Implement FPDFText_GetMatrix().
Add a public API to get the transformation matrix for characters in a
FPDF_TEXTPAGE. Include a new font_matrix.pdf test case, where fonts can
have a size of 1, even though they are visually rendered as though they
have a size of 12.
Bug: pdfium:1445
Change-Id: I2f091eeb839d94d5c93b13b8ae755d0e520f5f40
Reviewed-on: https://pdfium-review.googlesource.com/c/pdfium/+/65353
Commit-Queue: Lei Zhang <thestig@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
diff --git a/fpdfsdk/fpdf_text.cpp b/fpdfsdk/fpdf_text.cpp
index 7c18f17..078bc93 100644
--- a/fpdfsdk/fpdf_text.cpp
+++ b/fpdfsdk/fpdf_text.cpp
@@ -283,6 +283,21 @@
return true;
}
+FPDF_EXPORT FPDF_BOOL FPDF_CALLCONV FPDFText_GetMatrix(FPDF_TEXTPAGE text_page,
+ int index,
+ FS_MATRIX* matrix) {
+ if (!matrix)
+ return false;
+
+ CPDF_TextPage* textpage = GetTextPageForValidIndex(text_page, index);
+ if (!textpage)
+ return false;
+
+ const CPDF_TextPage::CharInfo& charinfo = textpage->GetCharInfo(index);
+ *matrix = FSMatrixFromCFXMatrix(charinfo.m_Matrix);
+ return true;
+}
+
FPDF_EXPORT FPDF_BOOL FPDF_CALLCONV
FPDFText_GetCharOrigin(FPDF_TEXTPAGE text_page,
int index,
@@ -298,7 +313,6 @@
return true;
}
-// select
FPDF_EXPORT int FPDF_CALLCONV
FPDFText_GetCharIndexAtPos(FPDF_TEXTPAGE text_page,
double x,
diff --git a/fpdfsdk/fpdf_text_embeddertest.cpp b/fpdfsdk/fpdf_text_embeddertest.cpp
index 001c0be..06ff4cc 100644
--- a/fpdfsdk/fpdf_text_embeddertest.cpp
+++ b/fpdfsdk/fpdf_text_embeddertest.cpp
@@ -1464,3 +1464,87 @@
FPDFText_ClosePage(text_page);
UnloadPage(page);
}
+
+TEST_F(FPDFTextEmbedderTest, GetMatrix) {
+ constexpr char kExpectedText[] = "A1\r\nA2\r\nA3";
+ constexpr size_t kExpectedTextSize = FX_ArraySize(kExpectedText);
+ constexpr FS_MATRIX kExpectedMatrices[] = {
+ {12.0f, 0.0f, 0.0f, 10.0f, 66.0f, 90.0f},
+ {12.0f, 0.0f, 0.0f, 10.0f, 66.0f, 90.0f},
+ {1.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f},
+ {1.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f},
+ {12.0f, 0.0f, 0.0f, 10.0f, 38.0f, 60.0f},
+ {12.0f, 0.0f, 0.0f, 10.0f, 38.0f, 60.0f},
+ {1.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f},
+ {1.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f},
+ {1.0f, 0.0f, 0.0f, 0.833333, 60.0f, 130.0f},
+ {1.0f, 0.0f, 0.0f, 0.833333, 60.0f, 130.0f},
+ };
+ constexpr size_t kExpectedCount = FX_ArraySize(kExpectedMatrices);
+ static_assert(kExpectedCount + 1 == kExpectedTextSize,
+ "Bad expected matrix size");
+
+ // For a size 12 letter 'A'.
+ constexpr double kExpectedCharWidth = 8.436;
+ constexpr double kExpectedCharHeight = 6.77;
+
+ ASSERT_TRUE(OpenDocument("font_matrix.pdf"));
+ FPDF_PAGE page = LoadPage(0);
+ ASSERT_TRUE(page);
+
+ {
+ ScopedFPDFTextPage text_page(FPDFText_LoadPage(page));
+ ASSERT_TRUE(text_page);
+ ASSERT_EQ(static_cast<int>(kExpectedCount),
+ FPDFText_CountChars(text_page.get()));
+
+ {
+ // Check the characters.
