+def make_wrapper_workflow(arvRunner, main, packed, project_uuid, name, git_info, tool):
+ col = arvados.collection.Collection(api_client=arvRunner.api,
+ keep_client=arvRunner.keep_client)
+
+ with col.open("workflow.json", "wt") as f:
+ json.dump(packed, f, sort_keys=True, indent=4, separators=(',',': '))
+
+ pdh = col.portable_data_hash()
+
+ toolname = tool.tool.get("label") or tool.metadata.get("label") or os.path.basename(tool.tool["id"])
+ if git_info and git_info.get("http://arvados.org/cwl#gitDescribe"):
+ toolname = "%s (%s)" % (toolname, git_info.get("http://arvados.org/cwl#gitDescribe"))
+
+ existing = arvRunner.api.collections().list(filters=[["portable_data_hash", "=", pdh], ["owner_uuid", "=", project_uuid]]).execute(num_retries=arvRunner.num_retries)
+ if len(existing["items"]) == 0:
+ col.save_new(name=toolname, owner_uuid=project_uuid, ensure_unique_name=True)
+
+ # now construct the wrapper
+
+ step = {
+ "id": "#main/" + toolname,
+ "in": [],
+ "out": [],
+ "run": "keep:%s/workflow.json#main" % pdh,
+ "label": name
+ }
+
+ newinputs = []
+ for i in main["inputs"]:
+ inp = {}
+ # Make sure to only copy known fields that are meaningful at
+ # the workflow level. In practice this ensures that if we're
+ # wrapping a CommandLineTool we don't grab inputBinding.
+ # Right now also excludes extension fields, which is fine,
+ # Arvados doesn't currently look for any extension fields on
+ # input parameters.
+ for f in ("type", "label", "secondaryFiles", "streamable",
+ "doc", "id", "format", "loadContents",
+ "loadListing", "default"):
+ if f in i:
+ inp[f] = i[f]
+ newinputs.append(inp)
+
+ wrapper = {
+ "class": "Workflow",
+ "id": "#main",
+ "inputs": newinputs,
+ "outputs": [],
+ "steps": [step]
+ }
+
+ for i in main["inputs"]:
+ step["in"].append({
+ "id": "#main/step/%s" % shortname(i["id"]),
+ "source": i["id"]
+ })
+
+ for i in main["outputs"]:
+ step["out"].append({"id": "#main/step/%s" % shortname(i["id"])})
+ wrapper["outputs"].append({"outputSource": "#main/step/%s" % shortname(i["id"]),
+ "type": i["type"],
+ "id": i["id"]})
+
+ wrapper["requirements"] = [{"class": "SubworkflowFeatureRequirement"}]
+
+ if main.get("requirements"):
+ wrapper["requirements"].extend(main["requirements"])
+ if main.get("hints"):
+ wrapper["hints"] = main["hints"]
+
+ doc = {"cwlVersion": "v1.2", "$graph": [wrapper]}
+
+ if git_info:
+ for g in git_info:
+ doc[g] = git_info[g]
+
+ return json.dumps(doc, sort_keys=True, indent=4, separators=(',',': '))
+
+
+def rel_ref(s, baseuri, urlexpander, merged_map, jobmapper):
+ if s.startswith("keep:"):
+ return s
+
+ uri = urlexpander(s, baseuri)
+
+ if uri.startswith("keep:"):
+ return uri
+
+ fileuri = urllib.parse.urldefrag(baseuri)[0]
+
+ for u in (baseuri, fileuri):
+ if u in merged_map:
+ replacements = merged_map[u].resolved
+ if uri in replacements:
+ return replacements[uri]
+
+ if uri in jobmapper:
+ return jobmapper.mapper(uri).target
+
+ p1 = os.path.dirname(uri_file_path(fileuri))
+ p2 = os.path.dirname(uri_file_path(uri))
+ p3 = os.path.basename(uri_file_path(uri))
+
+ r = os.path.relpath(p2, p1)
+ if r == ".":
+ r = ""
+
+ return os.path.join(r, p3)
+
+def is_basetype(tp):
+ basetypes = ("null", "boolean", "int", "long", "float", "double", "string", "File", "Directory", "record", "array", "enum")
+ for b in basetypes:
+ if re.match(b+"(\[\])?\??", tp):
+ return True
+ return False
+
+
+def update_refs(d, baseuri, urlexpander, merged_map, jobmapper, set_block_style, runtimeContext, prefix, replacePrefix):
+ if set_block_style and (isinstance(d, CommentedSeq) or isinstance(d, CommentedMap)):
+ d.fa.set_block_style()
+
+ if isinstance(d, MutableSequence):
+ for i, s in enumerate(d):
+ if prefix and isinstance(s, str):
+ if s.startswith(prefix):
+ d[i] = replacePrefix+s[len(prefix):]
+ else:
+ update_refs(s, baseuri, urlexpander, merged_map, jobmapper, set_block_style, runtimeContext, prefix, replacePrefix)
+ elif isinstance(d, MutableMapping):
+ for field in ("id", "name"):
+ if isinstance(d.get(field), str) and d[field].startswith("_:"):
+ # blank node reference, was added in automatically, can get rid of it.
