mmif-python 0.5.0__tar.gz → 0.5.2__tar.gz

{mmif-python-0.5.0/mmif_python.egg-info → mmif-python-0.5.2}/PKG-INFO

@@ -1,16 +1,12 @@
 Metadata-Version: 2.1
 Name: mmif-python
-Version: 0.5.0
+Version: 0.5.2
 Summary: Python implementation of MultiMedia Interchange Format specification. (https://mmif.clams.ai)
 Home-page: https://mmif.clams.ai
 Author: Brandeis Lab for Linguistics and Computation
 Author-email: admin@clams.ai
 License: UNKNOWN
-Description: # MMIF for python 
-        
-        **NOTE** that this project is in pre-alpha and being actively developed. Nothing is guaranteed to reliably work for the moment and developer need to be very careful when using APIs implemented here. Please use [the issue track](https://github.com/clamsproject/mmif/issues) to report bugs and malfunctions.
-        
-        ## MultiMedia Interchange Format
+Description: ## MultiMedia Interchange Format
         [MMIF](https://mmif.clams.ai) is a JSON(-LD)-based data format designed for transferring annotation data between computational analysis applications in [CLAMS project](https://clams.ai). 
         
         

mmif-python-0.5.2/README.md

@@ -0,0 +1,3 @@
+## MultiMedia Interchange Format
+[MMIF](https://mmif.clams.ai) is a JSON(-LD)-based data format designed for transferring annotation data between computational analysis applications in [CLAMS project](https://clams.ai). 
+

mmif-python-0.5.2/VERSION

@@ -0,0 +1 @@
+0.5.2

{mmif-python-0.5.0 → mmif-python-0.5.2}/mmif/serialize/annotation.py

@@ -113,7 +113,10 @@ class Document(Annotation):
     :param document_obj: the JSON data that defines the document
     """
     def __init__(self, doc_obj: Optional[Union[bytes, str, dict]] = None) -> None:
+        # to store the parent view ID
         self._parent_view_id = ''
+        self.reserved_names.add('_parent_view_id')
+        
         self._type: Union[ThingTypesBase, DocumentTypesBase] = ThingTypesBase('')
         self.properties: DocumentProperties = DocumentProperties()
         self.disallow_additional_properties()

{mmif-python-0.5.0 → mmif-python-0.5.2}/mmif/serialize/mmif.py

@@ -68,6 +68,31 @@ class Mmif(MmifObject):
             json_str = json.loads(json_str)
         jsonschema.validators.validate(json_str, schema)
 
+    def serialize(self, pretty: bool = False, sanitize: bool = False) -> str:
+        if sanitize:
+            self.sanitize()
+        return super().serialize(pretty)
+
+    def sanitize(self):
+        """
+        Sanitizes a Mmif object by running some safeguards.
+        Concretely, it performs the following before returning the JSON string.
+        
+        #. validating output using built-in MMIF jsonschema
+        #. remove non-existing annotation types from ``contains`` metadata
+    
+        """
+        for view in self.views:
+            existing_at_types = set(annotation.at_type for annotation in view.annotations)
+            to_pop = set()
+            for contains_at_type in view.metadata.contains.keys():
+                if contains_at_type not in existing_at_types:
+                    to_pop.add(contains_at_type)
+            for key in to_pop:
+                view.metadata.contains.pop(key)
+        serialized = self.serialize()
+        self.validate(serialized)
+
     def new_view_id(self) -> str:
         """
         Fetches an ID for a new view.
@@ -241,14 +266,14 @@ class Mmif(MmifObject):
         :return: a dict that keyed by view IDs (str) and has lists of alignment Annotation objects as values.
         """
         v_and_a = {}
-        at_type1 = ThingTypesBase.from_str(at_type1) if isinstance(at_type1, str) else at_type1
-        at_type2 = ThingTypesBase.from_str(at_type2) if isinstance(at_type2, str) else at_type2
+        at_type1, at_type2 = [ThingTypesBase.from_str(x) if isinstance(x, str) else x for x in (at_type1, at_type2)]
         assert at_type1 != at_type2, f"Alignment must be between two different types, given only one: {at_type1}"
         for alignment_view in self.get_all_views_contain(AnnotationTypes.Alignment):
             alignments = []
             contains_meta = alignment_view.metadata.contains[AnnotationTypes.Alignment]
             if 'sourceType' in contains_meta and 'targetType' in contains_meta:
-                aligned_types = list({contains_meta['sourceType'], contains_meta['targetType']})
+                aligned_types = [ThingTypesBase.from_str(x) 
+                                 for x in {contains_meta['sourceType'], contains_meta['targetType']}]
                 if len(aligned_types) == 2 and at_type1 in aligned_types and at_type2 in aligned_types:
                     alignments.extend(alignment_view.annotations)
             else:

{mmif-python-0.5.0 → mmif-python-0.5.2}/mmif/serialize/model.py

@@ -37,7 +37,7 @@ class MmifObject(object):
     an actual representation with a JSON formatted string or equivalent
     `dict` object argument.
 