+ unsigned short buffer[kExpectedTextSize];
+ ASSERT_EQ(static_cast<int>(kExpectedTextSize),
+ FPDFText_GetText(text_page.get(), 0, kExpectedCount, buffer));
+ EXPECT_TRUE(
+ check_unsigned_shorts(kExpectedText, buffer, kExpectedTextSize));
+ }
+
+ {
+ // Check the character box size.
+ double left;
+ double right;
+ double bottom;
+ double top;
+ ASSERT_TRUE(FPDFText_GetCharBox(text_page.get(), 0, &left, &right,
+ &bottom, &top));
+ EXPECT_NEAR(kExpectedCharWidth, right - left, 0.001);
+ EXPECT_NEAR(kExpectedCharHeight, top - bottom, 0.001);
+ ASSERT_TRUE(FPDFText_GetCharBox(text_page.get(), 4, &left, &right,
+ &bottom, &top));
+ EXPECT_NEAR(kExpectedCharWidth, right - left, 0.001);
+ EXPECT_NEAR(kExpectedCharHeight, top - bottom, 0.001);
+ ASSERT_TRUE(FPDFText_GetCharBox(text_page.get(), 8, &left, &right,
+ &bottom, &top));
+ EXPECT_NEAR(kExpectedCharWidth, right - left, 0.001);
+ EXPECT_NEAR(kExpectedCharHeight, top - bottom, 0.001);
+ }
+
+ // Check the character matrix.
+ FS_MATRIX matrix;
+ for (size_t i = 0; i < kExpectedCount; ++i) {
+ ASSERT_TRUE(FPDFText_GetMatrix(text_page.get(), i, &matrix)) << i;
+ EXPECT_FLOAT_EQ(kExpectedMatrices[i].a, matrix.a) << i;
+ EXPECT_FLOAT_EQ(kExpectedMatrices[i].b, matrix.b) << i;
+ EXPECT_FLOAT_EQ(kExpectedMatrices[i].c, matrix.c) << i;
+ EXPECT_FLOAT_EQ(kExpectedMatrices[i].d, matrix.d) << i;
+ EXPECT_FLOAT_EQ(kExpectedMatrices[i].e, matrix.e) << i;
+ EXPECT_FLOAT_EQ(kExpectedMatrices[i].f, matrix.f) << i;
+ }
+
+ // Check bad parameters.
+ EXPECT_FALSE(FPDFText_GetMatrix(nullptr, 0, &matrix));
+ EXPECT_FALSE(FPDFText_GetMatrix(text_page.get(), 10, &matrix));
+ EXPECT_FALSE(FPDFText_GetMatrix(text_page.get(), -1, &matrix));
+ EXPECT_FALSE(FPDFText_GetMatrix(text_page.get(), 0, nullptr));
+ }
+
+ UnloadPage(page);
+}
diff --git a/fpdfsdk/fpdf_view_c_api_test.c b/fpdfsdk/fpdf_view_c_api_test.c
index 89ca154..f8a3f1e 100644
--- a/fpdfsdk/fpdf_view_c_api_test.c
+++ b/fpdfsdk/fpdf_view_c_api_test.c
@@ -338,6 +338,7 @@
CHK(FPDFText_GetFontSize);
CHK(FPDFText_GetFontWeight);
CHK(FPDFText_GetLooseCharBox);
+ CHK(FPDFText_GetMatrix);
CHK(FPDFText_GetRect);
CHK(FPDFText_GetSchCount);
CHK(FPDFText_GetSchResultIndex);
diff --git a/public/fpdf_text.h b/public/fpdf_text.h
index 4129282..65554e4 100644
--- a/public/fpdf_text.h
+++ b/public/fpdf_text.h
@@ -262,6 +262,24 @@
FPDF_EXPORT FPDF_BOOL FPDF_CALLCONV
FPDFText_GetLooseCharBox(FPDF_TEXTPAGE text_page, int index, FS_RECTF* rect);
+// Experimental API.
+// Function: FPDFText_GetMatrix
+// Get the effective transformation matrix for a particular character.
+// Parameters:
+// text_page - Handle to a text page information structure.
+// Returned by FPDFText_LoadPage().
