Add FPDF_StructElement_GetObjType
Get the object type (/Type) for a given element. Acrobat does not
enforce /Type field to be 'StructElement' string. It renders any value
in /Type. Therefore adding this allows us to see and validate that.
Change-Id: I0808b0c072cbe561e0aa2a5d1a9128a0bcd41422
Reviewed-on: https://pdfium-review.googlesource.com/c/pdfium/+/90570
Reviewed-by: Lei Zhang <thestig@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
diff --git a/core/fpdfdoc/cpdf_structelement.cpp b/core/fpdfdoc/cpdf_structelement.cpp
index 56cd410..dda016b 100644
--- a/core/fpdfdoc/cpdf_structelement.cpp
+++ b/core/fpdfdoc/cpdf_structelement.cpp
@@ -49,6 +49,10 @@
CPDF_StructElement::~CPDF_StructElement() = default;
+ByteString CPDF_StructElement::GetObjType() const {
+ return GetDict()->GetStringFor("Type");
+}
+
WideString CPDF_StructElement::GetAltText() const {
return GetDict()->GetUnicodeTextFor("Alt");
}
diff --git a/core/fpdfdoc/cpdf_structelement.h b/core/fpdfdoc/cpdf_structelement.h
index 5aac005..efeb2dc 100644
--- a/core/fpdfdoc/cpdf_structelement.h
+++ b/core/fpdfdoc/cpdf_structelement.h
@@ -22,6 +22,7 @@
CONSTRUCT_VIA_MAKE_RETAIN;
ByteString GetType() const { return m_Type; }
+ ByteString GetObjType() const;
WideString GetAltText() const;
WideString GetActualText() const;
WideString GetTitle() const;
diff --git a/fpdfsdk/fpdf_structtree.cpp b/fpdfsdk/fpdf_structtree.cpp
index 5dc4ec1..a8800cd 100644
--- a/fpdfsdk/fpdf_structtree.cpp
+++ b/fpdfsdk/fpdf_structtree.cpp
@@ -166,6 +166,18 @@
}
FPDF_EXPORT unsigned long FPDF_CALLCONV
+FPDF_StructElement_GetObjType(FPDF_STRUCTELEMENT struct_element,
+ void* buffer,
+ unsigned long buflen) {
+ CPDF_StructElement* elem =
+ CPDFStructElementFromFPDFStructElement(struct_element);
+ return elem ? WideStringToBuffer(
+ WideString::FromUTF8(elem->GetObjType().AsStringView()),
+ buffer, buflen)
+ : 0;
+}
+
+FPDF_EXPORT unsigned long FPDF_CALLCONV
FPDF_StructElement_GetTitle(FPDF_STRUCTELEMENT struct_element,
void* buffer,
unsigned long buflen) {
diff --git a/fpdfsdk/fpdf_structtree_embeddertest.cpp b/fpdfsdk/fpdf_structtree_embeddertest.cpp
index 6a57631..bc26022 100644
--- a/fpdfsdk/fpdf_structtree_embeddertest.cpp
+++ b/fpdfsdk/fpdf_structtree_embeddertest.cpp
@@ -368,6 +368,58 @@
UnloadPage(page);
}
+TEST_F(FPDFStructTreeEmbedderTest, GetObjType) {
+ ASSERT_TRUE(OpenDocument("tagged_table_bad_elem.pdf"));
+ FPDF_PAGE page = LoadPage(0);
+ ASSERT_TRUE(page);
+
+ {
+ ScopedFPDFStructTree struct_tree(FPDF_StructTree_GetForPage(page));
+ ASSERT_TRUE(struct_tree);
+ ASSERT_EQ(1, FPDF_StructTree_CountChildren(struct_tree.get()));
+
+ FPDF_STRUCTELEMENT child =
+ FPDF_StructTree_GetChildAtIndex(struct_tree.get(), 0);
+ ASSERT_TRUE(child);
+
+ // test nullptr inputs
+ unsigned short buffer[28] = {};
+ ASSERT_EQ(0U,
+ FPDF_StructElement_GetObjType(nullptr, buffer, sizeof(buffer)));
+ ASSERT_EQ(0U, FPDF_StructElement_GetObjType(nullptr, nullptr, 0));
+ ASSERT_EQ(22U, FPDF_StructElement_GetObjType(child, nullptr, 0));
+
+ // Deliberately pass in a small buffer size to make sure `buffer` remains
+ // untouched.
