PyPDFForm 5.2.0__tar.gz → 5.2.2__tar.gz

{pypdfform-5.2.0 → pypdfform-5.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PyPDFForm
-Version: 5.2.0
+Version: 5.2.2
 Summary: The Python library & CLI for PDF forms.
 Author: Jinge Li
 License-Expression: MIT
@@ -24,12 +24,12 @@ Classifier: Topic :: Utilities
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: cryptography<49.0.0,>=48.0.0
+Requires-Dist: cryptography<50.0.0,>=48.0.0
 Requires-Dist: fonttools<5.0.0,>=4.63.0
 Requires-Dist: pikepdf<11.0.0,>=10.7.2
 Requires-Dist: pillow<13.0.0,>=12.2.0
 Requires-Dist: pypdf<7.0.0,>=6.12.2
-Requires-Dist: reportlab<5.0.0,>=4.5.1
+Requires-Dist: reportlab<6.0.0,>=4.5.1
 Provides-Extra: cli
 Requires-Dist: typer<1.0.0,>=0.26.1; extra == "cli"
 Requires-Dist: jsonschema<5.0.0,>=4.26.0; extra == "cli"
@@ -125,4 +125,10 @@ The official documentation can be found on [the GitHub page](https://chinapandam
 
 This project is maintained entirely in my spare time. If you like the project please consider starring the GitHub repository. It is the best way to keep me motivated and continue making the project better.
 
-[![Stargazers over time](https://starchart.cc/chinapandaman/PyPDFForm.svg?variant=adaptive)](https://starchart.cc/chinapandaman/PyPDFForm)
+<a href="https://www.star-history.com/?repos=chinapandaman%2FPyPDFForm&type=date&logscale=&legend=top-left">
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&theme=dark&legend=top-left" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&legend=top-left" />
+   <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&legend=top-left" />
+ </picture>
+</a>

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/__init__.py

@@ -18,7 +18,7 @@ from Python code or from the command line.
 
 import logging
 
-__version__ = "5.2.0"
+__version__ = "5.2.2"
 
 from .lib.annotations import Annotations
 from .lib.assets.blank import BlankPage

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/egress.py

@@ -12,6 +12,7 @@ called right before the final PDF byte stream is returned by the wrapper module.
 
 from functools import lru_cache
 from io import BytesIO
+from warnings import catch_warnings, simplefilter
 
 from pikepdf import Pdf
 from pypdf import PdfReader, PdfWriter
@@ -20,7 +21,7 @@ from pypdf.generic import DictionaryObject, NameObject, TextStringObject
 from .constants import JS, XFA, AcroForm, JavaScript, OpenAction, Root, S, Title
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def appearance_streams_handler(pdf: bytes, generate_appearance_streams: bool) -> bytes:
     """
     Handles appearance streams and the /NeedAppearances flag for a PDF form.
@@ -57,7 +58,9 @@ def appearance_streams_handler(pdf: bytes, generate_appearance_streams: bool) ->
         result = f.read()
 
     if generate_appearance_streams:
-        with Pdf.open(BytesIO(result)) as f:
+        # TODO: remove after fixing /Annots /Fields mismatch
+        with Pdf.open(BytesIO(result)) as f, catch_warnings():
+            simplefilter("ignore")
             f.generate_appearance_streams()
             with BytesIO() as r:
                 f.save(r)

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/filler.py

@@ -57,7 +57,9 @@ def signature_image_handler(
     any_image_to_draw = False
     if stream is not None:
         any_image_to_draw = True
-        image_width, image_height = get_image_dimensions(stream)
+        image_width, image_height = (
+            get_image_dimensions(stream) if middleware.preserve_aspect_ratio else (0, 0)
+        )
         x, y, width, height = get_draw_image_resolutions(
             widget, middleware.preserve_aspect_ratio, image_width, image_height
         )

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/font.py

@@ -66,7 +66,7 @@ from .raw.text import RawText
 from .watermark import create_watermarks_and_draw
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def validate_font(font_name: str, ttf_stream: bytes) -> bool:
     """
     Validates a TrueType font stream.
@@ -208,7 +208,7 @@ def temporary_font_registration(
                 del _fonts[rl_name]
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def _get_watermark_with_font(ttf_stream: bytes) -> bytes:
     """
     Creates a watermark PDF with a single space character using the specified font.
@@ -230,6 +230,20 @@ def _get_watermark_with_font(ttf_stream: bytes) -> bytes:
         )[0]
 