+// index - Zero-based index of the character.
+// matrix - Pointer to a FS_MATRIX receiving the transformation
+// matrix.
+// Return Value:
+// On success, return TRUE and fill in |matrix|. If |text_page| is
+// invalid, or if |index| is out of bounds, or if |matrix| is NULL,
+// then return FALSE, and |matrix| remains unmodified.
+//
+FPDF_EXPORT FPDF_BOOL FPDF_CALLCONV FPDFText_GetMatrix(FPDF_TEXTPAGE text_page,
+ int index,
+ FS_MATRIX* matrix);
+
// Function: FPDFText_GetCharOrigin
// Get origin of a particular character.
// Parameters:
diff --git a/testing/resources/font_matrix.in b/testing/resources/font_matrix.in
new file mode 100644
index 0000000..6057e86
--- /dev/null
+++ b/testing/resources/font_matrix.in
@@ -0,0 +1,59 @@
+{{header}}
+{{object 1 0}} <<
+ /Type /Catalog
+ /Pages 2 0 R
+>>
+endobj
+{{object 2 0}} <<
+ /Type /Pages
+ /Count 1
+ /Kids [ 3 0 R ]
+>>
+endobj
+{{object 3 0}} <<
+ /Type /Page
+ /Parent 2 0 R
+ /Resources <<
+ /Font <<
+ /F1 4 0 R
+ >>
+ >>
+ /Contents 5 0 R
+ /MediaBox [ 0 0 100 200 ]
+>>
+endobj
+{{object 4 0}} <<
+ /Type /Font
+ /Subtype /Type1
+ /BaseFont /Times-Roman
+>>
+endobj
+{{object 5 0}} <<
+ {{streamlen}}
+>>
+stream
+q
+12 0 0 10 30 40 cm
+BT
+3 5 Td /F1 1 Tf (A1) Tj
+ET
+Q
+q
+1 0 0 1 20 60 cm
+BT
+12 0 0 10 30 20 Tm
+-1 -2 Td /F1 1 Tf (A2) Tj
+ET
+Q
+q
+1 0 0 0.833333 10 80 cm
+BT
+50 60 Td /F1 12 Tf (A3) Tj
+ET
+Q
+endstream
+endobj
+{{xref}}
+{{trailer}}
+{{startxref}}
+%%EOF
diff --git a/testing/resources/font_matrix.pdf b/testing/resources/font_matrix.pdf
new file mode 100644
index 0000000..435a126
--- /dev/null
+++ b/testing/resources/font_matrix.pdf
@@ -0,0 +1,71 @@
+%PDF-1.7
+% ò¤ô
+1 0 obj <<
+ /Type /Catalog
+ /Pages 2 0 R
+>>
+endobj
+2 0 obj <<
+ /Type /Pages
+ /Count 1
+ /Kids [ 3 0 R ]
+>>
+endobj
+3 0 obj <<
+ /Type /Page
+ /Parent 2 0 R
+ /Resources <<
+ /Font <<
+ /F1 4 0 R
+ >>
+ >>
+ /Contents 5 0 R
+ /MediaBox [ 0 0 100 200 ]
+>>
+endobj
+4 0 obj <<
+ /Type /Font
+ /Subtype /Type1
+ /BaseFont /Times-Roman
+>>
+endobj
+5 0 obj <<
+ /Length 186
+>>
+stream
+q
+12 0 0 10 30 40 cm
+BT
+3 5 Td /F1 1 Tf (A1) Tj
+ET
+Q
+q
+1 0 0 1 20 60 cm
+BT
+12 0 0 10 30 20 Tm
+-1 -2 Td /F1 1 Tf (A2) Tj
+ET
+Q
+q
+1 0 0 0.833333 10 80 cm
+BT
+50 60 Td /F1 12 Tf (A3) Tj
+ET
+Q
+endstream
+endobj
+xref
+0 6
+0000000000 65535 f
+0000000015 00000 n
+0000000068 00000 n
+0000000133 00000 n
+0000000287 00000 n
+0000000365 00000 n
+trailer <<
+ /Root 1 0 R
+ /Size 6
+>>
+startxref
+603
+%%EOF
