Ocr showcase2

docs/ch9.rst

@@ -6,3 +6,4 @@ Chapter 9. Tasks on Other modalities
    toc/audio_processing.md
 
    toc/image_processing.md
+   notebook_tutorial/ocr.ipynb

forte/data/data_pack.py

@@ -32,7 +32,6 @@
 
 import numpy as np
 from sortedcontainers import SortedList
-
 from forte.common.exception import (
     ProcessExecutionException,
     UnknownOntologyClassException,
@@ -47,13 +46,13 @@
 from forte.data.ontology.core import EntryType
 from forte.data.ontology.top import (
     Annotation,
+    Grids,
     Link,
     Group,
     SinglePackEntries,
     Generics,
     AudioAnnotation,
     ImageAnnotation,
-    Grids,
     Payload,
 )
 
@@ -171,8 +170,8 @@ def __init__(self, pack_name: Optional[str] = None):
         self._data_store: DataStore = DataStore()
         self._entry_converter: EntryConverter = EntryConverter()
         self.image_annotations: List[ImageAnnotation] = []
-        self.grids: List[Grids] = []
 
+        self.grids: List[Grids] = []
         self.text_payloads: List[Payload] = []
         self.audio_payloads: List[Payload] = []
         self.image_payloads: List[Payload] = []
@@ -244,7 +243,7 @@ def text(self) -> str:
     @property
     def audio(self) -> Optional[np.ndarray]:
         r"""Return the audio of the data pack"""
-        return self.get_payload_data_at(Modality.Audio, 0)
+        return cast(np.ndarray, self.get_payload_data_at(Modality.Audio, 0))
 
     @property
     def all_annotations(self) -> Iterator[Annotation]:
@@ -448,15 +447,12 @@ def get_payload_at(
         supported_modality = [enum.name for enum in Modality]
 
         try:
-            # if modality.name == "text":
             if modality == Modality.Text:
                 payloads_length = len(self.text_payloads)
                 payload = self.text_payloads[payload_index]
-            # elif modality.name == "audio":
             elif modality == Modality.Audio:
                 payloads_length = len(self.audio_payloads)
                 payload = self.audio_payloads[payload_index]
-            # elif modality.name == "image":
             elif modality == Modality.Image:
                 payloads_length = len(self.image_payloads)
                 payload = self.image_payloads[payload_index]
@@ -569,7 +565,7 @@ def set_text(
         # temporary solution for backward compatibility
         # past API use this method to add a single text in the datapack
         if len(self.text_payloads) == 0 and text_payload_index == 0:
-            from ft.onto.base_ontology import (  # pylint: disable=import-outside-toplevel
+            from ft.onto.payload_ontology import (  # pylint: disable=import-outside-toplevel
                 TextPayload,
             )
 
@@ -601,7 +597,7 @@ def set_audio(
         # temporary solution for backward compatibility
         # past API use this method to add a single audio in the datapack
         if len(self.audio_payloads) == 0 and audio_payload_index == 0:
-            from ft.onto.base_ontology import (  # pylint: disable=import-outside-toplevel
+            from ft.onto.payload_ontology import (  # pylint: disable=import-outside-toplevel
                 AudioPayload,
             )
 

forte/data/ontology/top.py

@@ -63,6 +63,7 @@
     "Box",
     "BoundingBox",
     "Payload",
+    "Meta",
 ]
 
 QueryType = Union[Dict[str, Any], np.ndarray]
@@ -1253,7 +1254,7 @@ def __init__(
         payload_idx: int = 0,
         uri: Optional[str] = None,
     ):
-        from ft.onto.base_ontology import (  # pylint: disable=import-outside-toplevel
+        from ft.onto.payload_ontology import (  # pylint: disable=import-outside-toplevel
             TextPayload,
             AudioPayload,
             ImagePayload,
@@ -1282,6 +1283,12 @@ def __init__(
         self.replace_back_operations: Sequence[Tuple] = []
         self.processed_original_spans: Sequence[Tuple] = []
         self.orig_text_len: int = 0
+        self.payloading = None
+        self.meta = None
+
+    def set_meta(self, meta):
+        # there might be a better way to set meta
+        self.meta = meta
 
     def get_type(self) -> type:
         """
@@ -1324,11 +1331,22 @@ def payload_index(self) -> int:
 
     @property
     def uri(self) -> Optional[str]:
+        """
+        Universal resource identifier of the data source.
+
+        Returns:
+            Optional[str]: Universal resource identifier of the data source.
+        """
         return self._uri
 
+    def load(self):
+        fn = self.payloading.route(self.meta)
+        self._cache = fn(self.uri)
+
     def set_cache(self, data: Union[str, np.ndarray]):
         """
-        Load cache data into the payload.
+        Set cache data in the the payload. This method can be useful when users
+        want to set a new cache.
 
         Args:
             data: data to be set in the payload. It can be str for text data or
@@ -1365,6 +1383,14 @@ def __setstate__(self, state):
         self._modality = getattr(Modality, state["_modality"])
 
 
+class Meta(Generics):
+    def __init__(
+        self, pack: PackType
+    ):  # pylint: disable=useless-super-delegation
+        super().__init__(pack)
+        self.source_type = None  # data source type, it can be 'local` or 'web'
+
+
 SinglePackEntries = (
     Link,
     Group,
@@ -1373,5 +1399,6 @@ def __setstate__(self, state):
     AudioAnnotation,
     ImageAnnotation,
     Payload,
+    Meta,
 )
 MultiPackEntries = (MultiPackLink, MultiPackGroup, MultiPackGeneric)

forte/ontology_specs/base_ontology.json

@@ -444,29 +444,6 @@
           "type": "str"
         }
       ]
-    },
-    {
-      "entry_name": "ft.onto.base_ontology.AudioPayload",
-      "parent_entry": "forte.data.ontology.top.Payload",
-      "description": "A payload that caches audio data",
-      "attributes":[
-        {
-          "name": "sample_rate",
-          "type": "int"
-        }
-      ]
-    },
-    {
-      "entry_name": "ft.onto.base_ontology.TextPayload",
-      "parent_entry": "forte.data.ontology.top.Payload",
-      "description": "A payload that caches text data",
-      "attributes": []
-    },
-    {
-      "entry_name": "ft.onto.base_ontology.ImagePayload",
-      "parent_entry": "forte.data.ontology.top.Payload",
-      "description": "A payload that caches image data",
-      "attributes":[]
     }
   ]
 }

forte/ontology_specs/payload_ontology.json

@@ -0,0 +1,76 @@
+{
+    "name": "payload_ontology",
+    "definitions": [
+      {
+        "entry_name": "ft.onto.payload_ontology.AudioPayload",
+        "parent_entry": "forte.data.ontology.top.Payload",
+        "description": "A payload that caches audio data",
+        "attributes":[
+          {
+            "name": "sample_rate",
+            "type": "int"
+          }
+        ]
+      },
+      {
+        "entry_name": "ft.onto.payload_ontology.TextPayload",
+        "parent_entry": "forte.data.ontology.top.Payload",
+        "description": "A payload that caches text data",
+        "attributes": []
+      },
+      {
+        "entry_name": "ft.onto.payload_ontology.ImagePayload",
+        "parent_entry": "forte.data.ontology.top.Payload",
+        "description": "A payload that caches image data",
+        "attributes":[]
+      },
+        {
+            "entry_name": "ft.onto.payload_ontology.JpegMeta",
+            "parent_entry": "forte.data.ontology.top.Meta",
+            "attributes":[
+              {
+              "name": "extension",
+              "type": "str"
+            },
+              {"name": "mime",
+              "type": "str"},
+              {"name": "type_code",
+              "type": "str"},
+              {"name": "version",
+              "type": "str"}
+             ]
+          },
+          {
+            "entry_name": "ft.onto.payload_ontology.AudioMeta",
+            "parent_entry": "forte.data.ontology.top.Meta",
+            "attributes":[
+              {
+              "name": "sample_rate",
+              "type": "int"
+              },
+              {"name": "channels",
+              "type": "int"},
+              {"name": "bits_per_sample",
+              "type": "int"},
+              {"name": "duration",
+              "type": "float"},
+              {"name": "bitrate",
+              "type": "int"},
+              {"name": "encoding",
+              "type": "str"},
+              {"name": "dtype",
+                "type": "str"}
+             ]
+          },
+          {
+            "entry_name": "ft.onto.payload_ontology.JpegPayload",
+            "parent_entry": "ft.onto.payload_ontology.ImagePayload",
+            "attributes":[
+              {
+                "name": "meta",
+                "type": "ft.onto.payload_ontology.JpegMeta"
+              }
+            ]
+        }
+    ]
+}