-    This superclass has three specially designed instance variables, and these
+    This superclass has four specially designed instance variables, and these
     variable names cannot be used as attribute names for MMIF objects.
 
     1. _unnamed_attributes:
@@ -54,12 +54,15 @@ class MmifObject(object):
        When serialize, an object will skip its *empty* (e.g. zero-length, or None)
        attributes unless they are in this list. Otherwise, the serialized JSON
        string would have empty representations (e.g. ``""``, ``[]``).
+    4. _exclude_from_diff:
+       This is a simple list of names of attributes that should be excluded from
+       the diff calculation in ``__eq__``. 
 
     # TODO (krim @ 8/17/20): this dict is however, a duplicate with the type hints in the class definition.
     Maybe there is a better way to utilize type hints (e.g. getting them as a programmatically), but for now
     developers should be careful to add types to hints as well as to this dict.
 
-    Also note that those two special attributes MUST be set in the __init__()
+    Also note that those special attributes MUST be set in the __init__()
     before calling super method, otherwise deserialization will not work.
 
     And also, a subclass that has one or more *named* attributes, it must
@@ -74,17 +77,15 @@ class MmifObject(object):
      an ID value automatically generated, based on its parent object.
     """
     
-    reserved_names: Set = {
+    reserved_names: Set[str] = {
         'reserved_names',
         '_unnamed_attributes',
         '_attribute_classes',
         '_required_attributes',
-        # used in Document class to store parent view id
-        '_parent_view_id', 
-        # used in View class to autogenerate annotation ids
-        '_id_counts'
+        '_exclude_from_diff'
     }
     _unnamed_attributes: Optional[dict]
+    _exclude_from_diff: Set[str]
     _attribute_classes: Dict[str, Type] = {}  # Mapping: str -> Type
 
     def __init__(self, mmif_obj: Optional[Union[bytes, str, dict]] = None) -> None:
@@ -92,6 +93,8 @@ class MmifObject(object):
             mmif_obj = mmif_obj.decode('utf8')
         if not hasattr(self, '_required_attributes'):
             self._required_attributes = []
+        if not hasattr(self, '_exclude_from_diff'):
+            self._exclude_from_diff = set()
         if not hasattr(self, '_unnamed_attributes'):
             self._unnamed_attributes = {}
         if mmif_obj is not None:
@@ -253,7 +256,10 @@ class MmifObject(object):
 
     def __eq__(self, other) -> bool:
         return isinstance(other, type(self)) and \
-               len(DeepDiff(self, other, report_repetition=True, exclude_types=[datetime])) == 0
+               len(DeepDiff(self, other, report_repetition=True, exclude_types=[datetime],
+                            # https://github.com/clamsproject/mmif-python/issues/214
+                            exclude_paths=self._exclude_from_diff)
+                   ) == 0
 
     def __len__(self) -> int:
         """

{mmif-python-0.5.0 → mmif-python-0.5.2}/mmif/serialize/view.py

@@ -31,7 +31,11 @@ class View(MmifObject):
     """
 
     def __init__(self, view_obj: Optional[Union[bytes, str, dict]] = None) -> None:
+        # used to autogenerate annotation ids
         self._id_counts = {}
+        self.reserved_names.add("_id_counts")
+        self._exclude_from_diff = {"_id_counts"}
+        
         self.id: str = ''
         self.metadata: ViewMetadata = ViewMetadata()
         self.annotations: AnnotationsList = AnnotationsList()
@@ -246,12 +250,21 @@ class ViewMetadata(MmifObject):
             'error': ErrorDict,
             'contains': ContainsDict
         }
-        # in theory, either `contains` or `error` should appear in a `view`
+        # in theory, *oneOf* `contains`, `error`, or `warnings` should appear in a `view`
         # but with current implementation, there's no easy way to set a condition 
         # for `oneOf` requirement 
         # see MmifObject::_required_attributes in model.py 
+        # also see this class' `_serialize()` override implementation
         super().__init__(viewmetadata_obj)
 
