def pack(
document_loader: Loader,
processobj, # type: Union[Dict[str, Any], List[Dict[str, Any]]]
uri, # type: str
metadata, # type: Dict[str, str]
rewrite_out=None, # type: Optional[Dict[str, str]]
): # type: (...) -> Dict[str, Any]
document_loader = SubLoader(document_loader)
document_loader.idx = {}
if isinstance(processobj, MutableMapping):
document_loader.idx[processobj["id"]] = CommentedMap(
processobj.items())
elif isinstance(processobj, MutableSequence):
_, frag = urllib.parse.urldefrag(uri)
for po in processobj:
if not frag:
if po["id"].endswith("#main"):
uri = po["id"]
document_loader.idx[po["id"]] = CommentedMap(po.items())
document_loader.idx[metadata["id"]] = CommentedMap(metadata.items())
def loadref(base, uri):
# type: (Optional[str], str) -> Union[Dict[str, Any], List[Dict[str, Any]], str, None]
return document_loader.resolve_ref(uri, base_url=base)[0]
ids = set() # type: Set[str]
find_ids(processobj, ids)
runs = {uri}
find_run(processobj, loadref, runs)
for f in runs:
find_ids(document_loader.resolve_ref(f)[0], ids)
names = set() # type: Set[str]
if rewrite_out is None:
rewrite = {} # type: Dict[str, str]
else:
rewrite = rewrite_out
mainpath, _ = urllib.parse.urldefrag(uri)
def rewrite_id(r, mainuri):
# type: (str, str) -> None
if r == mainuri:
rewrite[r] = "#main"
elif r.startswith(mainuri) and r[len(mainuri)] in ("#", "/"):
if r[len(mainuri):].startswith("#main/"):
rewrite[r] = "#" + uniquename(r[len(mainuri) + 1:], names)
else:
rewrite[r] = "#" + uniquename("main/" + r[len(mainuri) + 1:],
names)
else:
path, frag = urllib.parse.urldefrag(r)
if path == mainpath:
rewrite[r] = "#" + uniquename(frag, names)
else:
if path not in rewrite:
rewrite[path] = "#" + uniquename(shortname(path), names)
sortedids = sorted(ids)
for r in sortedids:
rewrite_id(r, uri)
packed = CommentedMap(
(("$graph", CommentedSeq()), ("cwlVersion", metadata["cwlVersion"])))
namespaces = metadata.get("$namespaces", None)
schemas = set() # type: Set[str]
if "$schemas" in metadata:
for each_schema in metadata["$schemas"]:
schemas.add(each_schema)
for r in sorted(runs):
dcr, metadata = document_loader.resolve_ref(r)
if isinstance(dcr, CommentedSeq):
dcr = dcr[0]
dcr = cast(CommentedMap, dcr)
if not isinstance(dcr, MutableMapping):
continue
metadata = cast(Dict[str, Any], metadata)
if "$schemas" in metadata:
for s in metadata["$schemas"]:
schemas.add(s)
if dcr.get("class") not in ("Workflow", "CommandLineTool",
"ExpressionTool"):
continue
dc = cast(Dict[str, Any], copy.deepcopy(dcr))
v = rewrite[r]
dc["id"] = v
for n in ("name", "cwlVersion", "$namespaces", "$schemas"):
if n in dc:
del dc[n]
packed["$graph"].append(dc)
if schemas:
packed["$schemas"] = list(schemas)
for r in list(rewrite.keys()):
v = rewrite[r]
replace_refs(packed, rewrite, r + "/" if "#" in r else r + "#",
v + "/")
import_embed(packed, set())
if len(packed["$graph"]) == 1:
# duplicate 'cwlVersion' and $schemas inside $graph when there is only
# a single item because we will print the contents inside '$graph'
# rather than whole dict
packed["$graph"][0]["cwlVersion"] = packed["cwlVersion"]
if schemas:
packed["$graph"][0]["$schemas"] = list(schemas)
# always include $namespaces in the #main
if namespaces:
packed["$graph"][0]["$namespaces"] = namespaces
return packed
