diff options
Diffstat (limited to 'fpdfsdk')
-rw-r--r-- | fpdfsdk/include/fpdftext.h | 38 | ||||
-rw-r--r-- | fpdfsdk/src/fpdftext.cpp | 13 | ||||
-rw-r--r-- | fpdfsdk/src/fpdftext_embeddertest.cpp | 346 |
3 files changed, 374 insertions, 23 deletions
diff --git a/fpdfsdk/include/fpdftext.h b/fpdfsdk/include/fpdftext.h index 8b897798e6..d37715f6b8 100644 --- a/fpdfsdk/include/fpdftext.h +++ b/fpdfsdk/include/fpdftext.h @@ -16,32 +16,31 @@ extern "C" { // Function: FPDFText_LoadPage // Prepare information about all characters in a page. -// Parameters: -// page - Handle to the page. Returned by FPDF_LoadPage function (in FPDFVIEW module). +// Parameters: +// page - Handle to the page. Returned by FPDF_LoadPage function (in FPDFVIEW module). // Return value: // A handle to the text page information structure. // NULL if something goes wrong. // Comments: // Application must call FPDFText_ClosePage to release the text page information. -// If you don't purchase Text Module , this function will return NULL. -// +// DLLEXPORT FPDF_TEXTPAGE STDCALL FPDFText_LoadPage(FPDF_PAGE page); // Function: FPDFText_ClosePage // Release all resources allocated for a text page information structure. -// Parameters: +// Parameters: // text_page - Handle to a text page information structure. Returned by FPDFText_LoadPage function. // Return Value: // None. // DLLEXPORT void STDCALL FPDFText_ClosePage(FPDF_TEXTPAGE text_page); - + // Function: FPDFText_CountChars // Get number of characters in a page. -// Parameters: +// Parameters: // text_page - Handle to a text page information structure. Returned by FPDFText_LoadPage function. // Return value: -// Number of characters in the page. Return -1 for error. +// Number of characters in the page. Return -1 for error. // Generated characters, like additional space characters, new line characters, are also counted. // Comments: // Characters in a page form a "stream", inside the stream, each character has an index. @@ -52,7 +51,7 @@ DLLEXPORT int STDCALL FPDFText_CountChars(FPDF_TEXTPAGE text_page); // Function: FPDFText_GetUnicode // Get Unicode of a character in a page. -// Parameters: +// Parameters: // text_page - Handle to a text page information structure. Returned by FPDFText_LoadPage function. // index - Zero-based index of the character. // Return value: @@ -64,7 +63,7 @@ DLLEXPORT unsigned int STDCALL FPDFText_GetUnicode(FPDF_TEXTPAGE text_page, int // Function: FPDFText_GetFontSize // Get the font size of a particular character. -// Parameters: +// Parameters: // text_page - Handle to a text page information structure. Returned by FPDFText_LoadPage function. // index - Zero-based index of the character. // Return value: @@ -75,7 +74,7 @@ DLLEXPORT double STDCALL FPDFText_GetFontSize(FPDF_TEXTPAGE text_page, int index // Function: FPDFText_GetCharBox // Get bounding box of a particular character. -// Parameters: +// Parameters: // text_page - Handle to a text page information structure. Returned by FPDFText_LoadPage function. // index - Zero-based index of the character. // left - Pointer to a double number receiving left position of the character box. @@ -118,7 +117,7 @@ DLLEXPORT int STDCALL FPDFText_GetCharIndexAtPos(FPDF_TEXTPAGE text_page, // Number