Fix #811

doc/source/develop/add_solver.rst

@@ -92,13 +92,13 @@ Takes an :py:class:`pulp.pulp.LpProblem` as argument, solves it, stores the solu
             os.remove(tmpSol)
         except:
             pass
-        cmd = self.path
-        cmd += " %s" % tmpMps
-        cmd += " -solfile %s" % tmpSol
+        cmd = [self.path]
+        cmd.append(tmpMps)
+        cmd.extend(["-solfile", tmpSol])
         if self.timeLimit is not None:
-            cmd += " -time %s" % self.timeLimit
+            cmd.extend(["-time", f"{self.timeLimit}"])
         for option in self.options:
-            cmd += " " + option
+            cmd.append(option)
         if lp.isMIP():
             if not self.mip:
                 warnings.warn("MIPCL_CMD cannot solve the relaxation of a problem")
@@ -107,7 +107,7 @@ Takes an :py:class:`pulp.pulp.LpProblem` as argument, solves it, stores the solu
         else:
             pipe = open(os.devnull, "w")
 
-        return_code = subprocess.call(cmd.split(), stdout=pipe, stderr=pipe)
+        return_code = subprocess.call(cmd, stdout=pipe, stderr=pipe)
         # We need to undo the objective swap before finishing
         if lp.sense == constants.LpMaximize:
             lp += -lp.objective

pulp/apis/choco_api.py

@@ -89,13 +89,15 @@ def actualSolve(self, lp):
             os.remove(tmpSol)
         except:
             pass
-        cmd = java_path + ' -cp "' + self.path + '" org.chocosolver.parser.mps.ChocoMPS'
+        cmd = [java_path, "-cp", self.path, "org.chocosolver.parser.mps.ChocoMPS"]
         if self.timeLimit is not None:
-            cmd += f" -limit [-{self.timeLimit}s]"
-        cmd += " " + " ".join([f"{key} {value}" for key, value in self.options])
-        cmd += f" {tmpMps}"
+            cmd.extend(["-limit", f"[-{self.timeLimit}s]"])
+        cmd.extend(
+            [key_or_value for key_value in self.options for key_or_value in key_value]
+        )
+        cmd.append(tmpMps)
         if lp.sense == constants.LpMaximize:
-            cmd += " -max"
+            cmd.append("-max")
         if lp.isMIP():
             if not self.mip:
                 warnings.warn("CHOCO_CMD cannot solve the relaxation of a problem")

pulp/apis/gurobi_api.py

@@ -459,23 +459,23 @@ def actualSolve(self, lp):
             os.remove(tmpSol)
         except:
             pass
-        cmd = self.path
+        cmd = [self.path]
         options = self.options + self.getOptions()
         if self.timeLimit is not None:
             options.append(("TimeLimit", self.timeLimit))
-        cmd += " " + " ".join([f"{key}={value}" for key, value in options])
-        cmd += f" ResultFile={tmpSol}"
+        cmd.extend(f"{key}={value}" for key, value in options)
+        cmd.append(f"ResultFile={tmpSol}")
         if self.optionsDict.get("warmStart", False):
             self.writesol(filename=tmpMst, vs=vs)
-            cmd += f" InputFile={tmpMst}"
+            cmd.append(f"InputFile={tmpMst}")
 
         if lp.isMIP():
             if not self.mip:
                 warnings.warn("GUROBI_CMD does not allow a problem to be relaxed")
-        cmd += f" {tmpLp}"
+        cmd.append(tmpLp)
         pipe = self.get_pipe()
 
-        return_code = subprocess.call(cmd.split(), stdout=pipe, stderr=pipe)
+        return_code = subprocess.call(cmd, stdout=pipe, stderr=pipe)
 
         # Close the pipe now if we used it.
         if pipe is not None:

pulp/apis/mipcl_api.py

@@ -92,13 +92,13 @@ def actualSolve(self, lp):
             os.remove(tmpSol)
         except:
             pass
-        cmd = self.path
-        cmd += f" {tmpMps}"
-        cmd += f" -solfile {tmpSol}"
+        cmd = [self.path]
+        cmd.append(tmpMps)
+        cmd.extend(["-solfile", tmpSol])
         if self.timeLimit is not None:
-            cmd += f" -time {self.timeLimit}"
+            cmd.extend(["-time", f"{self.timeLimit}"])
         for option in self.options:
-            cmd += " " + option
+            cmd.append(option)
         if lp.isMIP():
             if not self.mip:
                 warnings.warn("MIPCL_CMD cannot solve the relaxation of a problem")
@@ -107,7 +107,7 @@ def actualSolve(self, lp):
         else:
             pipe = open(os.devnull, "w")
 
-        return_code = subprocess.call(cmd.split(), stdout=pipe, stderr=pipe)
+        return_code = subprocess.call(cmd, stdout=pipe, stderr=pipe)
         # We need to undo the objective swap before finishing
         if lp.sense == constants.LpMaximize:
             lp += -lp.objective

pulp/tests/test_pulp.py

@@ -181,6 +181,27 @@ def test_continuous(self):
                 prob, self.solver, [const.LpStatusOptimal], {x: 4, y: -1, z: 6, w: 0}
             )
 
+        def test_continuous_with_spaces_in_path(self):
+            with tempfile.TemporaryDirectory() as tmpdir:
+                prob = LpProblem(self._testMethodName, const.LpMinimize)
+                prob.tmpDir = tmpdir + "/folder with spaces"
+                x = LpVariable("x", 0, 4)
+                y = LpVariable("y", -1, 1)
+                z = LpVariable("z", 0)
+                w = LpVariable("w", 0)
+                prob += x + 4 * y + 9 * z, "obj"
+                prob += x + y <= 5, "c1"
+                prob += x + z >= 10, "c2"
+                prob += -y + z == 7, "c3"
+                prob += w >= 0, "c4"
+
+                pulpTestCheck(
+                    prob,
+                    self.solver,
+                    [const.LpStatusOptimal],
+                    {x: 4, y: -1, z: 6, w: 0},
+                )
+
         def test_non_intermediate_var(self):
             prob = LpProblem(self._testMethodName, const.LpMinimize)
             x_vars = {