More use of Inventory

ocfl/inventory.py

@@ -54,7 +54,7 @@ def __init__(self, data=None, filepath=None):
         elif isinstance(data, Inventory):
             self.data = copy.deepcopy(data.data)
         elif isinstance(data, dict):
-            self.data = copy.deepcopy(data)
+            self.data = data
         else:
             raise InventoryException("Bad data type supplied to Inventory() creator, " + str(type(data)))
 
@@ -161,8 +161,8 @@ def content(self):
         digests for each file. Essentially an inversion of the manifest.
         """
         files = {}
-        for digest in self.manifest:
-            for file in self.manifest[digest]:
+        for digest, files_for_digest in self.manifest.items():
+            for file in files_for_digest:
                 files[file] = digest
         return files
 
@@ -265,24 +265,27 @@ def content_path_for_digest(self, digest):
             return paths[0]
         return None
 
-    def add_version(self, vdir=None):
+    def add_version(self, vdir=None, metadata=None):
         """Add new version object to the versions block.
 
         Arguments:
             vdir: string with the version directory name (e.g. "v1"). If None
                 then will create the next version in sequence
+            metadata: dict to initialize version metadata with, else None to
+                create empty
 
         Returns a Version object to access version properties.
         """
-        highest_version = 0
-        for vvdir in self.version_directories:
-            highest_version = max(highest_version,
-                                  parse_version_directory(vvdir))
-        # FIXME - Need to deal with zero padding
-        vdir = "v" + str(highest_version + 1)
+        if vdir is None:
+            highest_version = 0
+            for vvdir in self.version_directories:
+                highest_version = max(highest_version,
+                                      parse_version_directory(vvdir))
+            # FIXME - Need to deal with zero padding
+            vdir = "v" + str(highest_version + 1)
         if "versions" not in self.data:
             self.data["versions"] = {}
-        self.data["versions"][vdir] = {}
+        self.data["versions"][vdir] = {} if metadata is None else metadata
         self.data["head"] = vdir
         return self.version(vdir)
 
@@ -311,6 +314,14 @@ def as_json(self):
         """Serlialize JSON representation."""
         return json.dumps(self.data, sort_keys=True, indent=2)
 
+    def write_json(self, fh):
+        """Serialise JSON representation to file.
+
+        Arguments:
+            fh - filehandle to write to
+        """
+        json.dump(self.data, fh, sort_keys=True, indent=2)
+
     def init_manifest_and_versions(self):
         """Initialize manifest and versions blocks for building new inventory."""
         self.manifest = {}
@@ -329,6 +340,22 @@ def add_fixity_type(self, digest_algorithm):
             self.data["fixity"] = {}
         self.data["fixity"][digest_algorithm] = {}
 
+    def add_fixity_data(self, digest_algorithm, digest, filepath):
+        """Add fixity information for a file.
+
+        Arguments:
+            digest_algorithm: string of the digest algorithm specifying this
+                fixity type
+
+        Assumes that there is already fixity block and within that a block for
+        the specific digest_algorithm.
+        """
+        fixities = self.fixity[digest_algorithm]
+        if digest not in fixities:
+            fixities[digest] = [filepath]
+        else:
+            fixities[digest].append(filepath)
+
 
 class Version():
     """Version class to represent version information in an Inventory.

ocfl/object.py

@@ -195,9 +195,9 @@ def add_version(self, *, inventory, src_fs, src_dir, vdir, metadata=None):
         """Add to inventory data for new version based on files in srcdir.
 