of characters written into the result buffer, including the trailing terminator. // Comments: // This function ignores characters without unicode information. -// +// DLLEXPORT int STDCALL FPDFText_GetText(FPDF_TEXTPAGE text_page, int start_index, int count, unsigned short* result); // Function: FPDFText_CountRects @@ -163,10 +162,15 @@ DLLEXPORT void STDCALL FPDFText_GetRect(FPDF_TEXTPAGE text_page, int rect_index, // buffer - A unicode buffer. // buflen - Number of characters (not bytes) for the buffer, excluding an additional terminator. // Return Value: -// If buffer is NULL or buflen is zero, return number of characters (not bytes) needed, -// otherwise, return number of characters copied into the buffer. +// If buffer is NULL or buflen is zero, return number of characters (not bytes) of text present within +// the rectangle, excluding a terminating NUL. Generally you should pass a buffer at least one larger +// than this if you want a terminating NUL, which will be provided if space is available. +// Otherwise, return number of characters copied into the buffer, including the terminating NUL +// when space for it is available. +// Comment: +// If the buffer is too small, as much text as will fit is copied into it. // -DLLEXPORT int STDCALL FPDFText_GetBoundedText(FPDF_TEXTPAGE text_page,double left, double top, +DLLEXPORT int STDCALL FPDFText_GetBoundedText(FPDF_TEXTPAGE text_page,double left, double top, double right, double bottom,unsigned short* buffer,int buflen); @@ -236,7 +240,7 @@ DLLEXPORT void STDCALL FPDFText_FindClose(FPDF_SCHHANDLE handle); // Prepare information about weblinks in a page. // Parameters: // text_page - Handle to a text page information structure. Returned by FPDFText_LoadPage function. -// Return Value: +// Return Value: // A handle to the page's links information structure. // NULL if something goes wrong. // Comments: @@ -296,7 +300,7 @@ DLLEXPORT int STDCALL FPDFLink_CountRects(FPDF_PAGELINK link_page, int link_inde // Return Value: // None. // -DLLEXPORT void STDCALL FPDFLink_GetRect(FPDF_PAGELINK link_page, int link_index, int rect_index, +DLLEXPORT void STDCALL FPDFLink_GetRect(FPDF_PAGELINK link_page, int link_index, int rect_index, double* left, double* top,double* right, double* bottom); // Function: FPDFLink_CloseWebLinks diff --git a/fpdfsdk/src/fpdftext.cpp b/fpdfsdk/src/fpdftext.cpp index 264631b630..63de4433b3 100644 --- a/fpdfsdk/src/fpdftext.cpp +++ b/fpdfsdk/src/fpdftext.cpp @@ -235,12 +235,13 @@ DLLEXPORT void STDCALL FPDFLink_GetRect(FPDF_PAGELINK link_page,int link_index, IPDF_LinkExtract* pageLink=(IPDF_LinkExtract*)link_page; CFX_RectArray rectArray; pageLink->GetRects(link_index,rectArray); - CFX_FloatRect rect; - rect=rectArray.GetAt(rect_index); - *left=rect.left; - *right=rect.right; - *top=rect.top; - *bottom=rect.bottom; + if (rect_index >= 0 && rect_index < rectArray.GetSize()) { + CFX_FloatRect rect=rectArray.GetAt(rect_index); + *left=rect.left; + *right=rect.right; + *top=rect.top; + *bottom=rect.bottom; + } } DLLEXPORT void STDCALL FPDFLink_CloseWebLinks(FPDF_PAGELINK link_page) { diff --git a/fpdfsdk/src/fpdftext_embeddertest.cpp b/fpdfsdk/src/fpdftext_embeddertest.cpp index 3198fe0f86..1457832658 100644 --- a/fpdfsdk/src/fpdftext_embeddertest.cpp +++ b/fpdfsdk/src/fpdftext_embeddertest.cpp @@ -7,17 +7,363 @@ #include "../../fpdfsdk/include/fpdftext.h" #include "testing/gtest/include/gtest/gtest.h" +namespace { + +static bool check_unsigned_shorts(const char* expected, + const unsigned short* actual, + size_t length) { + if (length > strlen(expected) + 1) { + return false; + } + for (size_t i = 0; i < length; ++i) { + if (actual[i] != static_cast<unsigned short>(expected[i])) { + return false; + } + } + return true; +} + +} // namespace + class FPDFTextEmbeddertest : public EmbedderTest { }; +TEST_F(FPDFTextEmbeddertest, Text) { + EXPECT_TRUE(OpenDocument("testing/resources/hello_world.pdf")); + FPDF_FORMHANDLE form_handle = SetFormFillEnvironment(); + FPDF_PAGE page = LoadPage(0, form_handle); + EXPECT_NE(nullptr, page); + + FPDF_TEXTPAGE textpage = FPDFText_LoadPage(page); + EXPECT_NE(nullptr, textpage); + + const char expected[] = "Hello, world!\r\nGoodbye, world!"; + unsigned short fixed_buffer[128]; + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + + // Check includes the terminating NUL that is provided. + EXPECT_EQ(sizeof(expected), FPDFText_GetText(textpage, 0, 128, fixed_buffer)); + EXPECT_TRUE(check_unsigned_shorts(expected, fixed_buffer, sizeof(expected))); + + // Count does not include the terminating NUL in the string literal. + EXPECT_EQ(sizeof(expected) - 1, FPDFText_CountChars(textpage)); + for (size_t i = 0; i < sizeof(expected) - 1; ++i) { + EXPECT_EQ(expected[i], FPDFText_GetUnicode(textpage, i)) << " at " << i; + } + + EXPECT_EQ(12.0, FPDFText_GetFontSize(textpage, 0)); + EXPECT_EQ(16.0, FPDFText_GetFontSize(textpage, 15)); + + double left = 0.0; + double right = 0.0; + double bottom = 0.0; + double top = 0.0; + FPDFText_GetCharBox(textpage, 4, &left, &right, &bottom, &top); + EXPECT_NEAR(41.071, left, 0.001); + EXPECT_NEAR(46.243, right, 0.001); + EXPECT_NEAR(49.844, bottom, 0.001); + EXPECT_NEAR(55.520, top, 0.001); + + EXPECT_EQ(4, FPDFText_GetCharIndexAtPos( + textpage, 42.0, 50.0, 1.0, 1.0)); + EXPECT_EQ(-1, FPDFText_GetCharIndexAtPos( + textpage, 0.0, 0.0, 1.0, 1.0)); + EXPECT_EQ(-1, FPDFText_GetCharIndexAtPos( + textpage, 199.0, 199.0, 1.0, 1.0)); + + // Test out of range indicies. + EXPECT_EQ(-1, FPDFText_GetCharIndexAtPos( + textpage, 42.0, 10000000.0, 1.0, 1.0)); + EXPECT_EQ(-1, FPDFText_GetCharIndexAtPos( + textpage, -1.0, 50.0, 1.0, 1.0)); + + // Count does not include the terminating NUL in the string literal. + EXPECT_EQ(2, FPDFText_CountRects(textpage, 0, sizeof(expected) - 1)); + + left = 0.0; + right = 0.0; + bottom = 0.0; + top = 0.0; + FPDFText_GetRect(textpage, 1, &left, &top, &right, &bottom); + EXPECT_NEAR(20.847, left, 0.001); + EXPECT_NEAR(135.167, right, 0.001); + EXPECT_NEAR(96.655, bottom, 0.001); + EXPECT_NEAR(116.000, top, 0.001); + + // Test out of range indicies set outputs to (0.0, 0.0, 0.0, 0.0). + left = -1.0; + right = -1.0; + bottom = -1.0; + top = -1.0; + FPDFText_GetRect(textpage, -1, &left, &top, &right, &bottom); + EXPECT_EQ(0.0, left); + EXPECT_EQ(0.0, right); + EXPECT_EQ(0.0, bottom); + EXPECT_EQ(0.0, top); + + left = -2.0; + right = -2.0; + bottom = -2.0; + top = -2.0; + FPDFText_GetRect(textpage, 