The PRC file header parsing logic trusts the constructed file structure description information, assumes that the underlying array contains elements and reads them, leading to out-of-bounds reads and application crashes.

During the PRC parsing stage, there is a lack of boundary verification for the PRC entity index, which leads to an out-of-bounds read of the entity array. As a result, the application crashes.

When the application opens a PDF and JavaScript modifies the properties of form fields, it causes the state of the underlying objects referenced by the program to become invalid. Eventually, it reads an illegal memory address, which leads to the crash of the application.

After JavaScript resetting the form, the synchronization process lacks re-entry protection and object lifecycle verification, resulting in the failure of the control pointer during the traversal process. After the pointer fails, it still continues to dereference, causing the application to crash.

When the application opens a PDF and JavaScript resets the form fields, the script re-enters the interface. The underlying native object is damaged, but the application does not perform validation. The function call on the damaged object leads to the application crashing.

The application re-enters the document structure via field processing and deletes the current page, and then continues using the field objects obtained before deletion, triggering an illegal read and crashing.

The application opened a PDF file containing an abnormal Unity 3D object. During parsing, the application incorrectly resolved a portion of the abnormal object as a pointer and used it as a valid address, ultimately causing the application to crash.

The application's list box calculate array logic keeps stale references to page or form objects after they are deleted or re-created, which allows crafted documents to trigger a use-after-free when the calculation runs and can potentially lead to arbitrary code execution.

The application does not validate the presence of required appearance (AP) data before accessing stamp annotation resources. When a PDF contains a stamp annotation missing its AP entry, the code continues to dereference the associated object without a prior null or validity check, which allows a crafted document to trigger a null pointer dereference and crash the application, resulting in denial of service.

The application does not properly validate the lifetime and validity of internal view cache pointers after JavaScript changes the document zoom and page state. When a script modifies the zoom property and then triggers a page change, the original view object may be destroyed while stale pointers are still kept and later dereferenced, which under crafted JavaScript and document structures can lead to a use-after-free condition and potentially allow arbitrary code execution.

The application does not detect or guard against cyclic PDF object references while handling JavaScript in PDF. When pages and annotations are crafted that reference each other in a loop, passing the document to APIs (e.g., SOAP) that perform deep traversal can cause uncontrolled recursion, stack exhaustion, and application crashes.