OpenSystemsDevelopment / qpdf

…update `CMakeLists.txt` accordingly.

Fix bug in #1595

- replace repetitive error handling with utility functions in BufferInputSource, OffsetBuffer, and QPDF_objects.

- Remove some "internal error" labels from errors that could be due to user logic errors.

- Updated warning messages to provide clearer descriptions of common issues ("object has offset 0", "xref entry missing") and note that these are handled gracefully by qpdf.
- Adjusted logic in `QPDF_objects.cc` to identify and process `xref` stream-specific anomalies accurately.

Fixes #1594
Fixes #1292

…sting return values for uninitialized objects, and cleaning up error handling logic.

…oved shared pointer handling, remove deprecated object methods, and update all references.

…sages, and clean up `QPDF.hh`.

… and `findEndstream` to `Objects`, remove `QPDF::PatternFinder` class, and update related logic for improved encapsulation.

…, replace direct calls with `QIntC` equivalents, and clean up `QPDF.hh`.

…:Doc`, rename related methods, and streamline usage across QPDFWriter and linearization modules.

…sed_after_compressed` in `QPDF::Doc::Objects`, update usage in `QPDF_linearization`, and streamline related logic.

… `attempt_recovery` to `password_is_hex_key` and `suppress_recovery` to align with the CLI, update usage across the codebase, and streamline related logic.

…fields, replace direct member access across the codebase, and streamline related logic.

…te usage throughout the codebase, and streamline configurations.

… streamline error handling, and improve encapsulation.

Relocate `Objects` to `QPDF::Doc` for improved encapsulation of object-related logic. Adjust all relevant methods and references to use the new placement.

…deliberately broken files more efficiently.

…`, and simplify member access.

…lize type conversions, and improve clarity across multiple modules.

…eams, improving robustness against damaged PDFs.

…string`, improving performance, simplifying stream data handling, and reducing dependency on `BufferInputSource`.

…std::string_view`, improving performance and code clarity.

…cess for consistency and readability.

Simplify `readObjectAtOffset` by splitting responsibilities into separate functions: `read_object_start` and a streamlined `readObjectAtOffset`. Remove redundant parameters and improve code readability, ensuring better modularity for object reading operations.

After xref reconstruction treat the input file as suspect and apply sanity checks to all subsequent object reads.

Remove `in_xref_reconstruction` flag and update references to use `reconstructed_xref` for simplified state management during xref processing. Adjust warnings for invalid dictionary keys in test output.

the xref table has been loaded

Detect corrupt xref streams early to avoid excessive runtime and memory usage.

Backup and restore the xref table during attempted startxref recovery to avoid corrupt states. Added a guard condition to handle invalid xref entries types, preventing processing of unsupported cases. These changes improve robustness during PDF parsing and recovery.

Replaced raw pointer defaults (e.g., `0`) with `nullptr` for better readability and type-safety. Simplified object initializer syntax using uniform initialization `{}`. These changes enhance code maintainability and ensure modern C++ standards compliance.

Simplified checks for empty containers and strings across the codebase using the `.empty()` method. This improves code readability and adheres to best practices for checking emptiness.

Converted multiple occurrences of `count()` to `contains()` throughout the codebase where the goal was to check key existence in containers. This improves code readability and aligns with modern C++ practices, particularly with C++20, making the intent more explicit and potentially aiding performance.

During xref table reconstruction ignore uncompressed object entries found
in xref streams. The xref table gets populated with entries for the
objects actually found in the file. The entries for uncompressed object in
xref streams are redundant and potentially incorrect.

Reject objects containing arrays or dictionaries with more than 5000
elements. We are by definition dealing with damaged files, and such
objects are extremely likely to be invalid or malicious.

As discussed in #1396.