Fixes for FSaccess in provenance

Author: mb1069

cwltool/executors.py

@@ -159,7 +159,8 @@ def run_jobs(self,
                 user_provenance=False,
                 orcid=runtime_context.orcid,
                 # single tool execution, so RO UUID = wf UUID = tool UUID
-                run_uuid=runtime_context.research_obj.ro_uuid)
+                run_uuid=runtime_context.research_obj.ro_uuid,
+                fsaccess=runtime_context.make_fs_access(''))
             process.parent_wf = process.provenance_object
         jobiter = process.job(job_order_object, self.output_callback,
                               runtime_context)

cwltool/provenance.py

@@ -305,9 +305,14 @@ def __init__(self,
                  host_provenance,        # type: bool
                  user_provenance,        # type: bool
                  orcid,                  # type: str
-                 run_uuid=None           # type: Optional[uuid.UUID]
+                 run_uuid=None,           # type: Optional[uuid.UUID],
+                 fsaccess=None           # type: Optional[StdFsAccess]
                 ):  # type: (...) -> None
         """Initialize the provenance profile."""
+        if not fsaccess:
+            fsaccess = StdFsAccess('')
+        self.fsaccess = fsaccess
+
         self.orcid = orcid
         self.research_object = research_object
         self.folder = self.research_object.folder
@@ -504,8 +509,7 @@ def declare_file(self, value):
         if not entity and 'location' in value:
             location = str(value['location'])
             # If we made it here, we'll have to add it to the RO
-            fsaccess = StdFsAccess("")
-            with fsaccess.open(location, "rb") as fhandle:
+            with self.fsaccess.open(location, "rb") as fhandle:
                 relative_path = self.research_object.add_data_file(fhandle)
                 # FIXME: This naively relies on add_data_file setting hash as filename
                 checksum = PurePath(relative_path).name
@@ -592,8 +596,7 @@ def declare_directory(self, value):  # type: (MutableMapping[Text, Any]) -> Prov
         is_empty = True
 
         if "listing" not in value:
-            fsaccess = StdFsAccess("")
-            get_listing(fsaccess, value)
+            get_listing(self.fsaccess, value)
         for entry in value.get("listing", []):
             is_empty = False
             # Declare child-artifacts
@@ -937,8 +940,9 @@ def finalize_prov_profile(self, name):
 class ResearchObject():
     """CWLProv Research Object."""
 
-    def __init__(self, temp_prefix_ro="tmp", orcid='', full_name=''):
+    def __init__(self, temp_prefix_ro="tmp", orcid='', full_name='', fsaccess=None):
         # type: (str, Text, Text) -> None
+        # type: Optional[StdFsAccess]
         """Initialize the ResearchObject."""
         self.temp_prefix = temp_prefix_ro
         self.orcid = '' if not orcid else _valid_orcid(orcid)
@@ -954,7 +958,7 @@ def __init__(self, temp_prefix_ro="tmp", orcid='', full_name=''):
         self._external_aggregates = []  # type: List[Dict[Text, Text]]
         self.annotations = []  # type: List[Dict[Text, Any]]
         self._content_types = {}  # type: Dict[Text,str]
-
+        self.fsaccess = fsaccess
         # These should be replaced by generate_prov_doc when workflow/run IDs are known:
         self.engine_uuid = "urn:uuid:%s" % uuid.uuid4()
         self.ro_uuid = uuid.uuid4()
@@ -1553,8 +1557,7 @@ def _relativise_files(self, structure):
                     # Register in RO; but why was this not picked
                     # up by used_artefacts?
                     _logger.info("[provenance] Adding to RO %s", structure["location"])
-                    fsaccess = StdFsAccess("")
-                    with fsaccess.open(structure["location"], "rb") as fp:
+                    with self.fsaccess.open(structure["location"], "rb") as fp:
                         relative_path = self.add_data_file(fp)
                         checksum = PurePosixPath(relative_path).name
                         structure["checksum"] = "%s$%s" % (SHA1, checksum)

cwltool/workflow.py

@@ -522,7 +522,8 @@ def __init__(self,
                 host_provenance=loadingContext.host_provenance,
                 user_provenance=loadingContext.user_provenance,
                 orcid=loadingContext.orcid,
-                run_uuid=run_uuid)  # inherit RO UUID for master wf run
+                run_uuid=run_uuid,
+                fsaccess=loadingContext.research_obj.fsaccess)  # inherit RO UUID for master wf run
             # TODO: Is Workflow(..) only called when we are the master workflow?
             self.parent_wf = self.provenance_object