+    def _serialize(self, alt_container: Optional[Dict] = None) -> dict:
+        serialized = super()._serialize()
+        # `_serialize()` eliminates any *empty* attributes, so 
+        # when no "contains", "errors", nor "warnings", at least add an empty contains back
+        if not (self.contains.items() or self.error or self.warnings):
+            serialized['contains'] = {}
+        return serialized
+
     def new_contain(self, at_type: Union[str, ThingTypesBase], **contains_metadata) -> Optional['Contain']:
         """
         Adds a new element to the ``contains`` dictionary.
@@ -364,4 +377,7 @@ class ContainsDict(DataDict[ThingTypesBase, Contain]):
         return self._items.get(key, default)
     
     def __contains__(self, item: Union[str, ThingTypesBase]):
-        return item in list(self._items.keys())
+        return item in self._items
+
+    def pop(self, key):
+        self._items.pop(key)

mmif-python-0.5.2/mmif/ver/__init__.py

@@ -0,0 +1,2 @@
+__version__ = "0.5.2"
+__specver__ = "0.5.0"

{mmif-python-0.5.0 → mmif-python-0.5.2}/mmif/vocabulary/__init__.py

@@ -5,3 +5,5 @@ from .base_types import AnnotationTypesBase
 from .base_types import DocumentTypesBase
 from .annotation_types import AnnotationTypes
 from .document_types import DocumentTypes
+
+typevers = {**ThingType.typevers, **AnnotationTypes.typevers, **DocumentTypes.typevers}

{mmif-python-0.5.0 → mmif-python-0.5.2}/mmif/vocabulary/base_types.py

@@ -1,6 +1,8 @@
 # This file is auto-generated by setup.py
+import warnings
 from typing import ClassVar
 
+
 class TypesBase(object):
     """
     Base class for arbitrary vocabulary type. 
@@ -10,8 +12,10 @@ class TypesBase(object):
     _prefixes = {}
     base_uri: str
     shortname: str
-    
+    fuzzy_eq: bool
+
     def __init__(self, type_uri: str):
+        self.fuzzy_eq = False
         self.parse_names(type_uri)
         if self.__repr__() not in self.__class__._prefixes:
             # prepare auto-inferred prefix for Annotations.[].id field
@@ -109,6 +113,10 @@ class ClamsTypesBase(TypesBase):
     version: str
     dev_version: ClassVar[str] = 'develop'
     
+    def __init__(self, type_uri, fuzzymode=True):
+        super().__init__(type_uri)
+        self.fuzzy_eq = fuzzymode
+    
     def parse_names(self, type_uri:str):
         if not type_uri:
             self.base_uri, self.shortname, self.version = "", "", ""
@@ -130,27 +138,38 @@ class ClamsTypesBase(TypesBase):
 
     def __hash__(self):
         return hash(str(self))
+    
+    def normalize_attype_vers(self):
+        if '.' in self.version and self._check_legacy_version(self.version):
+            # legacy mapping: see https://github.com/clamsproject/mmif/issues/14#issuecomment-1504439497
+            if self.version == '0.4.2' and self.shortname == 'Annotation':
+                return 'v2'
+            elif self.version.startswith('0.4.'):
+                return 'v1'
+        else:
+            return self.version
 
-    def __eq__(self, other):
-        if isinstance(other, str):
-            other = ThingTypesBase.from_str(other)
+    def _eq_internal(self, other, fuzzy):
         if isinstance(other, ClamsTypesBase):
-            vers = []
-            for o in (self, other):
-                if '.' in o.version and self._check_legacy_version(o.version):
-                    # legacy mapping: see https://github.com/clamsproject/mmif/issues/14#issuecomment-1504439497
-                    if o.version == '0.4.2' and o.shortname == 'Annotation':
-                        vers.append('v2')
-                    elif o.version.startswith('0.4.'):
-                        vers.append('v1')
-                else:
-                    vers.append(o.version)
-            return self.base_uri == other.base_uri \
-                and self.shortname == other.shortname \
-                and vers[0] == vers[1]
+            if fuzzy:
+                vers = [x.normalize_attype_vers() for x in (self, other)]
+                if self.base_uri == other.base_uri and self.shortname == other.shortname:
+                    if vers[0] != vers[1]:
+                        warnings.warn(f'version difference detected: {self.version}&{other.version}@{self.shortname}')
+                    return True
+                return False
+            else:
+                return self.base_uri == other.base_uri \
+                    and self.shortname == other.shortname \
+                    and self.version == other.version
         else:
             return False
     