         Arguments:
-            inventory: the inventory up to (vdir-1) which must include blocks
-                for ['manifest'] and ['versions']. It must also include
-                a ['fixity'][algorithm] block for every algorithm in self.fixity
+            inventory: an Invenory object with data up to versio (vdir-1) which
+                must include blocks for the manifest and versions. It must also
+                include a fixity block for every algorithm in self.fixity
             src_fs: pyfs filesystem where this new version exist
             src_dir: the version directory in src_fs that files are being added
                from
@@ -210,7 +210,7 @@ def add_version(self, *, inventory, src_fs, src_dir, vdir, metadata=None):
         content for this new version.
         """
         state = {}  # state for this new version
-        manifest = inventory['manifest']
+        manifest = inventory.manifest
         digests_in_version = {}
         manifest_to_srcfile = {}
         # Go through all files to find new files in manifest and state for this version
@@ -245,17 +245,14 @@ def add_version(self, *, inventory, src_fs, src_dir, vdir, metadata=None):
         # Add extra fixity entries if required
         if self.fixity is not None:
             for fixity_type in self.fixity:
-                fixities = inventory['fixity'][fixity_type]
                 for digest, vfilepaths in digests_in_version.items():
                     for vfilepath in vfilepaths:
                         fixity_digest = file_digest(manifest_to_srcfile[vfilepath], fixity_type, pyfs=src_fs)
-                        if fixity_digest not in fixities:
-                            fixities[fixity_digest] = [vfilepath]
-                        else:
-                            fixities[fixity_digest].append(vfilepath)
-        # Set head to this latest version, and add this version to inventory
-        inventory['head'] = vdir
-        inventory['versions'][vdir] = metadata.as_dict(state=state)
+                        inventory.add_fixity_data(digest_algorithm=fixity_type,
+                                                  digest=fixity_digest,
+                                                  filepath=vfilepath)
+        # Add this new version to inventory (also updates head)
+        inventory.add_version(vdir=vdir, metadata=metadata.as_dict(state=state))
         return manifest_to_srcfile
 
     def build_inventory(self, src_fs, versions_metadata=None):
@@ -274,7 +271,7 @@ def build_inventory(self, src_fs, versions_metadata=None):
         """
         if versions_metadata is None:
             versions_metadata = {}
-        inventory = self.start_inventory().data
+        inventory = self.start_inventory()
         # Find the versions
         versions = {}
         for vdir in src_fs.listdir('/'):
@@ -297,7 +294,7 @@ def build_inventory(self, src_fs, versions_metadata=None):
                                                    src_dir=vdir,
                                                    vdir=vdir,
                                                    metadata=metadata)
-            yield (vdir, inventory, manifest_to_srcfile)
+            yield (vdir, inventory.data, manifest_to_srcfile)
 
     def object_declaration_object(self):
         """NAMASTE object declaration Namaste object."""
@@ -310,21 +307,27 @@ def write_object_declaration(self):
         """
         self.object_declaration_object().write(pyfs=self.obj_fs)
 
-    def write_inventory_and_sidecar(self, inventory, vdir='', write_inventory=True):
+    def write_inventory_and_sidecar(self, inventory=None, vdir=''):
         """Write inventory and sidecar to vdir in the current object.
 
+        Arguments:
+            inventory: an Inventory object to write the inventory, else None
+                if only the sidecar should be written (default)
+            vdir: string of the directory name within self.obj_fs that the
+                inventory and sidecar should be written to. Default is ''
+
         Assumes self.obj_fs is open for this object. Will create vdir if that
         does not exist. If vdir is not specified then will write to root of
-        the object.
+        the object filesystem.
 
         Returns the inventory sidecar filename.
         """
         if not self.obj_fs.exists(vdir):
             self.obj_fs.makedir(vdir)
         invfile = fs.path.join(vdir, INVENTORY_FILENAME)
-        if write_inventory:
+        if inventory is not None:
             with self.obj_fs.open(invfile, 'w') as fh:
-                json.dump(inventory, fh, sort_keys=True, indent=2)
+                inventory.write_json(fh)
         digest = file_digest(invfile, self.digest_algorithm, pyfs=self.obj_fs)
         sidecar = fs.path.join(vdir, INVENTORY_FILENAME + '.' + self.digest_algorithm)
         with self.obj_fs.open(sidecar, 'w') as fh:
@@ -336,7 +339,7 @@ def write_inventory_sidecar(self):
 