2, &left, &top, &right, &bottom); + EXPECT_EQ(0.0, left); + EXPECT_EQ(0.0, right); + EXPECT_EQ(0.0, bottom); + EXPECT_EQ(0.0, top); + + EXPECT_EQ(9, FPDFText_GetBoundedText( + textpage, 41.0, 56.0, 82.0, 48.0, 0, 0)); + + // Extract starting at character 4 as above. + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(1, FPDFText_GetBoundedText( + textpage, 41.0, 56.0, 82.0, 48.0, fixed_buffer, 1)); + EXPECT_TRUE(check_unsigned_shorts(expected + 4, fixed_buffer, 1)); + EXPECT_EQ(0xbdbd, fixed_buffer[1]); + + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(9, FPDFText_GetBoundedText( + textpage, 41.0, 56.0, 82.0, 48.0, fixed_buffer, 9)); + EXPECT_TRUE(check_unsigned_shorts(expected + 4, fixed_buffer, 9)); + EXPECT_EQ(0xbdbd, fixed_buffer[9]); + + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(10, FPDFText_GetBoundedText( + textpage, 41.0, 56.0, 82.0, 48.0, fixed_buffer, 128)); + EXPECT_TRUE(check_unsigned_shorts(expected + 4, fixed_buffer, 9)); + EXPECT_EQ(0u, fixed_buffer[9]); + EXPECT_EQ(0xbdbd, fixed_buffer[10]); + + FPDFText_ClosePage(textpage); + ClearFormFillEnvironment(form_handle); +} + +TEST_F(FPDFTextEmbeddertest, TextSearch) { + EXPECT_TRUE(OpenDocument("testing/resources/hello_world.pdf")); + FPDF_FORMHANDLE form_handle = SetFormFillEnvironment(); + FPDF_PAGE page = LoadPage(0, form_handle); + EXPECT_NE(nullptr, page); + + FPDF_TEXTPAGE textpage = FPDFText_LoadPage(page); + EXPECT_NE(nullptr, textpage); + + // Avoid issues with system wchar_t width vs. FPDF_WideString. + const unsigned short nope[] = { 'n', 'o', 'p', 'e', '\0' }; + const unsigned short world[] = { 'w', 'o', 'r', 'l', 'd', '\0' }; + const unsigned short world_caps[] = { 'W', 'O', 'R', 'L', 'D', '\0' }; + const unsigned short world_substr[] = { 'o', 'r', 'l', 'd', '\0' }; + + // No occurences of "nope" in test page. + FPDF_SCHHANDLE search = FPDFText_FindStart(textpage, nope, 0, 0); + EXPECT_NE(nullptr, search); + EXPECT_EQ(0, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(0, FPDFText_GetSchCount(search)); + + // Advancing finds nothing. + EXPECT_FALSE(FPDFText_FindNext(search)); + EXPECT_EQ(0, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(0, FPDFText_GetSchCount(search)); + + // Retreating finds nothing. + EXPECT_FALSE(FPDFText_FindPrev(search)); + EXPECT_EQ(0, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(0, FPDFText_GetSchCount(search)); + FPDFText_FindClose(search); + + // Two occurences of "world" in test page. + search = FPDFText_FindStart(textpage, world, 0, 2); + EXPECT_NE(nullptr, search); + + // Remains not found until advanced. + EXPECT_EQ(0, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(0, FPDFText_GetSchCount(search)); + + // First occurence of "world" in this test page. + EXPECT_TRUE(FPDFText_FindNext(search)); + EXPECT_EQ(7, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + + // Last occurence of "world" in this test page. + EXPECT_TRUE(FPDFText_FindNext(search)); + EXPECT_EQ(24, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + + // Found position unchanged when fails to advance. + EXPECT_FALSE(FPDFText_FindNext(search)); + EXPECT_EQ(24, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + + // Back to first occurence. + EXPECT_TRUE(FPDFText_FindPrev(search)); + EXPECT_EQ(7, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + + // Found position unchanged when fails to retreat. + EXPECT_FALSE(FPDFText_FindPrev(search)); + EXPECT_EQ(7, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + FPDFText_FindClose(search); + + // Exact search unaffected by case sensitiity and whole word flags. + search = FPDFText_FindStart( + textpage, world, FPDF_MATCHCASE | FPDF_MATCHWHOLEWORD, 0); + EXPECT_NE(nullptr, search); + EXPECT_TRUE(FPDFText_FindNext(search)); + EXPECT_EQ(7, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + FPDFText_FindClose(search); + + // Default is case-insensitive, so matching agaist caps works. + search = FPDFText_FindStart(textpage, world_caps, 0, 0); + EXPECT_NE(nullptr, search); + EXPECT_TRUE(FPDFText_FindNext(search)); + EXPECT_EQ(7, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(5, FPDFText_GetSchCount(search)); + FPDFText_FindClose(search); + + // But can be made case sensitive, in which case this fails. + search = FPDFText_FindStart(textpage, world_caps, FPDF_MATCHCASE, 0); + EXPECT_FALSE(FPDFText_FindNext(search)); + EXPECT_EQ(0, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(0, FPDFText_GetSchCount(search)); + FPDFText_FindClose(search); + + // Default is match anywhere within word, so matching substirng works. + search = FPDFText_FindStart(textpage, world_substr, 0, 0); + EXPECT_TRUE(FPDFText_FindNext(search)); + EXPECT_EQ(8, FPDFText_GetSchResultIndex(search)); + EXPECT_EQ(4, FPDFText_GetSchCount(search)); + FPDFText_FindClose(search); + + // But can be made to mach word boundaries, in which case this fails. + search = FPDFText_FindStart(textpage, world_substr, FPDF_MATCHWHOLEWORD, 0); + EXPECT_FALSE(FPDFText_FindNext(search)); + // TODO(tsepez): investigate strange index/count values in this state. + FPDFText_FindClose(search); + + FPDFText_ClosePage(textpage); + ClearFormFillEnvironment(form_handle); +} + // Test that the page has characters despite a bad stream length. TEST_F(FPDFTextEmbeddertest, StreamLengthPastEndOfFile) { EXPECT_TRUE(OpenDocument("testing/resources/bug_57.pdf")); FPDF_FORMHANDLE form_handle = SetFormFillEnvironment(); FPDF_PAGE page = LoadPage(0, form_handle); EXPECT_NE(nullptr, page); + FPDF_TEXTPAGE textpage = FPDFText_LoadPage(page); EXPECT_NE(nullptr, textpage); EXPECT_EQ(13, FPDFText_CountChars(textpage)); + + FPDFText_ClosePage(textpage); + ClearFormFillEnvironment(form_handle); +} + +TEST_F(FPDFTextEmbeddertest, WebLinks) { + EXPECT_TRUE(OpenDocument("testing/resources/weblinks.pdf")); + FPDF_FORMHANDLE form_handle = SetFormFillEnvironment(); + FPDF_PAGE page = LoadPage(0, form_handle); + EXPECT_NE(nullptr, page); + + FPDF_TEXTPAGE textpage = FPDFText_LoadPage(page); + EXPECT_NE(nullptr, textpage); + + FPDF_PAGELINK pagelink = FPDFLink_LoadWebLinks(textpage); + EXPECT_NE(nullptr, pagelink); + + // Page contains two HTTP-style URLs. + EXPECT_EQ(2, FPDFLink_CountWebLinks(pagelink)); + + // Only a terminating NUL required for bogus links. + EXPECT_EQ(1, FPDFLink_GetURL(pagelink, 2, nullptr, 0)); + EXPECT_EQ(1, FPDFLink_GetURL(pagelink, 1400, nullptr, 0)); + EXPECT_EQ(1, FPDFLink_GetURL(pagelink, -1, nullptr, 0)); + + // Query the number of characters required for each link (incl