        elif flag == 2:
+            # multi and coords
+            s = line.split()
+            if len(s) == 2:
+                pass
+            elif len(s) == 4:
+                if s[0] == "TV":
+                    cells.append(list(map(float, s[1:4])))
+                else:
+                    # element
+                    elements.append(re.sub("\\(.*?\\)|\\{.*?}|\\[.*?]", "", s[0]))
+                    coords.append(list(map(float, s[1:4])))
+        elif flag == 3:
+            # end
+            break
+    atom_names, atom_types, atom_numbs = np.unique(
+        elements, return_inverse=True, return_counts=True
+    )
+    if len(cells):
+        nopbc = False
+    else:
+        nopbc = True
+        cells = np.array([np.eye(3)]) * 100
+    return {
+        "atom_names": list(atom_names),
+        "atom_numbs": list(atom_numbs),
+        "atom_types": atom_types,
+        "cells": np.array(cells).reshape(1, 3, 3),
+        "nopbc": nopbc,
+        "coords": np.array(coords).reshape(1, -1, 3),
+        "orig": np.zeros(3),
+    }

dpdata/plugins/gaussian.py

@@ -27,6 +27,20 @@ def from_labeled_system(self, file_name, **kwargs):
 class GaussiaGJFFormat(Format):
     """Gaussian input file."""
 
+    def from_system(self, file_name: str, **kwargs):
+        """Read Gaussian input file.
+
+        Parameters
+        ----------
+        file_name : str
+            file name
+        **kwargs : dict
+            keyword arguments
+        """
+        with open(file_name) as fp:
+            text = fp.read()
+        return dpdata.gaussian.gjf.read_gaussian_input(text)
+
     def to_system(self, data: dict, file_name: str, **kwargs):
         """Generate Gaussian input file.
 

tests/test_gaussian_gjf.py

@@ -1,6 +1,7 @@
 import os
 import unittest
 
+from comp_sys import CompSys
 from context import dpdata
 
 
@@ -11,3 +12,16 @@ def setUp(self):
     def test_dump_gaussian_gjf(self):
         self.system.to_gaussian_gjf("tmp.gjf", keywords="force b3lyp/6-31g*")
         os.remove("tmp.gjf")
+
+
+class TestGaussianGJFComp(unittest.TestCase, CompSys):
+    def setUp(self):
+        self.system_1 = dpdata.LabeledSystem(
+            "poscars/OUTCAR.h2o.md", fmt="vasp/outcar"
+        )[0]
+        self.system_1.to_gaussian_gjf("tmp.gjf", keywords="force b3lyp/6-31g*")
+        self.system_2 = dpdata.System(
+            "tmp.gjf", fmt="gaussian/gjf", type_map=self.system_1.get_atom_names()
+        )
+        os.remove("tmp.gjf")
+        self.places = 6