         Returns the inventory sidecar filename.
         """
-        return self.write_inventory_and_sidecar(None, write_inventory=False)
+        return self.write_inventory_and_sidecar(inventory=None)
 
     def build(self, srcdir, versions_metadata=None, objdir=None):
         """Build an OCFL object with multiple versions.
@@ -367,15 +370,15 @@ def build(self, srcdir, versions_metadata=None, objdir=None):
         for (vdir, inventory, manifest_to_srcfile) in self.build_inventory(src_fs, versions_metadata):
             num_versions += 1
             if objdir is not None:
-                self.write_inventory_and_sidecar(inventory, vdir)
+                self.write_inventory_and_sidecar(Inventory(inventory), vdir)
                 # Copy files into this version
                 for (path, srcfile) in manifest_to_srcfile.items():
                     self.copy_into_object(src_fs, srcfile, path, create_dirs=True)
         # Finally populate the object root
         if objdir is not None:
             # Write object declaration, inventory and sidecar
             self.write_object_declaration()
-            self.write_inventory_and_sidecar(inventory)
+            self.write_inventory_and_sidecar(Inventory(inventory))
             logging.info("Built object %s at %s with %s versions", self.id, objdir, num_versions)
         # Whether object written or not, return the last inventory
         return inventory
@@ -397,16 +400,16 @@ def create(self, srcdir, metadata=None, objdir=None):
             self.open_fs(objdir, create=True)
         inventory = self.start_inventory()
         vdir = 'v1'
-        manifest_to_srcfile = self.add_version(inventory=inventory.data, src_fs=src_fs,
+        manifest_to_srcfile = self.add_version(inventory=inventory, src_fs=src_fs,
                                                src_dir='', vdir=vdir,
                                                metadata=metadata)
         if objdir is None:
             return inventory.data
         # Write out v1 object
-        self.write_inventory_and_sidecar(inventory.data, vdir)
+        self.write_inventory_and_sidecar(inventory, vdir)
         # Write object root with object declaration, inventory and sidecar
         self.write_object_declaration()
-        self.write_inventory_and_sidecar(inventory.data)
+        self.write_inventory_and_sidecar(inventory)
         # Write version files
         for path in inventory.content_paths:
             srcfile = manifest_to_srcfile[path]
@@ -501,15 +504,15 @@ def update(self, objdir, srcdir=None, metadata=None):
             inventory['versions'][head] = metadata.as_dict(state=state)
         else:
             src_fs = open_fs(srcdir)
-            manifest_to_srcfile = self.add_version(inventory=inventory, src_fs=src_fs,
+            manifest_to_srcfile = self.add_version(inventory=Inventory(inventory), src_fs=src_fs,
                                                    src_dir='', vdir=head,
                                                    metadata=metadata)
             # Copy files into this version
             for (path, srcfile) in manifest_to_srcfile.items():
                 self.copy_into_object(src_fs, srcfile, path, create_dirs=True)
         # Write inventory in both root and head version
-        self.write_inventory_and_sidecar(inventory, head)
-        self.write_inventory_and_sidecar(inventory)
+        self.write_inventory_and_sidecar(Inventory(inventory), head)
+        self.write_inventory_and_sidecar(Inventory(inventory))
         # Delete old root inventory sidecar if we changed digest algorithm
         if digest_algorithm != old_digest_algorithm:
             self.obj_fs.remove(INVENTORY_FILENAME + '.' + old_digest_algorithm)

tests/test_object.py

@@ -7,6 +7,7 @@
 import fs
 import fs.tempfs
 
+from ocfl.inventory import Inventory
 from ocfl.object import Object, ObjectException
 from ocfl.version_metadata import VersionMetadata
 
@@ -69,42 +70,41 @@ def test05_add_version(self):
         """Test add_version method."""
         self.maxDiff = None
         oo = Object(digest_algorithm="md5")
-        inventory = {'manifest': {}, 'versions': {}}
+        inventory = Inventory({'manifest': {}, 'versions': {}})
         with open('fixtures/1.0/content/spec-ex-full/v1_inventory.json', 'r', encoding="utf-8") as fh:
             v_inventory = json.load(fh)
         metadata = VersionMetadata(inventory=v_inventory, version='v1')
         src_fs = fs.open_fs('fixtures/1.0/content/spec-ex-full')
         oo.add_version(inventory=inventory, src_fs=src_fs,
                        src_dir='v1', vdir='v1', metadata=metadata)
-        self.assertEqual(inventory['head'], 'v1')
-        self.assertEqual(inventory['manifest'],
+        self.assertEqual(inventory.head, 'v1')
+        self.assertEqual(inventory.manifest,
                          {'184f84e28cbe75e050e9c25ea7f2e939': ['v1/content/foo/bar.xml'],
                           'c289c8ccd4bab6e385f5afdd89b5bda2': ['v1/content/image.tiff'],
                           'd41d8cd98f00b204e9800998ecf8427e': ['v1/content/empty.txt']})
-        self.assertEqual(inventory['versions'],
-                         {"v1":
-                          {'created': '2018-01-01T01:01:01Z',
-                           'message': 'Initial import',
-                           'state': {
-                               '184f84e28cbe75e050e9c25ea7f2e939': ['foo/bar.xml'],
-                               'c289c8ccd4bab6e385f5afdd89b5bda2': ['image.tiff'],
-                               'd41d8cd98f00b204e9800998ecf8427e': ['empty.txt']},
-                           'user': {'address': '[email protected]', 'name': 'Alice'}}})
-        self.assertNotIn('fixity', inventory)
+        self.assertEqual(inventory.versiondata("v1"),
+                         {'created': '2018-01-01T01:01:01Z',
+                          'message': 'Initial import',
+                          'state': {
+                              '184f84e28cbe75e050e9c25ea7f2e939': ['foo/bar.xml'],
+                              'c289c8ccd4bab6e385f5afdd89b5bda2': ['image.tiff'],
+                              'd41d8cd98f00b204e9800998ecf8427e': ['empty.txt']},
+                          'user': {'address': '[email protected]', 'name': 'Alice'}})
+        self.assertNotIn('fixity', inventory.data)
         # Now add second version to check forward delta
         with open('fixtures/1.0/content/spec-ex-full/v2_inventory.json', 'r', encoding="utf-8") as fh:
             v_inventory = json.load(fh)
         metadata = VersionMetadata(inventory=v_inventory, version='v2')
         src_fs = fs.open_fs('fixtures/1.0/content/spec-ex-full/v2')
         oo.add_version(inventory=inventory, src_fs=src_fs,
                        src_dir='', vdir='v2', metadata=metadata)
-        self.assertEqual(inventory['head'], 'v2')
-        self.assertEqual(inventory['manifest'],
+        self.assertEqual(inventory.head, 'v2')
+        self.assertEqual(inventory.manifest,
                          {'184f84e28cbe75e050e9c25ea7f2e939': ['v1/content/foo/bar.xml'],
                           '2673a7b11a70bc7ff960ad8127b4adeb': ['v2/content/foo/bar.xml'],
                           'c289c8ccd4bab6e385f5afdd89b5bda2': ['v1/content/image.tiff'],
                           'd41d8cd98f00b204e9800998ecf8427e': ['v1/content/empty.txt']})
-        self.assertEqual(inventory['versions']['v2'],
+        self.assertEqual(inventory.versiondata('v2'),
                          {'created': '2018-02-02T02:02:02Z',
