Merge pull request #787 from chinapandaman/PPF-786

PPF-786: implement use full widget name

PyPDFForm/middleware/base.py

@@ -16,6 +16,7 @@ def __init__(
 
         super().__init__()
         self._name = name
+        self.full_name = None
         self._value = value
         self.desc = None
 

PyPDFForm/template.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.pdfmetrics import stringWidth
 
 from .constants import (COMB, DEFAULT_FONT_SIZE, MULTILINE, NEW_LINE_SYMBOL,
-                        WIDGET_TYPES, Annots, MaxLen, Rect)
+                        WIDGET_TYPES, Annots, MaxLen, Rect, Parent, T)
 from .font import (adjust_paragraph_font_size, adjust_text_field_font_size,
                    auto_detect_font, get_text_field_font_color,
                    get_text_field_font_size, text_field_font_size)
@@ -43,7 +43,7 @@ def set_character_x_paddings(
     return widgets
 
 
-def build_widgets(pdf_stream: bytes) -> Dict[str, WIDGET_TYPES]:
+def build_widgets(pdf_stream: bytes, use_full_widget_name: bool) -> Dict[str, WIDGET_TYPES]:
     """Builds a widget dict given a PDF form stream."""
 
     results = {}
@@ -53,6 +53,7 @@ def build_widgets(pdf_stream: bytes) -> Dict[str, WIDGET_TYPES]:
             key = get_widget_key(widget)
             _widget = construct_widget(widget, key)
             if _widget is not None:
+                _widget.full_name = get_widget_full_key(widget)
                 _widget.desc = get_widget_description(widget)
                 if isinstance(_widget, Text):
                     _widget.max_length = get_text_field_max_length(widget)
@@ -73,6 +74,8 @@ def build_widgets(pdf_stream: bytes) -> Dict[str, WIDGET_TYPES]:
                     continue
 
                 results[key] = _widget
+                if _widget.full_name is not None and use_full_widget_name:
+                    results[_widget.full_name] = results[key]
     return results
 
 
@@ -190,6 +193,20 @@ def get_widget_key(widget: dict) -> Union[str, list, None]:
     return result
 
 
+def get_widget_full_key(widget: dict) -> Union[str, None]:
+    """
+    Returns a PDF widget's full annotated key by prepending its
+    parent widget's key.
+    """
+
+    key = get_widget_key(widget)
+
+    if Parent in widget and T in widget[Parent].get_object() and widget[Parent][T] != key:
+        return f"{widget[Parent][T]}.{key}"
+
+    return None
+
+
 def get_widget_alignment(widget: dict) -> Union[str, list, None]:
     """Finds a PDF widget's alignment by pattern matching."""
 

PyPDFForm/wrapper.py

@@ -52,7 +52,7 @@ def fill(
     ) -> FormWrapper:
         """Fills a PDF form."""
 
-        widgets = build_widgets(self.stream) if self.stream else {}
+        widgets = build_widgets(self.stream, False) if self.stream else {}
 
         for key, value in data.items():
             if key in widgets:
@@ -86,13 +86,15 @@ def __init__(
         self.global_font_size = kwargs.get("global_font_size")
         self.global_font_color = kwargs.get("global_font_color")
 
+        self.use_full_widget_name = kwargs.get("use_full_widget_name", False)
+
         self._init_helper()
 
     def _init_helper(self, key_to_refresh: str = None) -> None:
         """Updates all attributes when the state of the PDF stream changes."""
 
         refresh_not_needed = {}
-        new_widgets = build_widgets(self.read()) if self.read() else {}
+        new_widgets = build_widgets(self.read(), self.use_full_widget_name) if self.read() else {}
         for k, v in self.widgets.items():
             if k in new_widgets:
                 new_widgets[k] = v
@@ -254,6 +256,9 @@ def update_widget_key(
     ) -> PdfWrapper:
         """Updates the key of an existed widget on a PDF form."""
 
+        if self.use_full_widget_name:
+            raise NotImplementedError
+
         if defer:
             self._keys_to_update.append((old_key, new_key, index))
             return self
@@ -268,6 +273,9 @@ def update_widget_key(
     def commit_widget_key_updates(self) -> PdfWrapper:
         """Commits all deferred widget key updates on a PDF form."""
 