+    def __eq__(self, other):
+        if isinstance(other, str):
+            other = ThingTypesBase.from_str(other)
+        return self._eq_internal(other, self.fuzzy_eq and other.fuzzy_eq)
+    
     def __repr__(self):
         if self.version[0] == 'v':
             return f'{self.base_uri}/vocabulary/{self.shortname}/{self.version}'

{mmif-python-0.5.0 → mmif-python-0.5.2/mmif_python.egg-info}/PKG-INFO

@@ -1,16 +1,12 @@
 Metadata-Version: 2.1
 Name: mmif-python
-Version: 0.5.0
+Version: 0.5.2
 Summary: Python implementation of MultiMedia Interchange Format specification. (https://mmif.clams.ai)
 Home-page: https://mmif.clams.ai
 Author: Brandeis Lab for Linguistics and Computation
 Author-email: admin@clams.ai
 License: UNKNOWN
-Description: # MMIF for python 
-        
-        **NOTE** that this project is in pre-alpha and being actively developed. Nothing is guaranteed to reliably work for the moment and developer need to be very careful when using APIs implemented here. Please use [the issue track](https://github.com/clamsproject/mmif/issues) to report bugs and malfunctions.
-        
-        ## MultiMedia Interchange Format
+Description: ## MultiMedia Interchange Format
         [MMIF](https://mmif.clams.ai) is a JSON(-LD)-based data format designed for transferring annotation data between computational analysis applications in [CLAMS project](https://clams.ai). 
         
         

{mmif-python-0.5.0 → mmif-python-0.5.2}/setup.py

@@ -134,7 +134,7 @@ def generate_vocabulary(spec_version, clams_types_vers):
             f"from .{mod_name} import {class_name}"
             for mod_name, classes in types.items()
             for class_name in classes
-        )+'\n'
+        ) + '\n\n' + "typevers = {**ThingType.typevers, **AnnotationTypes.typevers, **DocumentTypes.typevers}" + '\n'
     )
     document_types = []
     annotation_types = []
@@ -161,7 +161,7 @@ def generate_vocabulary(spec_version, clams_types_vers):
     return vocabulary_dir
 
 
-def get_latest_mmif_gittag(version: str):
+def get_latest_mmif_gittag():
     # vmaj, vmin, vpat = version.split('.')[0:3]
     res = request.urlopen('https://api.github.com/repos/clamsproject/mmif/git/refs/tags')
     body = json.loads(res.read())
@@ -200,7 +200,7 @@ def prep_ext_files(setuptools_cmd):
     def mod_run(self):
         # will infer the `spec_ver` from the latest git tag available either on GH or local `mmif` repository.
         # NOTE that when `make develop`, it will use specification files from upstream "develop" branch of `mmif` repo
-        latest_mmif_gittag = get_latest_mmif_gittag(version)
+        latest_mmif_gittag = get_latest_mmif_gittag()
         spec_file_gitref = latest_mmif_gittag if '.dev' not in version else 'develop'
         # legacy version tags were formatted as xx-a.b.c (e.g., vocab-0.0.1)
         spec_version = latest_mmif_gittag.split('-')[-1]

mmif-python-0.5.0/README.md

@@ -1,7 +0,0 @@
-# MMIF for python 
-
-**NOTE** that this project is in pre-alpha and being actively developed. Nothing is guaranteed to reliably work for the moment and developer need to be very careful when using APIs implemented here. Please use [the issue track](https://github.com/clamsproject/mmif/issues) to report bugs and malfunctions.
-
-## MultiMedia Interchange Format
-[MMIF](https://mmif.clams.ai) is a JSON(-LD)-based data format designed for transferring annotation data between computational analysis applications in [CLAMS project](https://clams.ai). 
-

mmif-python-0.5.0/VERSION

@@ -1 +0,0 @@
-0.5.0

mmif-python-0.5.0/mmif/ver/__init__.py

@@ -1,2 +0,0 @@
-__version__ = "0.5.0"
-__specver__ = "0.5.0"