From 0c3846f496aeb5aae405fda8b91174ceb8c8e4ed Mon Sep 17 00:00:00 2001
From: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>
Date: Sat, 6 Jan 2024 01:53:30 -0500
Subject: [PATCH] breaking: remove implicit Hs from RDKit SMILES

Fix #2087.

Signed-off-by: Jinzhe Zeng <jinzhe.zeng@rutgers.edu>
---
 reacnetgenerator/_path.py     | 35 +++++++++++++++++++++++++++++++++-
 reacnetgenerator/_reachtml.py | 36 ++---------------------------------
 tests/test.json               | 11 +++++++----
 tests/test_reacnetgen.py      | 10 +++++-----
 4 files changed, 48 insertions(+), 44 deletions(-)

diff --git a/reacnetgenerator/_path.py b/reacnetgenerator/_path.py
index 4fa330549..146117511 100644
--- a/reacnetgenerator/_path.py
+++ b/reacnetgenerator/_path.py
@@ -19,6 +19,7 @@
 """
 
 import itertools
+import re
 from abc import ABCMeta, abstractmethod
 from collections import Counter, defaultdict
 
@@ -215,6 +216,38 @@ def _printatomroute(self, atomeach, timeaxis=None):
             allmoleculeroute = np.unique(allmoleculeroute, axis=0)
         return allmoleculeroute
 
+    def _re(self, smi):
+        """If you use RDkit to convert a methyl radical to SMILES, you will get something
+        like [H]C([H])[H]. However, OpenBabel will consider it as a methane molecule. So,
+        you have to use [H][C]([H])[H], if you need to process some radicals.
+
+        Examples
+        --------
+        >>> self._re('C')
+        [C]
+        >>> self._re('[C]')
+        [C]
+        >>> self._re('[CH]')
+        [CH]
+        >>> self._re('Na')
+        [Na]
+        >>> self._re('[H]c(Cl)C([H])Cl')
+        [H][c]([Cl])[C]([H])[Cl]
+        """
+        if "_unknownSMILES" in smi:
+            # not SMILES
+            return smi
+        Satom = sorted(self.atomname, key=len, reverse=True)
+        elements = "|".join(
+            [
+                ((an.upper() + "|" + an.lower()) if len(an) == 1 else an)
+                for an in Satom
+                if an != "H"
+            ]
+        )
+        smi = re.sub(r"(?<!\[)(" + elements + r")(?!H)", r"[\1]", smi)
+        return smi.replace("[HH]", "[H]")
+
     def convertSMILES(self, atoms, bonds):
         """Convert atoms and bonds information to SMILES.
 
@@ -230,7 +263,7 @@ def convertSMILES(self, atoms, bonds):
         for atom1, atom2, level in bonds:
             m.AddBond(d[atom1], d[atom2], Chem.BondType(level))  # type: ignore
         name = Chem.MolToSmiles(m)  # type: ignore
-        return name
+        return self._re(name)
 
     def _getatomsandbonds(self, line):
         atoms = np.array(bytestolist(line[0]), dtype=int)
diff --git a/reacnetgenerator/_reachtml.py b/reacnetgenerator/_reachtml.py
index ae245e272..2f418ee75 100644
--- a/reacnetgenerator/_reachtml.py
+++ b/reacnetgenerator/_reachtml.py
@@ -66,43 +66,11 @@ def report(self):
             f"Report is generated. Please see {self.resultfilename} for more details."
         )
 
-    def _re(self, smi):
-        """If you use RDkit to convert a methyl radical to SMILES, you will get something
-        like [H]C([H])[H]. However, OpenBabel will consider it as a methane molecule. So,
-        you have to use [H][C]([H])[H], if you need to process some radicals.
-
-        Examples
-        --------
-        >>> self._re('C')
-        [C]
-        >>> self._re('[C]')
-        [C]
-        >>> self._re('[CH]')
-        [CH]
-        >>> self._re('Na')
-        [Na]
-        >>> self._re('[H]c(Cl)C([H])Cl')
-        [H][c]([Cl])[C]([H])[Cl]
-        """
-        if "_unknownSMILES" in smi:
-            # not SMILES
-            return smi
-        Satom = sorted(self.atomname, key=lambda i: len(i), reverse=True)
-        elements = "|".join(
-            [
-                ((an.upper() + "|" + an.lower()) if len(an) == 1 else an)
-                for an in Satom
-                if an != "H"
-            ]
-        )
-        smi = re.sub(r"(?<!\[)(" + elements + r")(?!H)", r"[\1]", smi)
-        return smi.replace("[HH]", "[H]")
-
     def _handlereaction(self, line):
         sx = line.split()
         left, right = sx[1].split("->")
-        left = [self._re(spec) for spec in left.split("+")]
-        right = [self._re(spec) for spec in right.split("+")]
+        left = left.split("+")
+        right = right.split("+")
         num = int(sx[0])
         return left, right, num
 