+        if self.use_full_widget_name:
+            raise NotImplementedError
+
         old_keys = [each[0] for each in self._keys_to_update]
         new_keys = [each[1] for each in self._keys_to_update]
         indices = [each[2] for each in self._keys_to_update]

docs/install.md

@@ -59,6 +59,25 @@ with open("sample_template.pdf", "rb+") as template:
 This adaptation is universal across all APIs of PyPDFForm. So in later sections of the documentation whenever you see 
 a function parameter that's a file path you can safely switch them for a file object or file stream.
 
+## Use full widget name in PDF wrapper (beta)
+
+**NOTE:** This is a beta feature, meaning it still needs to be tested against more PDF forms and may not work for 
+some of them.
+
+According to section 12.7.3.2 found on page 434 of [the PDF standard](https://opensource.adobe.com/dc-acrobat-sdk-docs/pdfstandards/PDF32000_2008.pdf), each PDF form widget can have a fully qualified name that is not explicitly defined but can be constructed following the pattern `<parent_widget_name>.<widget_name>`.
+
+PyPDFForm supports accessing widgets through their full names by simply setting the optional parameter `use_full_widget_name` to `True` when a `PdfWrapper` object is instantiated. Consider [this PDF](https://github.com/chinapandaman/PyPDFForm/raw/master/pdf_samples/sample_template_with_full_key.pdf):
+
+```python
+from PyPDFForm import PdfWrapper
+
+pdf = PdfWrapper("sample_template_with_full_key.pdf", use_full_widget_name=True)
+```
+
+The checkbox widget on the second page with texts `Gain de 2 classes` has a partial name of `0` and a full name of `Gain de 2 classes.0`. By constructing the object like above, you can access the same checkbox through both the partial name and the full name.
+
+**NOTE:** Because each full widget name involves both the widget itself and its parent widget, the methods `update_widget_key` and `commit_widget_key_updates` are disabled and will raise a `NotImplementedError` when invoked through an object that uses full widget names.
+
 ## Write to a file
 
 Lastly, `PdfWrapper` also implements itself similar to an open file object. So you can write the PDF it holds to another 

tests/conftest.py

@@ -188,6 +188,14 @@ def sample_template_with_image_field(pdf_samples):
         return f.read()
 
 
+@pytest.fixture
+def sample_template_with_full_key(pdf_samples):
+    with open(
+        os.path.join(pdf_samples, "sample_template_with_full_key.pdf"), "rb+"
+    ) as f:
+        return f.read()
+
+
 @pytest.fixture
 def dropdown_alignment(pdf_samples):
     with open(

tests/test_use_full_widget_name.py

@@ -0,0 +1,52 @@
+# -*- coding: utf-8 -*-
+
+import pytest
+
+from PyPDFForm import PdfWrapper
+
+
+def test_init(sample_template_with_full_key):
+    obj = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+    assert "Gain de 2 classes.0" in obj.widgets
+    assert obj.widgets["Gain de 2 classes.0"] is obj.widgets["0"]
+
+
+def test_sample_data(sample_template_with_full_key):
+    obj = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+    assert "Gain de 2 classes.0" in obj.sample_data
+    assert obj.sample_data["Gain de 2 classes.0"] == obj.sample_data["0"]
+
+
+def test_fill(sample_template_with_full_key):
+    obj_1 = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+    obj_2 = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+
+    assert obj_1.fill(
+        {
+            "Gain de 2 classes.0": True
+        }
+    ).read() == obj_2.fill(
+        {
+            "0": True
+        }
+    ).read()
+
+
+def test_update_widget_key(sample_template_with_full_key):
+    obj = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+
+    with pytest.raises(NotImplementedError):
+        obj.update_widget_key("0", "foo")
+
+
+def test_commit_widget_key_updates(sample_template_with_full_key):
+    obj = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+
+    with pytest.raises(NotImplementedError):
+        obj.commit_widget_key_updates()
+
+
+def test_schema(sample_template_with_full_key):
+    obj = PdfWrapper(sample_template_with_full_key, use_full_widget_name=True)
+    assert "Gain de 2 classes.0" in obj.schema["properties"]
+    assert obj.schema["properties"]["Gain de 2 classes.0"] == obj.schema["properties"]["0"]