python-pdffiller 2.0.3__py3-none-any.whl → 2.0.4__py3-none-any.whl

pdffiller/_version.py

@@ -3,4 +3,4 @@ __copyright__ = "Copyright 2025 SISMIC"
 __email__ = "jraphanel@sismic.fr"
 __license__ = "MIT"
 __title__ = "pdffiller"
-__version__ = "2.0.3"
+__version__ = "2.0.4"

pdffiller/cli/commands/sanitize.py

@@ -0,0 +1,100 @@
+import json
+import os
+
+from pdffiller.cli.args import add_global_arguments
+from pdffiller.cli.command import pdffiller_command, PdfFillerArgumentParser
+from pdffiller.cli.once_argument import OnceArgument
+from pdffiller.exceptions import (
+    AbortExecution,
+    CommandLineError,
+    FileNotExistsError,
+    PdfFillerException,
+)
+from pdffiller.io.output import cli_out_write, PdfFillerOutput
+from pdffiller.pdf import Pdf
+from pdffiller.typing import Any, Dict, List
+
+from ..exit_codes import ERROR_ENCOUNTERED
+
+
+def sanitize_text_formatter(sanitized: List[Dict[str, Any]]) -> None:
+    """Print sanitized fields as plain text"""
+    if not sanitized:
+        cli_out_write("No fields were sanitized.")
+        return
+    for field in sanitized:
+        cli_out_write("----------")
+        for key, value in field.items():
+            cli_out_write(f"{key}: {value}")
+
+
+def sanitize_json_formatter(sanitized: List[Dict[str, Any]]) -> None:
+    """Print sanitized fields as JSON"""
+    if not sanitized:
+        return
+    cli_out_write(json.dumps(sanitized, indent=4, ensure_ascii=False))
+
+
+@pdffiller_command(
+    group=None,
+    formatters={"text": sanitize_text_formatter, "json": sanitize_json_formatter},
+)
+def sanitize(parser: PdfFillerArgumentParser, *args: Any) -> Any:
+    """
+    Sanitize a PDF by removing MaxLen constraints from text fields.
+    This fixes PDFs where MaxLen is incorrectly set, causing text truncation.
+    """
+    options_group = parser.add_argument_group("options")
+
+    options_group.add_argument(
+        "-o",
+        "--output",
+        metavar="OUTPUT_PATH",
+        type=str,
+        help="""Path to the output PDF file.""",
+        action=OnceArgument,
+    )
+
+    options_group.add_argument(
+        "--deep",
+        action="store_true",
+        default=False,
+        help="Perform a deep sanitization: also clear the Comb flag from text fields.",
+    )
+
+    parser.add_argument(
+        "file",
+        metavar="INPUT_PATH",
+        type=str,
+        nargs="?",
+        help="""Path to the input PDF file.""",
+        action=OnceArgument,
+    )
+
+    add_global_arguments(options_group, True, parser)
+
+    opts = parser.parse_args(*args)
+
+    output = PdfFillerOutput()
+    if not opts.file:
+        raise CommandLineError("no input file given")
+
+    if not opts.output:
+        raise CommandLineError("no output file path given")
+
+    if not os.path.isfile(opts.file):
+        raise FileNotExistsError(opts.file)
+
+    try:
+        pdf = Pdf()
+        sanitized = pdf.sanitize(opts.file, opts.output, opts.deep)
+        output.info(f"sanitized pdf saved to {opts.output}")
+        return sanitized
+    except PdfFillerException as exp:
+        output.error(str(exp))
+    except Exception as exg:  # pylint: disable=broad-except # pragma: no cover
+        output.error(f"unexpected error when sanitizing {opts.file} with the following error:")
+        output.error(exg)
+        raise AbortExecution(ERROR_ENCOUNTERED) from exg
+
+    return None

pdffiller/pdf.py

@@ -88,7 +88,7 @@ class Pdf:
             ) from ex
 
         for i, page in enumerate(doc.pages()):
-            output.verbose(f"loading page {i+1}/{doc.page_count}")
+            output.verbose(f"loading page {i + 1}/{doc.page_count}")
             for widget in page.widgets():
                 button_states = widget.button_states()
                 choices = button_states["normal"] if button_states else None
@@ -105,7 +105,7 @@ class Pdf:
                             choices.insert(0, "Off")
                         new_widget.choices = [choice.replace("#20", " ") for choice in choices]
                     elif isinstance(new_widget, TextWidget):
-                        new_widget.max_length = widget.text_maxlen
+                        new_widget.max_length = self._resolve_text_maxlen(doc, widget)
                     loaded_widgets[widget.field_name] = new_widget
                 else:
                     new_widget = loaded_widgets[widget.field_name]
@@ -128,6 +128,186 @@ class Pdf:
 
         self.widgets = loaded_widgets
 
+    @staticmethod
+    def _resolve_text_maxlen(doc: pymupdf.Document, widget: Any) -> Optional[int]:
+        """
+        Resolve the MaxLen value for a text widget by walking up the PDF object hierarchy.
+
+        PyMuPDF's widget.text_maxlen may not reflect the actual MaxLen when it is
+        defined on a parent field dictionary rather than on the widget annotation itself.
+        This method checks the widget xref and its parents for /MaxLen.
+
+        Args:
+            doc: The pymupdf Document.
+            widget: The pymupdf Widget.
+
+        Returns:
+            The resolved max length, or None if not found.
+        """
+        max_length: Optional[int] = widget.text_maxlen
+        xref = widget.xref
+        while xref > 0:
+            key_type, value = doc.xref_get_key(xref, "MaxLen")
+            if key_type != "null" and value:
+                try:
+                    int_value = int(value)
+                    if max_length is None or max_length < int_value:
+                        max_length = int_value
+                    # return int(value)
+                except (ValueError, TypeError):
+                    pass
+            # Walk up to parent
+            key_type, value = doc.xref_get_key(xref, "Parent")
+            if key_type == "xref":
+                parent_xref = int(value.split()[0])
+                if parent_xref == xref:
+                    break
+                xref = parent_xref
+            else:
+                break
+
+        return max_length
+
+    COMB_FLAG = 1 << 24  # Bit 25 in PDF spec (Comb option)
+
+    @staticmethod
+    def _has_comb_flag(doc: pymupdf.Document, widget: Any) -> bool:
+        """
+        Check if the Comb flag is set on the widget or any of its parents.
+
+        Args:
+            doc: The pymupdf Document.
+            widget: The pymupdf Widget.
+
+        Returns:
+            True if the Comb flag is set on any node in the hierarchy.
+        """
+        xref = widget.xref
+        while xref > 0:
+            key_type, ff_value = doc.xref_get_key(xref, "Ff")
+            if key_type != "null" and ff_value:
+                try:
+                    if int(ff_value) & Pdf.COMB_FLAG:
+                        return True
+                except (ValueError, TypeError):
+                    pass
+            key_type, value = doc.xref_get_key(xref, "Parent")
+            if key_type == "xref":
+                parent_xref = int(value.split()[0])
+                if parent_xref == xref:
+                    break
+                xref = parent_xref
+            else:
+                break
+        return False
+
+    @staticmethod
+    def _remove_maxlen(doc: pymupdf.Document, widget: Any, deep: bool = False) -> None:
+        """
+        Remove /MaxLen from the widget and all its parent field dictionaries.
+
+        In deep mode, also clears the Comb flag (bit 25 of /Ff).
+        In normal mode, /MaxLen is only removed on nodes where Comb is not set.
+
+        Args:
+            doc: The pymupdf Document.
+            widget: The pymupdf Widget.
+            deep: If True, also clear the Comb flag from /Ff.
+        """
+        xref = widget.xref
+        while xref > 0:
+            # Check Comb flag on this node
+            has_comb = False
+            key_type, ff_value = doc.xref_get_key(xref, "Ff")
+            if key_type != "null" and ff_value:
+                try:
+                    flags = int(ff_value)
+                    has_comb = bool(flags & Pdf.COMB_FLAG)
+                except (ValueError, TypeError):
+                    pass
+
+            if deep:
+                # Deep mode: remove /MaxLen and clear Comb flag
+                if doc.xref_get_key(xref, "MaxLen")[0] != "null":
+                    doc.xref_set_key(xref, "MaxLen", "null")
+                if has_comb:
+                    doc.xref_set_key(xref, "Ff", str(flags & ~Pdf.COMB_FLAG))
+            else:
+                # Normal mode: only remove /MaxLen if Comb is not set on this node
+                if not has_comb and doc.xref_get_key(xref, "MaxLen")[0] != "null":
+                    doc.xref_set_key(xref, "MaxLen", "null")
+
+            # Walk up to parent
+            key_type, value = doc.xref_get_key(xref, "Parent")
+            if key_type == "xref":
+                parent_xref = int(value.split()[0])
+                if parent_xref == xref:
+                    break
+                xref = parent_xref
+            else:
+                break
+        widget.text_maxlen = 0
+
+    def sanitize(
+        self,
+        input_file: PathLike,
+        output_file: PathLike,
+        deep: bool = False,
+    ) -> List[Dict[str, Any]]:
+        """
+        Sanitize a PDF by removing /MaxLen constraints from all text fields.
+
+        This fixes PDFs where /MaxLen is set incorrectly (e.g. 4 instead of 27),
+        which causes text truncation when filling the form.
+
+        Args:
+            input_file (PathLike): The input file path.
+            output_file (PathLike): The output file path.
+            deep: If True, also clear the Comb flag from /Ff.
+
+        Returns:
+            A list of dicts describing each sanitized field:
+            [{"FieldName": ..., "FieldType": "text", "MaxLen": <old_value>}, ...]
+        """
+        try:
+            document = pymupdf.open(filename=input_file)
+        except Exception as ex:
+            PdfFillerOutput().error(str(ex))
+            raise PdfFillerException(f"failed to open {input_file}") from ex
+
+        output = PdfFillerOutput()
+        output.info("sanitizing pdf text fields")
+
+        sanitized: List[Dict[str, Any]] = []
+        for page in document:
+            for field in page.widgets():
+                if (
+                    field.field_type == pymupdf.PDF_WIDGET_TYPE_TEXT  # pylint: disable=no-member
+                    and field.text_maxlen
+                ):
+                    # Skip fields with Comb flag in normal mode
+                    if not deep and self._has_comb_flag(document, field):
+                        output.verbose(f"skipping {field.field_name} (Comb flag active)")
+                        continue
+                    old_maxlen = field.text_maxlen
+                    output.verbose(f"removing MaxLen={old_maxlen} from {field.field_name}")
+                    self._remove_maxlen(document, field, deep)
+                    field.update()
+                    sanitized.append(
+                        {
+                            "FieldName": field.field_name,
+                            "FieldType": field.field_type_string,
+                            "MaxLen": old_maxlen,
+                        }
+                    )
+
+        try:
+            document.save(output_file)
+        except Exception:  # pylint: disable=broad-exception-caught
+            output.warning("an error occurs when saving file")
+
+        return sanitized
+
     @property
     def schema(self) -> List[Dict[str, Any]]:
         """
@@ -217,11 +397,12 @@ class Pdf:
 
                     # Handling other fields types
                     else:
+                        # Remove MaxLen constraint from the PDF object to avoid truncation
+                        self._remove_maxlen(document, field)
                         output.verbose(
                             f"updating {field.field_name} with {value} from {field.field_value}"
                         )
                         field.field_value = value
-
                     # Update the widget!
                     field.update()
         try:

{python_pdffiller-2.0.3.dist-info → python_pdffiller-2.0.4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: python-pdffiller
-Version: 2.0.3
+Version: 2.0.4
 Summary: Interact with PDF by inspecting or filling it
 Author-email: Jacques Raphanel <jraphanel@sismic.fr>
 License-Expression: MIT

{python_pdffiller-2.0.3.dist-info → python_pdffiller-2.0.4.dist-info}/RECORD

@@ -1,9 +1,9 @@
 pdffiller/__init__.py,sha256=0HtgXhEV1fKTKAcOXGcq4UsCFflDIPCDQvckshZf-1k,195
 pdffiller/__main__.py,sha256=7NPQgZVx6VSZS7OrmyJQ_O1vL4wiSqhiILi-outwUqM,107
-pdffiller/_version.py,sha256=pFoKdvINjWe5nnOYr0ui5BcN-wdF-rn7wCYxgGvVkBQ,172
+pdffiller/_version.py,sha256=4kVX4SuWyAER9NxjJsz0eyvUdNcPcxAXfDBhA6Min2Q,172
 pdffiller/const.py,sha256=if_j5I8ftczpjrzZjA7idv-XpvIj1-XBF4oe6VtQvF0,434
 pdffiller/exceptions.py,sha256=CdN93bZ0mBBS5vLxg14FYZUy4xkYqoD3_SzqtSkZr4g,1624
-pdffiller/pdf.py,sha256=A9-GSnQHUNR4i0men27zKeM2VzMpwds9QHt4Kxnmw8A,9254
+pdffiller/pdf.py,sha256=Y-GWZC7jomgqdPS5mIhfEx12xtPKlEsaWptDaAVq2mQ,16217
 pdffiller/py.typed.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 pdffiller/typing.py,sha256=4GJGevlU-YOR9fmuPT8jmyg5MBhn-2TB1K88b6C8VRw,937
 pdffiller/utils.py,sha256=pmGf3QwkhKwosk_eFCauzHM-XHp_WGVQAtZlxa7taYY,827
@@ -19,6 +19,7 @@ pdffiller/cli/smart_formatter.py,sha256=59hwF07nKbp-P9IfbqKgMFsfbvjIw5SACCZpUF4n
 pdffiller/cli/commands/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 pdffiller/cli/commands/dump_data_fields.py,sha256=o7LmFBdl9GfJOyZM1tOgeesx004QWIIXVp2_pX6FB1Q,2292
 pdffiller/cli/commands/fill_form.py,sha256=5wjbjwYLsytY6ea-n_KDbU9UDmkN00NYj_gm25MQqJ8,4758
+pdffiller/cli/commands/sanitize.py,sha256=wfcsgaTVVQKfnlHzJoRrYYYKczYjamk2uDbCUb8j-4U,2986
 pdffiller/io/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 pdffiller/io/colors.py,sha256=QCBEWksTVNurOJQYO0zh1X_xxIOXxXmYSJhbCqnNjI8,1710
 pdffiller/io/output.py,sha256=QMASWRWmfZGG9DdtlfpWXM3VJAMgWGQwPUzoYp_9FFY,11298
@@ -27,10 +28,10 @@ pdffiller/widgets/base.py,sha256=omGVQsQgMa-ALESnUd3_94oVIYScAMl0SPhHC_DG8Lg,361
 pdffiller/widgets/checkbox.py,sha256=iijStLAsY1G4cljW3a9NxVS_8qxJewFEw-B8jU2aKXk,1711
 pdffiller/widgets/radio.py,sha256=Db9Oc3Q8ge8qqTVPLoz3I1_SJBGyJ8KfA33ixZMr78c,1070
 pdffiller/widgets/text.py,sha256=SiuyBvZPZ6idCmtZ_05zE26iN6Rz67OfOj1fUm98YQI,2397
-python_pdffiller-2.0.3.dist-info/licenses/AUTHORS.rst,sha256=1_hVzMKgmoXvGgrcZC7sIbU_6PvvkB6vwqevAqzrIkQ,205
-python_pdffiller-2.0.3.dist-info/licenses/COPYING,sha256=ADPe-bH2wYq8nFf6EPJyovzTJyl3jSPnm09mGI8FSTo,1074
-python_pdffiller-2.0.3.dist-info/METADATA,sha256=DDZg0ASQb7K4D-5W2POxLQJ-3eaCtheEIIt1kxTN4Tw,4175
-python_pdffiller-2.0.3.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
-python_pdffiller-2.0.3.dist-info/entry_points.txt,sha256=RESKKpPPdWl0wDET96ntuFoUydALx9j0mxtbt-MEBjU,49
-python_pdffiller-2.0.3.dist-info/top_level.txt,sha256=5MGWCBFYlu_Ax-I5PgQkV9Xw7O48maPe9z8Qj_yVPL4,10
-python_pdffiller-2.0.3.dist-info/RECORD,,
+python_pdffiller-2.0.4.dist-info/licenses/AUTHORS.rst,sha256=1_hVzMKgmoXvGgrcZC7sIbU_6PvvkB6vwqevAqzrIkQ,205
+python_pdffiller-2.0.4.dist-info/licenses/COPYING,sha256=ADPe-bH2wYq8nFf6EPJyovzTJyl3jSPnm09mGI8FSTo,1074
+python_pdffiller-2.0.4.dist-info/METADATA,sha256=4on4QrrWJGfDsaakE1InU3ptSd2kQNECI8KTNeLyizM,4175
+python_pdffiller-2.0.4.dist-info/WHEEL,sha256=aeYiig01lYGDzBgS8HxWXOg3uV61G9ijOsup-k9o1sk,91
+python_pdffiller-2.0.4.dist-info/entry_points.txt,sha256=RESKKpPPdWl0wDET96ntuFoUydALx9j0mxtbt-MEBjU,49
+python_pdffiller-2.0.4.dist-info/top_level.txt,sha256=5MGWCBFYlu_Ax-I5PgQkV9Xw7O48maPe9z8Qj_yVPL4,10
+python_pdffiller-2.0.4.dist-info/RECORD,,