diff --git a/tests/test.json b/tests/test.json
index 52836a9e9..17c65fa0d 100644
--- a/tests/test.json
+++ b/tests/test.json
@@ -40,7 +40,7 @@
         ]
       ]
     },
-    "reaction_sha256": "5eb802a5a4f5c537ea556528a7db9402ab401f9cbd650e75b2a03e809ab9c8ad"
+    "reaction_sha256": "12a2c53d81db733405d24dfc67dd390c85195c4011cef53305629e8083a5d2f4"
   },
   {
     "rngparams": {
@@ -64,7 +64,8 @@
       "speciescenter": "[H][H]"
     },
     "reaction_sha256": [
-      "3ad36fbbbe2f52f4533b42e19e9efa393129d505860f00efb544aac7820caac7"
+      "7b26f8b66b9134a726e1aadc4adf200f42f5921692d30a1c7867fcf437b1fb88",
+      "1f3a88d3281b7627bc9bed1a533d7b08f1edd4d7811291d4d07cc5bbf7ac9bd4"
     ]
   },
   {
@@ -88,8 +89,10 @@
       "miso": 1
     },
     "reaction_sha256": [
-      "ec826635846cbe4d70a2454056ae973aaaf2c7bb760514017ebd6c804a85b6c2",
-      "2b1f9276faeabd4dbbd3349a9ddd5ffbcb64ebc11dcf3f16d661c34372f36dca"
+      "c7d2d7b905d247aaeba36b0c272506ab287c557c5365b7090bd9a04a45e1fc48",
+      "3d99317c870758bd9d33e29d474d9496643b74bd7b4b07fa1e31afff33dbcd07",
+      "88e88f1fd4e83f37f560b6786cee0feadf1139bfb7de8551b2946a11e4b767f9",
+      "e1f445eab53f7208ac9b176a09ac79360999957edbf2e5a9d9e8990bc20b91d7"
     ]
   },
   {
diff --git a/tests/test_reacnetgen.py b/tests/test_reacnetgen.py
index fc6f38828..8a50d6ab9 100644
--- a/tests/test_reacnetgen.py
+++ b/tests/test_reacnetgen.py
@@ -15,7 +15,7 @@
 from reacnetgenerator import ReacNetGenerator
 from reacnetgenerator._detect import _Detect
 from reacnetgenerator._hmmfilter import _HMMFilter
-from reacnetgenerator._reachtml import _HTMLResult
+from reacnetgenerator._path import _CollectSMILESPaths
 from reacnetgenerator.commandline import parm2cmd
 from reacnetgenerator.gui import GUI
 from reacnetgenerator.utils import checksha256, download_multifiles, listtobytes
@@ -103,17 +103,17 @@ def test_gui_openandrun(self, reacnetgengui, mocker, reacnetgen_param):
 
     def test_commandline_help(self, script_runner):
         """Test commandline of ReacNetGenerator."""
-        ret = script_runner.run("reacnetgenerator", "-h")
+        ret = script_runner.run(["reacnetgenerator", "-h"])
         assert ret.success
 
     def test_commandline_version(self, script_runner):
         """Test commandline of ReacNetGenerator."""
-        ret = script_runner.run("reacnetgenerator", "--version")
+        ret = script_runner.run(["reacnetgenerator", "--version"])
         assert ret.success
 
     def test_commandline_run(self, script_runner, reacnetgen_param):
         """Test commandline of ReacNetGenerator."""
-        ret = script_runner.run(*parm2cmd(reacnetgen_param["rngparams"]))
+        ret = script_runner.run(parm2cmd(reacnetgen_param["rngparams"]))
         assert ret.success
 
     def test_benchmark_detect(self, benchmark, reacnetgen_param):
@@ -146,7 +146,7 @@ def test_benchmark_hmm(self, benchmark, reacnetgen_param):
     def test_re(self, reacnetgen_param):
         """Test regular expression of _HTMLResult."""
         reacnetgen = ReacNetGenerator(**reacnetgen_param["rngparams"])
-        r = _HTMLResult(reacnetgen)
+        r = _CollectSMILESPaths(reacnetgen)
         r.atomname = ["C", "H", "O", "Na", "Cl"]
         assert r._re("C"), "[C]"
         assert r._re("[C]"), "[C]"