NUL). + EXPECT_EQ(25, FPDFLink_GetURL(pagelink, 0, nullptr, 0)); + EXPECT_EQ(26, FPDFLink_GetURL(pagelink, 1, nullptr, 0)); + + const char expected_url[] = "http://example.com?q=foo"; + unsigned short fixed_buffer[128]; + + // Retrieve a link with too small a buffer. Buffer will not be + // NUL-terminated, but must not be modified past indicated length, + // so pre-fill with a pattern to check write bounds. + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(1, FPDFLink_GetURL(pagelink, 0, fixed_buffer, 1)); + EXPECT_TRUE(check_unsigned_shorts(expected_url, fixed_buffer, 1)); + EXPECT_EQ(0xbdbd, fixed_buffer[1]); + + // Check buffer that doesn't have space for a terminating NUL. + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(sizeof(expected_url) - 1, FPDFLink_GetURL( + pagelink, 0, fixed_buffer, sizeof(expected_url) - 1)); + EXPECT_TRUE(check_unsigned_shorts( + expected_url, fixed_buffer, sizeof(expected_url) - 1)); + EXPECT_EQ(0xbdbd, fixed_buffer[sizeof(expected_url) - 1]); + + // Retreive link with exactly-sized buffer. + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(sizeof(expected_url), FPDFLink_GetURL( + pagelink, 0, fixed_buffer, sizeof(expected_url))); + EXPECT_TRUE(check_unsigned_shorts( + expected_url, fixed_buffer, sizeof(expected_url))); + EXPECT_EQ(0u, fixed_buffer[sizeof(expected_url) - 1]); + EXPECT_EQ(0xbdbd, fixed_buffer[sizeof(expected_url)]); + + // Retreive link with ample-sized-buffer. + memset(fixed_buffer, 0xbd, sizeof(fixed_buffer)); + EXPECT_EQ(sizeof(expected_url), FPDFLink_GetURL( + pagelink, 0, fixed_buffer, 128)); + EXPECT_TRUE(check_unsigned_shorts( + expected_url, fixed_buffer, sizeof(expected_url))); + EXPECT_EQ(0u, fixed_buffer[sizeof(expected_url) - 1]); + EXPECT_EQ(0xbdbd, fixed_buffer[sizeof(expected_url)]); + + // Each link rendered in a single rect in this test page. + EXPECT_EQ(1, FPDFLink_CountRects(pagelink, 0)); + EXPECT_EQ(1, FPDFLink_CountRects(pagelink, 1)); + + // Each link rendered in a single rect in this test page. + EXPECT_EQ(0, FPDFLink_CountRects(pagelink, -1)); + EXPECT_EQ(0, FPDFLink_CountRects(pagelink, 2)); + EXPECT_EQ(0, FPDFLink_CountRects(pagelink, 10000)); + + // Check boundary of valid link index with valid rect index. + double left = 0.0; + double right = 0.0; + double top = 0.0; + double bottom = 0.0; + FPDFLink_GetRect(pagelink, 0, 0, &left, &top, &right, &bottom); + EXPECT_NEAR(50.791, left, 0.001); + EXPECT_NEAR(187.963, right, 0.001); + EXPECT_NEAR(97.624, bottom, 0.001); + EXPECT_NEAR(108.736, top, 0.001); + + // Check that valid link with invalid rect index leaves parameters unchanged. + left = -1.0; + right = -1.0; + top = -1.0; + bottom = -1.0; + FPDFLink_GetRect(pagelink, 0, 1, &left, &top, &right, &bottom); + EXPECT_EQ(-1.0, left); + EXPECT_EQ(-1.0, right); + EXPECT_EQ(-1.0, bottom); + EXPECT_EQ(-1.0, top); + + // Check that invalid link index leaves parameters unchanged. + left = -2.0; + right = -2.0; + top = -2.0; + bottom = -2.0; + FPDFLink_GetRect(pagelink, -1, 0, &left, &top, &right, &bottom); + EXPECT_EQ(-2.0, left); + EXPECT_EQ(-2.0, right); + EXPECT_EQ(-2.0, bottom); + EXPECT_EQ(-2.0, top); + + FPDFLink_CloseWebLinks(pagelink); + FPDFText_ClosePage(textpage); ClearFormFillEnvironment(form_handle); } |