                           'message': 'Fix bar.xml, remove image.tiff, add empty2.txt',
                           'state': {
@@ -113,7 +113,7 @@ def test05_add_version(self):
                           'user': {'address': '[email protected]', 'name': 'Bob'}})
         # Now with fixity
         oo = Object(digest_algorithm="md5", fixity=['sha1'])
-        inventory = {'manifest': {}, 'versions': {}, 'fixity': {'sha1': {}}}
+        inventory = Inventory({'manifest': {}, 'versions': {}, 'fixity': {'sha1': {}}})
         md1 = VersionMetadata()
         with open('fixtures/1.0/content/spec-ex-full/v1_inventory.json', 'r', encoding="utf-8") as fh:
             v_inventory = json.load(fh)
@@ -126,10 +126,10 @@ def test05_add_version(self):
             'v1/content/empty.txt': 'empty.txt',
             'v1/content/foo/bar.xml': 'foo/bar.xml'
         })
-        self.assertEqual(len(inventory['fixity']['sha1']), 3)
+        self.assertEqual(len(inventory.fixity['sha1']), 3)
         # Test dedupe=False and forward_delta=False settings
         oo = Object(dedupe=False, forward_delta=False, fixity=['md5'])
-        inventory = {'manifest': {}, 'versions': {}, 'fixity': {'md5': {}}}
+        inventory = Inventory({'manifest': {}, 'versions': {}, 'fixity': {'md5': {}}})
         md1 = VersionMetadata(inventory={
             "id": "http://example.org/dedupe_content",
             "versions": {
@@ -150,7 +150,7 @@ def test05_add_version(self):
             'v1/content/empty1.txt': 'v1/empty1.txt',
             'v1/content/empty2.txt': 'v1/empty2.txt',
             'v1/content/empty3.txt': 'v1/empty3.txt'})
-        self.assertEqual(inventory['fixity']['md5'], {"d41d8cd98f00b204e9800998ecf8427e": [
+        self.assertEqual(inventory.fixity['md5'], {"d41d8cd98f00b204e9800998ecf8427e": [
             "v1/content/empty1.txt", "v1/content/empty2.txt", "v1/content/empty3.txt"]})
         # Add a second version which will test for forward_delta=False
         md2 = VersionMetadata(inventory={
@@ -170,7 +170,7 @@ def test05_add_version(self):
         # Because of forward_delta=False we will have an additional copy of the empty file
         self.assertEqual(manifest_to_srcfile, {
             'v2/content/empty4.txt': 'v2/empty4.txt'})
-        self.assertEqual(inventory['fixity']['md5'], {"d41d8cd98f00b204e9800998ecf8427e": [
+        self.assertEqual(inventory.fixity['md5'], {"d41d8cd98f00b204e9800998ecf8427e": [
             "v1/content/empty1.txt", "v1/content/empty2.txt",
             "v1/content/empty3.txt", "v2/content/empty4.txt"]})
 
@@ -201,7 +201,7 @@ def test08_write_inventory_and_sidecar(self):
         """Test write_object_and_sidecar."""
         tmpfs = fs.tempfs.TempFS(identifier='test_write_inventory_and_sidecar')
         oo = Object(obj_fs=tmpfs)
-        oo.write_inventory_and_sidecar({'abc': 'def'})
+        oo.write_inventory_and_sidecar(Inventory({'abc': 'def'}))
         self.assertEqual(set(tmpfs.listdir('')),
                          set(['inventory.json', 'inventory.json.sha512']))
         with tmpfs.open('inventory.json') as fh:
@@ -212,7 +212,7 @@ def test08_write_inventory_and_sidecar(self):
         # and now making directory
         oo = Object(obj_fs=tmpfs)
         invdir = 'xxx'
-        oo.write_inventory_and_sidecar({'gh': 'ik'}, invdir)
+        oo.write_inventory_and_sidecar(Inventory({'gh': 'ik'}), invdir)
         self.assertEqual(set(tmpfs.listdir(invdir)),
                          set(['inventory.json', 'inventory.json.sha512']))
         with tmpfs.open(fs.path.join(invdir, 'inventory.json')) as fh: