Adds Form Element (#4272)

The form class currently maps to NarrativeText. This updates it to Form
and adds a new class for form.

Author: aadland6

CHANGELOG.md

@@ -1,3 +1,6 @@
+## 0.21.10
+- **Add Form Class**: Adds a new form class in elements.py to deal with forms
+
 ## 0.21.9
 - Add a fallback to use the filetype library to recover from incorrect results form libmagic
 

test_unstructured/staging/test_base.py

@@ -21,6 +21,7 @@
     ElementMetadata,
     ElementType,
     FigureCaption,
+    Form,
     Image,
     Link,
     ListItem,
@@ -115,6 +116,16 @@ def test_elements_from_dicts():
     ]
 
 
+def test_elements_from_dicts_form():
+    element_dicts = [
+        {"text": "Applicant Name: Jane Doe", "type": "Form"},
+    ]
+
+    elements = base.elements_from_dicts(element_dicts)
+
+    assert elements == [Form(text="Applicant Name: Jane Doe")]
+
+
 def test_convert_to_csv(tmp_path: str):
     output_csv_path = os.path.join(tmp_path, "isd_data.csv")
     elements = [Title(text="Title 1"), NarrativeText(text="Narrative 1")]

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.21.9"  # pragma: no cover
+__version__ = "0.21.10"  # pragma: no cover

unstructured/documents/elements.py

@@ -900,6 +900,12 @@ class NarrativeText(Text):
     category = "NarrativeText"
 
 
+class Form(Text):
+    """An element for capturing form text."""
+
+    category = "Form"
+
+
 class ListItem(Text):
     """ListItem is a NarrativeText element that is part of a list."""
 
@@ -1000,7 +1006,7 @@ class DocumentData(Text):
     # this mapping favors ensures yolox produces backward compatible categories
     ElementType.ABSTRACT: NarrativeText,
     ElementType.THREADING: NarrativeText,
-    ElementType.FORM: NarrativeText,
+    ElementType.FORM: Form,
     ElementType.VALUE: NarrativeText,
     ElementType.LINK: NarrativeText,
     ElementType.LIST_ITEM: ListItem,