+ del d[field]
+
+ if "id" in d:
+ baseuri = urlexpander(d["id"], baseuri, scoped_id=True)
+ elif "name" in d and isinstance(d["name"], str):
+ baseuri = urlexpander(d["name"], baseuri, scoped_id=True)
+
+ if d.get("class") == "DockerRequirement":
+ dockerImageId = d.get("dockerImageId") or d.get("dockerPull")
+ d["http://arvados.org/cwl#dockerCollectionPDH"] = runtimeContext.cached_docker_lookups.get(dockerImageId)
+
+ for field in d:
+ if field in ("location", "run", "name") and isinstance(d[field], str):
+ d[field] = rel_ref(d[field], baseuri, urlexpander, merged_map, jobmapper)
+ continue
+
+ if field in ("$include", "$import") and isinstance(d[field], str):
+ d[field] = rel_ref(d[field], baseuri, urlexpander, {}, jobmapper)
+ continue
+
+ for t in ("type", "items"):
+ if (field == t and
+ isinstance(d[t], str) and
+ not is_basetype(d[t])):
+ d[t] = rel_ref(d[t], baseuri, urlexpander, merged_map, jobmapper)
+ continue
+
+ if field == "inputs" and isinstance(d["inputs"], MutableMapping):
+ for inp in d["inputs"]:
+ if isinstance(d["inputs"][inp], str) and not is_basetype(d["inputs"][inp]):
+ d["inputs"][inp] = rel_ref(d["inputs"][inp], baseuri, urlexpander, merged_map, jobmapper)
+ if isinstance(d["inputs"][inp], MutableMapping):
+ update_refs(d["inputs"][inp], baseuri, urlexpander, merged_map, jobmapper, set_block_style, runtimeContext, prefix, replacePrefix)
+ continue
+
+ if field == "$schemas":
+ for n, s in enumerate(d["$schemas"]):
+ d["$schemas"][n] = rel_ref(d["$schemas"][n], baseuri, urlexpander, merged_map, jobmapper)
+ continue
+
+ update_refs(d[field], baseuri, urlexpander, merged_map, jobmapper, set_block_style, runtimeContext, prefix, replacePrefix)
+
+
+def fix_schemadef(req, baseuri, urlexpander, merged_map, jobmapper, pdh):
+ req = copy.deepcopy(req)
+
+ for f in req["types"]:
+ r = f["name"]
+ path, frag = urllib.parse.urldefrag(r)
+ rel = rel_ref(r, baseuri, urlexpander, merged_map, jobmapper)
+ merged_map.setdefault(path, FileUpdates({}, {}))
+ rename = "keep:%s/%s" %(pdh, rel)
+ for mm in merged_map:
+ merged_map[mm].resolved[r] = rename
+ return req
+
+def drop_ids(d):
+ if isinstance(d, MutableSequence):
+ for i, s in enumerate(d):
+ drop_ids(s)
+ elif isinstance(d, MutableMapping):
+ if "id" in d and d["id"].startswith("file:"):
+ del d["id"]
+
+ for field in d:
+ drop_ids(d[field])
+
+