[ctml2yaml] Allow reactions to be filtered

Reaction entries can now be filtered as specified in the include node of
the reactionArray node. Fix typo in transport key name at phase level.

interfaces/cython/cantera/ctml2yaml.py

@@ -350,7 +350,12 @@ def get_species_array(speciesArray_node):
 def get_reaction_array(reactionArray_node):
     """Process reactions from a reactionArray node in a phase definition."""
     datasrc = reactionArray_node.get("datasrc", "")
+    has_filter = reactionArray_node.find("include") is not None
     if not datasrc.startswith("#"):
+        if has_filter:
+            raise ValueError(
+                "Filtering reaction lists is not possible with external data sources"
+            )
         filename, location = datasrc.split("#", 1)
         name = str(Path(filename).with_suffix(".yaml"))
         if location == "reaction_data":
@@ -361,9 +366,18 @@ def get_reaction_array(reactionArray_node):
             species_skip = skip.get("species", "")
             if species_skip == "undeclared":
                 reactions = {datasrc: "declared-species"}
+            else:
+                raise ValueError("Unknown value in skip parameter for reactionArray")
+        else:
+            raise ValueError(
+                "Missing skip node in reactionArray with external data source"
+            )
         return {"reactions": [reactions]}
     elif datasrc == "#reaction_data":
-        return {"reactions": "all"}
+        if has_filter:
+            return {"reactions": []}
+        else:
+            return {"reactions": "all"}
     else:
         return {}
 
@@ -396,8 +410,10 @@ def convert(inpfile, outfile):
 
     # Phases
     phases = []
+    reaction_filters = []
     for phase in ctml_tree.iterfind("phase"):
-        phase_attribs = {"name": phase.get("id")}
+        phase_name = phase.get("id")
+        phase_attribs = {"name": phase_name}
         phase_thermo = phase.find("thermo")
         phase_attribs["thermo"] = thermo_model_mapping[
             phase_thermo.get("model").lower()
@@ -416,20 +432,30 @@ def convert(inpfile, outfile):
             if element_skip == "undeclared":
                 phase_attribs["skip-undeclared-elements"] = True
 
-        kinetics_model = kinetics_model_mapping[
-            phase.find("kinetics").get("model").lower()
-        ]
-        if kinetics_model is not None:
-            phase_attribs["kinetics"] = kinetics_model
-
         transport_model = transport_model_mapping[
             phase.find("transport").get("model").lower()
         ]
         if transport_model is not None:
-            phase_attribs["transport-model"] = transport_model
+            phase_attribs["transport"] = transport_model
 
         if phase.find("reactionArray") is not None:
+            # The kinetics model should only be specified if reactions
+            # are associated with the phase
+            kinetics_model = kinetics_model_mapping[
+                phase.find("kinetics").get("model").lower()
+            ]
+            if kinetics_model is not None:
+                phase_attribs["kinetics"] = kinetics_model
+
             phase_attribs.update(get_reaction_array(phase.find("reactionArray")))
+            reaction_filter = phase.find("reactionArray").find("include")
+            if reaction_filter is not None:
+                phase_attribs["reactions"].append("{}-reactions".format(phase_name))
+                if reaction_filter.get("min") != reaction_filter.get("max"):
+                    raise ValueError("Can't handle differing reaction filter criteria")
+                reaction_filters.append(
+                    ("{}-reactions".format(phase_name), reaction_filter.get("min"))
+                )
 
         state_node = phase.find("state")
         if state_node is not None:
@@ -549,7 +575,23 @@ def convert(inpfile, outfile):
 
         reaction_data.append(reaction_attribs)
 
-    yaml_doc = {"phases": phases, "species": species_data, "reactions": reaction_data}
+    output_reactions = {}
+    for phase_name, pattern in reaction_filters:
+        pattern = re.compile(pattern.replace("*", ".*"))
+        hits = []
+        misses = []
+        for reaction in reaction_data:
+            if pattern.match(reaction.get("id", "")):
+                hits.append(reaction)
+            else:
+                misses.append(reaction)
+        reaction_data = misses
+        output_reactions[phase_name] = hits
+    if reaction_data:
+        output_reactions["reactions"] = reaction_data
+
+    yaml_doc = {"phases": phases, "species": species_data}
+    yaml_doc.update(output_reactions)
     yaml_obj = yaml.YAML(typ="safe")
     yaml_obj.dump(yaml_doc, Path(outfile))