 
+@lru_cache(maxsize=128)
+def _compress_ttf(ttf_stream: bytes) -> bytes:
+    """
+    Compresses a TrueType font stream for embedding in a PDF.
+
+    Args:
+        ttf_stream (bytes): The font file data in TTF format.
+
+    Returns:
+        bytes: The compressed font stream.
+    """
+    return compress(ttf_stream)
+
+
 def register_font_acroform(
     pdf: bytes, ttf_stream: bytes, need_appearances: bool
 ) -> tuple:
@@ -266,7 +280,7 @@ def register_font_acroform(
         )
 
     font_file_stream = StreamObject()
-    compressed_ttf = compress(ttf_stream)
+    compressed_ttf = _compress_ttf(ttf_stream)
     font_file_stream.set_data(compressed_ttf)
     font_file_stream.update(
         {
@@ -341,7 +355,7 @@ def register_font_acroform(
         return f.read(), new_font_name
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def _get_base_font_name(ttf_stream: bytes) -> str:
     """
     Extracts the base font name from a TrueType font stream.
@@ -383,7 +397,7 @@ def _get_new_font_name(fonts: dict) -> str:
     return f"{FONT_NAME_PREFIX}{n}"
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def get_all_available_fonts(pdf: bytes) -> dict:
     """
     Retrieves all available fonts from a PDF document's AcroForm.

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/image.py

@@ -7,6 +7,7 @@ calculating the resolutions for drawing an image on a PDF page, taking into
 account whether to preserve the aspect ratio.
 """
 
+from functools import lru_cache
 from io import BytesIO
 from typing import Tuple
 
@@ -15,6 +16,7 @@ from PIL import Image
 from .constants import Rect
 
 
+@lru_cache(maxsize=128)
 def rotate_image(image_stream: bytes, rotation: float | int) -> bytes:
     """
     Rotates an image by a specified angle in degrees.
@@ -32,10 +34,7 @@ def rotate_image(image_stream: bytes, rotation: float | int) -> bytes:
     Returns:
         bytes: The rotated image data as bytes.
     """
-    buff = BytesIO()
-    buff.write(image_stream)
-    buff.seek(0)
-
+    buff = BytesIO(image_stream)
     image = Image.open(buff)
 
     rotated_buff = BytesIO()
@@ -50,12 +49,13 @@ def rotate_image(image_stream: bytes, rotation: float | int) -> bytes:
     return result
 
 
+@lru_cache(maxsize=128)
 def get_image_dimensions(image_stream: bytes) -> Tuple[float, float]:
     """
     Retrieves the width and height of an image from its byte stream.
 
-    This function uses the PIL library to open the image from the provided byte stream
-    and returns its dimensions (width and height) as a tuple of floats.
+    This cached function uses the PIL library to open the image from the provided
+    byte stream and returns its dimensions (width and height) as a tuple of floats.
 
     Args:
         image_stream (bytes): The image data as bytes.
@@ -63,13 +63,8 @@ def get_image_dimensions(image_stream: bytes) -> Tuple[float, float]:
     Returns:
         Tuple[float, float]: The width and height of the image in pixels.
     """
-    buff = BytesIO()
-    buff.write(image_stream)
-    buff.seek(0)
-
-    image = Image.open(buff)
-
-    return image.size
+    with Image.open(BytesIO(image_stream)) as image:
+        return image.size
 
 
 def get_draw_image_resolutions(

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/patterns.py

@@ -333,7 +333,8 @@ def get_dropdown_value(annot: DictionaryObject, widget: Dropdown) -> None:
     This function determines the current value of the dropdown, considering
     whether it's a child annotation or a top-level one. It then iterates
     through the widget's choices to find a match and sets the widget's
-    value to the index of the matched choice.
+    value to the index of the matched choice, or None when the first choice
+    is selected.
 
     Args:
         annot (DictionaryObject): The dropdown annotation dictionary.

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/raw/text.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 """
-Contains the RawText class, which represents a text annotation
+Contains the RawText class, which represents a text element
 that can be drawn directly onto a PDF page without relying on existing form fields.
 """
 
@@ -38,7 +38,7 @@ class RawText:
             y: The y-coordinate (vertical position) of the text.
             font: The name of the font to use for the text (defaults to DEFAULT_FONT).
             font_size: The size of the font (defaults to DEFAULT_FONT_SIZE).
-            font_color: The color of the text as an RGB tuple (0-255 for each channel).
+            font_color: The color of the text as an RGB tuple (0-1 for each channel).
         """
         super().__init__()
 

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/template.py

@@ -8,6 +8,7 @@ and defines specific patterns for identifying and constructing different
 types of widgets.
 """
 
+from copy import deepcopy
 from functools import lru_cache
 from io import BytesIO
 from typing import Dict, List, cast
@@ -41,7 +42,7 @@ from .patterns import (
 from .utils import extract_widget_property, find_pattern_match
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def get_metadata(pdf: bytes) -> dict:
     """
     Retrieves the metadata of a PDF.
@@ -64,13 +65,11 @@ def build_widgets(
     use_full_widget_name: bool,
 ) -> Dict[str, WIDGET_TYPES]:
     """
-    Builds a dictionary of widgets from a PDF stream.
+    Builds an independent dictionary of widgets from a PDF stream.
 
-    This function parses a PDF stream to identify and construct widgets
-    present in the PDF form. It iterates through each page and its annotations,
-    extracting widget properties such as key, description, max length (for text fields),
-    and choices (for dropdowns). The constructed widgets are stored in a dictionary
-    where the keys are the widget keys and the values are the widget objects.
+    Widget discovery and construction are cached internally, then deep-copied
+    before returning so callers can safely mutate widget attributes without
+    changing cached objects or widgets returned by other calls.
 
     Args:
         pdf_stream (bytes): The PDF stream to parse.
@@ -81,6 +80,28 @@ def build_widgets(
         Dict[str, WIDGET_TYPES]: A dictionary of widgets, where keys are widget
             keys and values are widget objects.
     """
+    return deepcopy(_build_widget_cache(pdf_stream, use_full_widget_name))
+
+
+@lru_cache(maxsize=128)
+def _build_widget_cache(
+    pdf_stream: bytes,
+    use_full_widget_name: bool,
+) -> Dict[str, WIDGET_TYPES]:
+    """
+    Builds and caches reusable widget objects from a PDF stream.
+
+    The cached widgets must be treated as prototypes only. Use `build_widgets`
+    to get independent copies that are safe to mutate.
+
+    Args:
+        pdf_stream (bytes): The PDF stream to parse.
+        use_full_widget_name (bool): Whether to use the full widget name
+            (including parent names) as the widget key.
+
+    Returns:
+        Dict[str, WIDGET_TYPES]: Cached widget prototypes keyed by widget name.
+    """
     results = {}
 
     for page_num, widgets in get_widgets_by_page(pdf_stream).items():
@@ -220,7 +241,7 @@ def _handle_radio_widget(
         radio.value = radio.number_of_options - 1
 
 
-@lru_cache()
+@lru_cache(maxsize=128)
 def get_widgets_by_page(pdf: bytes) -> Dict[int, List[dict]]:
     """
     Retrieves widgets from a PDF stream, organized by page number.
@@ -398,6 +419,7 @@ def remove_widgets_by_keys(
     if not keys:
         return pdf
 
+    key_set = set(keys)
     writer = PdfWriter(BytesIO(pdf))
 
     for page in writer.pages:
@@ -407,7 +429,7 @@ def remove_widgets_by_keys(
         for annot in page.get(Annots, []):
             annot = cast(DictionaryObject, annot.get_object())
             key = get_widget_key(annot.get_object(), use_full_widget_name)
-            if key not in keys:
+            if key not in key_set:
                 page_annots.append(annot)
             else:
                 needs_update = True
@@ -449,8 +471,7 @@ def update_widget_keys(
     out = PdfWriter()
     out.append(pdf)
 
-    for i, old_key in enumerate(old_keys):
-        _update_single_widget_key(out, widgets, old_key, new_keys[i], indices[i])
+    _apply_widget_key_updates(out, widgets, old_keys, new_keys, indices)
 
     with BytesIO() as f:
         out.write(f)
@@ -458,35 +479,44 @@ def update_widget_keys(
         return f.read()
 
 
-def _update_single_widget_key(
+def _apply_widget_key_updates(
     writer: PdfWriter,
     widgets: Dict[str, WIDGET_TYPES],
-    old_key: str,
-    new_key: str,
-    index: int,
+    old_keys: List[str],
+    new_keys: List[str],
+    indices: List[int],
 ) -> None:
     """
-    Updates a single widget key in a PDF template.
+    Applies queued widget key updates to matching annotations.
+
+    The update queue is converted into a lookup keyed by old widget name, then
+    each annotation is checked against that lookup as pages are traversed.
+    Non-radio widgets honor the requested occurrence index, while radio widgets
+    update every annotation in the radio group.
 
     Args:
         writer (PdfWriter): The PDF writer object.
         widgets (Dict[str, WIDGET_TYPES]): A dictionary of widgets in the template.
-        old_key (str): The old widget key to be replaced.
-        new_key (str): The new widget key to replace the old key.
-        index (int): The index of the widget to update if multiple widgets have the same name.
+        old_keys (List[str]): The old widget keys to replace.
+        new_keys (List[str]): The new widget keys to apply.
+        indices (List[int]): Widget occurrence indices for duplicate field names.
     """
-    tracker = -1
+    updates = {old_key: (new_keys[i], indices[i]) for i, old_key in enumerate(old_keys)}
+    trackers = {}
+
     for page in writer.pages:
         for annot in page.get(Annots, []):
             annot = cast(DictionaryObject, annot.get_object())
             key = get_widget_key(annot.get_object(), False)
 
-            widget = widgets.get(key)
-            if widget is None or old_key != key:
+            if key not in updates:
                 continue
 
-            tracker += 1
-            if not isinstance(widget, Radio) and tracker != index:
-                continue
+            widget = widgets.get(key)
+            if widget is not None:
+                trackers[key] = trackers.get(key, -1) + 1
+                new_key, index = updates[key]
+                if not isinstance(widget, Radio) and trackers[key] != index:
+                    continue
 
-            update_annotation_name(annot, new_key)
+                update_annotation_name(annot, new_key)

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/types.py

@@ -18,7 +18,7 @@ class PdfArray(list):
 
     When sliced, this list automatically merges the contained PdfWrapper
     objects using the PdfWrapper.__add__ method, returning a single
-    merged PdfWrapper object. If the slice is empty, it returns an empty list.
+    merged PdfWrapper object. If the slice is empty, it returns None.
     For non-slice indexing, it behaves like a standard list.
     """
 

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/utils.py

@@ -26,7 +26,7 @@ from pypdf.generic import ArrayObject, DictionaryObject, NameObject
 from .constants import SLASH, UNIQUE_SUFFIX_LENGTH, Annots
 
 
-@lru_cache
+@lru_cache(maxsize=128)
 def remove_all_widgets(pdf: bytes) -> bytes:
     """
     Removes all widgets (form fields) from a PDF, effectively flattening the form.

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/watermark.py

@@ -9,6 +9,7 @@ and to copy specific widgets from the watermarks to the original PDF.
 """
 
 from collections import defaultdict
+from functools import lru_cache
 from io import BytesIO
 from typing import Any, Dict, List, Optional
 
@@ -21,6 +22,20 @@ from .constants import Annots
 from .patterns import get_widget_key
 
 
+@lru_cache(maxsize=128)
+def _get_image_reader(image_stream: bytes) -> ImageReader:
+    """
+    Creates a cached ReportLab image reader for an image byte stream.
+
+    Args:
+        image_stream (bytes): The image data as a byte stream.
+
+    Returns:
+        ImageReader: The cached ReportLab image reader.
+    """
+    return ImageReader(BytesIO(image_stream))
+
+
 def draw_text(canvas: Canvas, **kwargs) -> None:
     """
     Draws a text string on the given canvas using the specified font, size, and color.
@@ -206,12 +221,8 @@ def draw_image(canvas: Canvas, **kwargs) -> None:
     width = kwargs["width"]
     height = kwargs["height"]
 
-    image_buff = BytesIO()
-    image_buff.write(image_stream)
-    image_buff.seek(0)
-
     canvas.drawImage(
-        ImageReader(image_buff),
+        _get_image_reader(image_stream),
         coordinate_x,
         coordinate_y,
         width=width,
@@ -219,8 +230,6 @@ def draw_image(canvas: Canvas, **kwargs) -> None:
         mask="auto",
     )
 
-    image_buff.close()
-
 
 def create_watermarks_and_draw(
     pdf: bytes, to_draw: List[dict], font_mapping: Optional[Dict[str, str]] = None
@@ -235,8 +244,9 @@ def create_watermarks_and_draw(
     Args:
         pdf (bytes): The original PDF file as a byte stream.
         to_draw (List[dict]): A list of drawing instructions, where each dictionary
-            must contain a "page_number" key (1-based) and a "type" key ("image", "text", or "line")
-            along with type-specific parameters.
+            must contain a "page_number" key (1-based) and a "type" key
+            ("image", "text", "line", "rect", "circle", or "ellipse") along
+            with type-specific parameters.
         font_mapping (Optional[Dict[str, str]]): A dictionary mapping original font names
             to temporary unique font names used by ReportLab.
 
@@ -253,20 +263,21 @@ def create_watermarks_and_draw(
         "ellipse": draw_ellipse,
     }
 
-    result = []
-
     page_to_to_draw = defaultdict(list)
     for each in to_draw:
         page_to_to_draw[each["page_number"]].append(each)
 
     pdf_file = PdfReader(BytesIO(pdf))
+    page_count = len(pdf_file.pages)
+    result = [b""] * page_count
+    font_mapping = font_mapping or {}
 
-    for i, page in enumerate(pdf_file.pages):
-        elements = page_to_to_draw[i + 1]
+    for page_num in range(1, page_count + 1):
+        elements = page_to_to_draw.get(page_num, [])
         if not elements:
-            result.append(b"")
             continue
 
+        page = pdf_file.pages[page_num - 1]
         buff = BytesIO()
 
         canvas = Canvas(
@@ -278,13 +289,11 @@ def create_watermarks_and_draw(
         )
 
         for element in elements:
-            type_to_func[element["type"]](
-                canvas, **element, font_mapping=font_mapping or {}
-            )
+            type_to_func[element["type"]](canvas, **element, font_mapping=font_mapping)
 
         canvas.save()
         buff.seek(0)
-        result.append(buff.read())
+        result[page_num - 1] = buff.read()
 
     return result
 
@@ -325,7 +334,7 @@ def merge_watermarks_with_pdf(
 def _clone_page_widgets(
     writer: PdfWriter,
     page: PageObject,
-    keys: Optional[List[str]],
+    keys: Optional[set[str]],
 ) -> List[Any]:
     """
     Clones matching widgets from a single PDF page.
@@ -333,7 +342,7 @@ def _clone_page_widgets(
     Args:
         writer (PdfWriter): The PdfWriter for cloning.
         page (PageObject): The source PDF page object.
-        keys (Optional[List[str]]): Keys of widgets to clone.
+        keys (Optional[set[str]]): Keys of widgets to clone.
 
     Returns:
         List[Any]: A list of cloned widget objects.
@@ -349,7 +358,7 @@ def _clone_page_widgets(
 def _collect_from_single_watermark_specific_page(
     writer: PdfWriter,
     watermark: bytes,
-    keys: Optional[List[str]],
+    keys: Optional[set[str]],
     page_num: int,
 ) -> Dict[int, List[Any]]:
     """
@@ -358,7 +367,7 @@ def _collect_from_single_watermark_specific_page(
     Args:
         writer (PdfWriter): The PdfWriter for cloning.
         watermark (bytes): The watermark PDF byte stream.
-        keys (Optional[List[str]]): Keys of widgets to clone.
+        keys (Optional[set[str]]): Keys of widgets to clone.
         page_num (int): The page index within the watermark PDF.
 
     Returns:
@@ -376,7 +385,7 @@ def _collect_from_single_watermark_specific_page(
 def _collect_from_single_watermark_1_to_1(
     writer: PdfWriter,
     watermark: bytes,
-    keys: Optional[List[str]],
+    keys: Optional[set[str]],
 ) -> Dict[int, List[Any]]:
     """
     Maps pages 1:1 between a single watermark PDF and the output PDF.
@@ -384,7 +393,7 @@ def _collect_from_single_watermark_1_to_1(
     Args:
         writer (PdfWriter): The PdfWriter for cloning.
         watermark (bytes): The watermark PDF byte stream.
-        keys (Optional[List[str]]): Keys of widgets to clone.
+        keys (Optional[set[str]]): Keys of widgets to clone.
 
     Returns:
         Dict[int, List[Any]]: A dictionary mapping output page indices to cloned widgets.
@@ -399,7 +408,7 @@ def _collect_from_single_watermark_1_to_1(
 def _collect_from_multiple_watermarks(
     writer: PdfWriter,
     watermarks: List[bytes],
-    keys: Optional[List[str]],
+    keys: Optional[set[str]],
     page_num: Optional[int],
 ) -> Dict[int, List[Any]]:
     """
@@ -408,7 +417,7 @@ def _collect_from_multiple_watermarks(
     Args:
         writer (PdfWriter): The PdfWriter for cloning.
         watermarks (List[bytes]): A list of watermark PDF byte streams.
-        keys (Optional[List[str]]): Keys of widgets to clone.
+        keys (Optional[set[str]]): Keys of widgets to clone.
         page_num (Optional[int]): The page index within each watermark PDF.
 
     Returns:
@@ -443,17 +452,19 @@ def _collect_widgets_to_copy(
     Returns:
         Dict[int, List[Any]]: A dictionary mapping output page indices to lists of cloned widgets.
     """
+    key_set = set(keys) if keys is not None else None
+
     if isinstance(watermarks, bytes):
         if page_num is not None:
             # Case: Single watermark PDF, extracting a specific page to the first output page.
             return _collect_from_single_watermark_specific_page(
-                writer, watermarks, keys, page_num
+                writer, watermarks, key_set, page_num
             )
         # Case: Single watermark PDF, mapping pages 1:1 to output pages.
-        return _collect_from_single_watermark_1_to_1(writer, watermarks, keys)
+        return _collect_from_single_watermark_1_to_1(writer, watermarks, key_set)
 
     # Case: List of watermark PDFs, each corresponding to an output page.
-    return _collect_from_multiple_watermarks(writer, watermarks, keys, page_num)
+    return _collect_from_multiple_watermarks(writer, watermarks, key_set, page_num)
 
 
 def _apply_widgets_to_pages(

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/widgets/base.py

@@ -15,6 +15,7 @@ Classes:
 from __future__ import annotations
 
 from dataclasses import dataclass
+from functools import lru_cache
 from inspect import signature
 from io import BytesIO
 from typing import List, Optional
@@ -102,6 +103,28 @@ class Widget:
             if each in kwargs:
                 self.hook_params.append((each, kwargs.get(each)))
 
+    @staticmethod
+    @lru_cache(maxsize=128)
+    def _get_default_field_flags(acro_form_class: type, acro_form_func: str) -> tuple:
+        """
+        Retrieves the default field flags for a ReportLab AcroForm method.
+
+        Args:
+            acro_form_class (type): The ReportLab AcroForm class.
+            acro_form_func (str): The AcroForm method name.
+
+        Returns:
+            tuple: The default field flags split into individual flag names.
+        """
+        default_flags = signature(
+            getattr(acro_form_class, acro_form_func)
+        ).parameters.get(fieldFlags)
+        return (
+            tuple(default_flags.default.split(" "))
+            if default_flags and default_flags.default
+            else ()
+        )
+
     def _required_handler(self, canvas: Canvas) -> None:
         """
         Handles the 'Required' flag for the widget's AcroForm field.
@@ -115,13 +138,8 @@ class Widget:
         Args:
             canvas (Canvas): The ReportLab canvas object used for PDF operations.
         """
-        default_flags = signature(
-            getattr(canvas.acroForm, self.ACRO_FORM_FUNC)
-        ).parameters.get(fieldFlags)
-        default_flags = (
-            default_flags.default.split(" ")
-            if default_flags and default_flags.default
-            else []
+        default_flags = list(
+            self._get_default_field_flags(type(canvas.acroForm), self.ACRO_FORM_FUNC)
         )
 
         if self.acro_form_params.get(required):
@@ -174,9 +192,9 @@ class Widget:
                          watermark for that page. Pages without any widgets will
                          have an empty byte string (b"").
         """
-        result = []
-
         pdf = PdfReader(BytesIO(stream))
+        page_count = len(pdf.pages)
+        result = [b""] * page_count
 
         widgets_by_page = {}
         for widget in widgets:
@@ -184,15 +202,14 @@ class Widget:
                 widgets_by_page[widget.page_number] = []
             widgets_by_page[widget.page_number].append(widget)
 
-        for i, page in enumerate(pdf.pages):
-            page_num = i + 1
+        for page_num in range(1, page_count + 1):
             if page_num not in widgets_by_page:
-                result.append(b"")
                 continue
 
             # Use a fresh buffer per page to avoid stale trailing bytes
             # when the current page watermark is smaller than a previous page.
             watermark = BytesIO()
+            page = pdf.pages[page_num - 1]
 
             canvas = Canvas(
                 watermark,
@@ -209,7 +226,7 @@ class Widget:
             canvas.showPage()
             canvas.save()
             watermark.seek(0)
-            result.append(watermark.read())
+            result[page_num - 1] = watermark.read()
 
         return result
 

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/widgets/signature.py

@@ -104,13 +104,13 @@ class SignatureWidget:
             List[bytes]: A list of watermark PDF streams. Each element corresponds to
             a page in the input PDF.
         """
-        result = []
-
         page_to_widgets = defaultdict(list)
         for widget in widgets:
             page_to_widgets[widget.page_number].append(widget)
 
         input_pdf = PdfReader(BytesIO(stream))
+        page_count = len(input_pdf.pages)
+        result = [b""] * page_count
 
         bedrock = PdfReader(BytesIO(BEDROCK_PDF))
         page = bedrock.pages[0]
@@ -119,9 +119,14 @@ class SignatureWidget:
             key = get_widget_key(annot.get_object(), False)
             annot_type_to_annot[key] = annot.get_object()
 
-        for i, p in enumerate(input_pdf.pages):
+        for page_num in range(1, page_count + 1):
+            page_widgets = page_to_widgets.get(page_num, [])
+            if not page_widgets:
+                continue
+
             # pylint: disable=R0801
             watermark = BytesIO()
+            p = input_pdf.pages[page_num - 1]
             canvas = Canvas(
                 watermark,
                 pagesize=(
@@ -135,8 +140,6 @@ class SignatureWidget:
 
             out = PdfWriter(watermark)
 
-            page_widgets = page_to_widgets.get(i + 1, [])
-
             widgets_to_copy = []
             for widget in page_widgets:
                 widget_to_copy = annot_type_to_annot[
@@ -166,7 +169,7 @@ class SignatureWidget:
             with BytesIO() as f:
                 out.write(f)
                 f.seek(0)
-                result.append(f.read())
+                result[page_num - 1] = f.read()
 
         return result
 

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm/lib/wrapper.py

@@ -20,7 +20,7 @@ from __future__ import annotations
 
 from collections import defaultdict
 from dataclasses import asdict
-from functools import cached_property
+from functools import lru_cache
 from os import PathLike
 from typing import (
     TYPE_CHECKING,
@@ -131,7 +131,7 @@ class PdfWrapper:
                 - str: The file path to the PDF.
                 - BinaryIO: An open file-like object containing the PDF data.
                 - BlankPage: A blank page object.
-                Defaults to an empty byte string (b""), which creates a blank PDF.
+                Defaults to an empty byte string (b"").
             **kwargs: Additional keyword arguments to configure the `PdfWrapper`.
                 These arguments are used to set the user-configurable parameters defined in `USER_PARAMS`.
                 For example: `use_full_widget_name=True` or `need_appearances=False`.
@@ -146,6 +146,7 @@ class PdfWrapper:
         )
         self._on_open_javascript = None
         self._available_fonts = {}  # for setting /F1
+        self._available_fonts_loaded = False
         self._font_register_events = []  # for reregister
         self._key_update_tracker = {}  # for update key preserve old key attrs
         self._keys_to_update = []  # for bulk update keys
@@ -208,19 +209,21 @@ class PdfWrapper:
 
     def _init_helper(self) -> None:
         """
-        Helper method to initialize widgets and available fonts.
+        Helper method to initialize widgets.
 
         This method is called during initialization and after certain operations
         that modify the PDF content (e.g., filling, creating widgets, updating keys).
-        It rebuilds the widget dictionary and updates the available fonts.
+        It rebuilds the widget dictionary and invalidates the lazily loaded font cache.
         """
 
+        stream = self._read()
+        self._available_fonts_loaded = False
         new_widgets = (
             build_widgets(
-                self._read(),
+                stream,
                 getattr(self, "use_full_widget_name"),
             )
-            if self._read()
+            if stream
             else {}
         )
         # ensure old widgets don't get overwritten
@@ -238,8 +241,44 @@ class PdfWrapper:
 
         self.widgets = new_widgets
 
-        if self._read():
-            self._available_fonts.update(**get_all_available_fonts(self._read()))
+    def _ensure_available_fonts_loaded(self) -> dict:
+        """
+        Loads AcroForm fonts from the PDF stream the first time they are needed.
+
+        Custom fonts registered through `register_font` are stored in the same
+        mapping, so loading updates the existing dictionary instead of replacing it.
+
+        Returns:
+            dict: A mapping from font names to internal PDF font identifiers.
+        """
+
+        if not self._available_fonts_loaded:
+            if self._stream:
+                self._available_fonts.update(**get_all_available_fonts(self._stream))
+            self._available_fonts_loaded = True
+
+        return self._available_fonts
+
+    @staticmethod
+    @lru_cache(maxsize=128)
+    def _get_page_streams_with_widgets(stream: bytes) -> tuple[bytes, ...]:
+        """
+        Extracts page streams while preserving the original page widgets.
+
+        Args:
+            stream (bytes): The PDF stream to split into pages.
+
+        Returns:
+            tuple[bytes, ...]: Single-page PDF streams with widgets preserved.
+        """
+
+        return tuple(
+            # Case: Single watermark PDF, extracting a specific page to the first output page.
+            copy_watermark_widgets(page_stream, stream, None, i)
+            for i, page_stream in enumerate(
+                get_page_streams(remove_all_widgets(stream))
+            )
+        )
 
     def _reregister_font(self) -> PdfWrapper:
         """
@@ -328,13 +367,16 @@ class PdfWrapper:
         """
         Returns a list of the names of the currently registered fonts.
 
+        Accessing this property loads AcroForm fonts from the PDF stream if
+        they have not already been loaded.
+
         Returns:
             list: A list of font names (str).
         """
 
-        return list(self._available_fonts.keys())
+        return list(self._ensure_available_fonts_loaded().keys())
 
-    @cached_property
+    @property
     def pages(self) -> Sequence[PdfWrapper]:
         """
         Returns a list of `PdfWrapper` objects, each representing a single page in the PDF document.
@@ -347,11 +389,10 @@ class PdfWrapper:
 
         result = [
             self.__class__(
-                # Case: Single watermark PDF, extracting a specific page to the first output page.
-                copy_watermark_widgets(each, self._read(), None, i),
+                each,
                 **{param: getattr(self, param) for param, _ in self.USER_PARAMS},
             )
-            for i, each in enumerate(get_page_streams(remove_all_widgets(self._read())))
+            for each in self._get_page_streams_with_widgets(self._read())
         ]
 
         # because copy_watermark_widgets and remove_all_widgets
@@ -433,15 +474,15 @@ class PdfWrapper:
         """
 
         if any(widget.hooks_to_trigger for widget in self.widgets.values()):
+            available_fonts = self._ensure_available_fonts_loaded()
             for widget in self.widgets.values():
                 if (
                     isinstance(widget, (Text, Dropdown))
-                    and widget.font not in self._available_fonts.values()
-                    and widget.font in self._available_fonts
+                    and widget.font not in available_fonts.values()
+                    and widget.font in available_fonts
                 ):
-                    widget.font = self._available_fonts.get(
-                        widget.font
-                    )  # from `new_font` to `/F1`
+                    # from `new_font` to `/F1`
+                    widget.font = available_fonts.get(widget.font)
 
             self._stream = trigger_widget_hooks(
                 self._stream,
@@ -776,7 +817,7 @@ class PdfWrapper:
         Draws raw elements (text, images, etc.) directly onto the PDF pages.
 
         This method is the primary mechanism for drawing non-form field content.
-        It takes a list of `RawText` or `RawImage` objects and renders them
+        It takes a list of raw element objects and renders them
         onto the PDF document as watermarks.
 
         Args:
@@ -824,6 +865,7 @@ class PdfWrapper:
         ttf_file = fp_or_f_obj_or_stream_to_stream(ttf_file)
 
         if validate_font(font_name, ttf_file) if ttf_file is not None else False:
+            self._ensure_available_fonts_loaded()
             self._stream, new_font_name = register_font_acroform(
                 self._read(), ttf_file, getattr(self, "need_appearances")
             )

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PyPDFForm
-Version: 5.2.0
+Version: 5.2.2
 Summary: The Python library & CLI for PDF forms.
 Author: Jinge Li
 License-Expression: MIT
@@ -24,12 +24,12 @@ Classifier: Topic :: Utilities
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: cryptography<49.0.0,>=48.0.0
+Requires-Dist: cryptography<50.0.0,>=48.0.0
 Requires-Dist: fonttools<5.0.0,>=4.63.0
 Requires-Dist: pikepdf<11.0.0,>=10.7.2
 Requires-Dist: pillow<13.0.0,>=12.2.0
 Requires-Dist: pypdf<7.0.0,>=6.12.2
-Requires-Dist: reportlab<5.0.0,>=4.5.1
+Requires-Dist: reportlab<6.0.0,>=4.5.1
 Provides-Extra: cli
 Requires-Dist: typer<1.0.0,>=0.26.1; extra == "cli"
 Requires-Dist: jsonschema<5.0.0,>=4.26.0; extra == "cli"
@@ -125,4 +125,10 @@ The official documentation can be found on [the GitHub page](https://chinapandam
 
 This project is maintained entirely in my spare time. If you like the project please consider starring the GitHub repository. It is the best way to keep me motivated and continue making the project better.
 
-[![Stargazers over time](https://starchart.cc/chinapandaman/PyPDFForm.svg?variant=adaptive)](https://starchart.cc/chinapandaman/PyPDFForm)
+<a href="https://www.star-history.com/?repos=chinapandaman%2FPyPDFForm&type=date&logscale=&legend=top-left">
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&theme=dark&legend=top-left" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&legend=top-left" />
+   <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&legend=top-left" />
+ </picture>
+</a>

{pypdfform-5.2.0 → pypdfform-5.2.2}/PyPDFForm.egg-info/requires.txt

@@ -1,9 +1,9 @@
-cryptography<49.0.0,>=48.0.0
+cryptography<50.0.0,>=48.0.0
 fonttools<5.0.0,>=4.63.0
 pikepdf<11.0.0,>=10.7.2
 pillow<13.0.0,>=12.2.0
 pypdf<7.0.0,>=6.12.2
-reportlab<5.0.0,>=4.5.1
+reportlab<6.0.0,>=4.5.1
 
 [cli]
 typer<1.0.0,>=0.26.1

{pypdfform-5.2.0 → pypdfform-5.2.2}/README.md

@@ -77,4 +77,10 @@ The official documentation can be found on [the GitHub page](https://chinapandam
 
 This project is maintained entirely in my spare time. If you like the project please consider starring the GitHub repository. It is the best way to keep me motivated and continue making the project better.
 
-[![Stargazers over time](https://starchart.cc/chinapandaman/PyPDFForm.svg?variant=adaptive)](https://starchart.cc/chinapandaman/PyPDFForm)
+<a href="https://www.star-history.com/?repos=chinapandaman%2FPyPDFForm&type=date&logscale=&legend=top-left">
+ <picture>
+   <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&theme=dark&legend=top-left" />
+   <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&legend=top-left" />
+   <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=chinapandaman/PyPDFForm&type=date&legend=top-left" />
+ </picture>
+</a>

{pypdfform-5.2.0 → pypdfform-5.2.2}/pyproject.toml

@@ -29,12 +29,12 @@ classifiers = [
 ]
 requires-python = ">=3.10"
 dependencies = [
-    "cryptography>=48.0.0,<49.0.0",
+    "cryptography>=48.0.0,<50.0.0",
     "fonttools>=4.63.0,<5.0.0",
     "pikepdf>=10.7.2,<11.0.0",
     "pillow>=12.2.0,<13.0.0",
     "pypdf>=6.12.2,<7.0.0",
-    "reportlab>=4.5.1,<5.0.0",
+    "reportlab>=4.5.1,<6.0.0",
 ]
 
 [project.urls]