+ ASSERT_EQ(22U, FPDF_StructElement_GetObjType(child, buffer, 1));
+ for (size_t i = 0; i < pdfium::size(buffer); ++i)
+ EXPECT_EQ(0U, buffer[i]);
+
+ ASSERT_EQ(22U,
+ FPDF_StructElement_GetObjType(child, buffer, sizeof(buffer)));
+ EXPECT_EQ(L"StructElem", GetPlatformWString(buffer));
+
+ ASSERT_EQ(1, FPDF_StructElement_CountChildren(child));
+ FPDF_STRUCTELEMENT gchild = FPDF_StructElement_GetChildAtIndex(child, 0);
+ memset(buffer, 0, sizeof(buffer));
+ // Missing /Type in `gchild`
+ ASSERT_EQ(0U,
+ FPDF_StructElement_GetObjType(gchild, buffer, sizeof(buffer)));
+ // Buffer is untouched.
+ for (size_t i = 0; i < pdfium::size(buffer); ++i)
+ EXPECT_EQ(0U, buffer[i]);
+
+ ASSERT_EQ(1, FPDF_StructElement_CountChildren(gchild));
+ FPDF_STRUCTELEMENT ggchild = FPDF_StructElement_GetChildAtIndex(gchild, 0);
+ ASSERT_EQ(28U,
+ FPDF_StructElement_GetObjType(ggchild, buffer, sizeof(buffer)));
+ // Reading bad elem also works.
+ EXPECT_EQ(L"NotStructElem", GetPlatformWString(buffer));
+ }
+
+ UnloadPage(page);
+}
+
TEST_F(FPDFStructTreeEmbedderTest, GetTitle) {
ASSERT_TRUE(OpenDocument("tagged_alt_text.pdf"));
FPDF_PAGE page = LoadPage(0);
diff --git a/fpdfsdk/fpdf_view_c_api_test.c b/fpdfsdk/fpdf_view_c_api_test.c
index 99c53a3..3d75908 100644
--- a/fpdfsdk/fpdf_view_c_api_test.c
+++ b/fpdfsdk/fpdf_view_c_api_test.c
@@ -359,6 +359,7 @@
CHK(FPDF_StructElement_GetID);
CHK(FPDF_StructElement_GetLang);
CHK(FPDF_StructElement_GetMarkedContentID);
+ CHK(FPDF_StructElement_GetObjType);
CHK(FPDF_StructElement_GetStringAttribute);
CHK(FPDF_StructElement_GetTitle);
CHK(FPDF_StructElement_GetType);
diff --git a/public/fpdf_structtree.h b/public/fpdf_structtree.h
index 4784bd4..8daccdb 100644
--- a/public/fpdf_structtree.h
+++ b/public/fpdf_structtree.h
@@ -190,6 +190,27 @@
void* buffer,
unsigned long buflen);
+// Experimental API.
+// Function: FPDF_StructElement_GetObjType
+// Get the object type (/Type) for a given element.
+// Parameters:
+// struct_element - Handle to the struct element.
+// buffer - A buffer for output. May be NULL.
+// buflen - The length of the buffer, in bytes. May be 0.
+// Return value:
+// The number of bytes in the object type, including the terminating
+// NUL character. The number of bytes is returned regardless of the
+// |buffer| and |buflen| parameters.
+// Comments:
+// Regardless of the platform, the |buffer| is always in UTF-16LE
+// encoding. The string is terminated by a UTF16 NUL character. If
+// |buflen| is less than the required length, or |buffer| is NULL,
+// |buffer| will not be modified.
+FPDF_EXPORT unsigned long FPDF_CALLCONV
+FPDF_StructElement_GetObjType(FPDF_STRUCTELEMENT struct_element,
+ void* buffer,
+ unsigned long buflen);
+
// Function: FPDF_StructElement_GetTitle
// Get the title (/T) for a given element.
// Parameters:
