diff options
author | Ashod Nakashian <ashod.nakashian@collabora.co.uk> | 2018-04-22 10:48:51 -0400 |
---|---|---|
committer | Jan Holesovsky <kendy@collabora.com> | 2018-06-07 10:45:29 +0200 |
commit | 67bdfba807b257ce977d61239568877f2bff1429 (patch) | |
tree | 25a149f3ced1278396efa8811113ee52810e9ce2 /svx | |
parent | d28e9e10a893b60b4bad5a1d0bae47a986dd2bdb (diff) |
svx: import processed PDF text
Some PDFs don't include spaces in the text.
Instead, they rely on the explicit positioning
of each character to render visually separated words.
Latex seems to be prone to this approach, though not
exclusively.
Luckily, PDFium does process text and inserts
"generated" spaces where necessary, which is what
we retrieve and use as the text string while importing.
Change-Id: Ic21fe6c8416ecaba66f06b6260f1d6b040ff12af
Diffstat (limited to 'svx')
-rw-r--r-- | svx/source/svdraw/svdpdf.cxx | 30 | ||||
-rw-r--r-- | svx/source/svdraw/svdpdf.hxx | 11 |
2 files changed, 24 insertions, 17 deletions
diff --git a/svx/source/svdraw/svdpdf.cxx b/svx/source/svdraw/svdpdf.cxx index f2fbd7a835ed..1946b61c29bb 100644 --- a/svx/source/svdraw/svdpdf.cxx +++ b/svx/source/svdraw/svdpdf.cxx @@ -225,13 +225,18 @@ void ImpSdrPdfImport::DoLoopActions(SvdProgressInfo* pProgrInfo, sal_uInt32* pAc << ", height: " << dPageHeight); SetupPageScale(dPageWidth, dPageHeight); + // Load the page text to extract it when we get text elements. + FPDF_TEXTPAGE pTextPage = FPDFText_LoadPage(pPdfPage); + const int nPageObjectCount = FPDFPage_CountObject(pPdfPage); for (int nPageObjectIndex = 0; nPageObjectIndex < nPageObjectCount; ++nPageObjectIndex) { FPDF_PAGEOBJECT pPageObject = FPDFPage_GetObject(pPdfPage, nPageObjectIndex); - ImportPdfObject(pPageObject, nPageObjectIndex); + ImportPdfObject(pPageObject, pTextPage, nPageObjectIndex); } + FPDFText_ClosePage(pTextPage); + #if 0 // Now do the text. FPDF_TEXTPAGE pTextPage = FPDFText_LoadPage(pPdfPage); @@ -990,8 +995,8 @@ void ImpSdrPdfImport::checkClip() } bool ImpSdrPdfImport::isClip() const { return !maClip.getB2DRange().isEmpty(); } - -void ImpSdrPdfImport::ImportPdfObject(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex) +void ImpSdrPdfImport::ImportPdfObject(FPDF_PAGEOBJECT pPageObject, FPDF_TEXTPAGE pTextPage, + int nPageObjectIndex) { if (pPageObject == nullptr) return; @@ -1000,7 +1005,7 @@ void ImpSdrPdfImport::ImportPdfObject(FPDF_PAGEOBJECT pPageObject, int nPageObje switch (nPageObjectType) { case FPDF_PAGEOBJ_TEXT: - ImportText(pPageObject, nPageObjectIndex); + ImportText(pPageObject, pTextPage, nPageObjectIndex); break; case FPDF_PAGEOBJ_PATH: ImportPath(pPageObject, nPageObjectIndex); @@ -1012,7 +1017,7 @@ void ImpSdrPdfImport::ImportPdfObject(FPDF_PAGEOBJECT pPageObject, int nPageObje SAL_WARN("sd.filter", "Got page object SHADING: " << nPageObjectIndex); break; case FPDF_PAGEOBJ_FORM: - ImportForm(pPageObject, nPageObjectIndex); + ImportForm(pPageObject, pTextPage, nPageObjectIndex); break; default: SAL_WARN("sd.filter", "Unknown PDF page object #" << nPageObjectIndex @@ -1021,7 +1026,8 @@ void ImpSdrPdfImport::ImportPdfObject(FPDF_PAGEOBJECT pPageObject, int nPageObje } } -void ImpSdrPdfImport::ImportForm(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex) +void ImpSdrPdfImport::ImportForm(FPDF_PAGEOBJECT pPageObject, FPDF_TEXTPAGE pTextPage, + int nPageObjectIndex) { SAL_WARN("sd.filter", "Got page object FORM: " << nPageObjectIndex); @@ -1036,14 +1042,15 @@ void ImpSdrPdfImport::ImportForm(FPDF_PAGEOBJECT pPageObject, int nPageObjectInd for (int nIndex = 0; nIndex < nCount; ++nIndex) { FPDF_PAGEOBJECT pFormObject = FPDFFormObj_GetSubObject(pPageObject, nIndex); - ImportPdfObject(pFormObject, -1); + ImportPdfObject(pFormObject, pTextPage, -1); } // Restore the old one. mCurMatrix = aOldMatrix; } -void ImpSdrPdfImport::ImportText(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex) +void ImpSdrPdfImport::ImportText(FPDF_PAGEOBJECT pPageObject, FPDF_TEXTPAGE pTextPage, + int nPageObjectIndex) { SAL_WARN("sd.filter", "Got page object TEXT: " << nPageObjectIndex); float left; @@ -1075,14 +1082,15 @@ void ImpSdrPdfImport::ImportText(FPDF_PAGEOBJECT pPageObject, int nPageObjectInd SAL_WARN("sd.filter", "Got TEXT origin: " << aPos); SAL_WARN("sd.filter", "Got TEXT Bounds: " << aRect); - const int nChars = FPDFTextObj_CountChars(pPageObject); + const int nChars = FPDFTextObj_CountChars(pPageObject) * 2; std::unique_ptr<sal_Unicode[]> pText(new sal_Unicode[nChars + 1]); // + terminating null unsigned short* pShortText = reinterpret_cast<unsigned short*>(pText.get()); - const int nActualChars = FPDFTextObj_GetText(pPageObject, 0, nChars, pShortText); + const int nActualChars + = FPDFTextObj_GetTextProcessed(pPageObject, pTextPage, 0, nChars, pShortText); if (nActualChars <= 0) { - SAL_WARN("sd.filter", "Got not TEXT"); + SAL_WARN("sd.filter", "Got no TEXT"); return; } diff --git a/svx/source/svdraw/svdpdf.hxx b/svx/source/svdraw/svdpdf.hxx index 460b508e83a8..da54b9a40fa8 100644 --- a/svx/source/svdraw/svdpdf.hxx +++ b/svx/source/svdraw/svdpdf.hxx @@ -42,6 +42,7 @@ class SdrObject; class SvdProgressInfo; typedef void* FPDF_DOCUMENT; typedef void* FPDF_PAGEOBJECT; +typedef void* FPDF_TEXTPAGE; // Helper Class to import PDF class ImpSdrPdfImport final @@ -86,7 +87,6 @@ class ImpSdrPdfImport final double d() const { return md; } double e() const { return me; } double f() const { return mf; } - /// Mutliply this * other. void Concatinate(const Matrix& other) { @@ -156,7 +156,6 @@ class ImpSdrPdfImport final /// Correct the vertical coordinate to start at the top. /// PDF coordinate system has orign at the bottom right. double correctVertOrigin(double offsetPts) const { return mdPageHeightPts - offsetPts; } - /// Convert PDF points to logic (twips). tools::Rectangle PointsToLogic(double left, double right, double top, double bottom) const; Point PointsToLogic(double x, double y) const; @@ -165,11 +164,12 @@ class ImpSdrPdfImport final void checkClip(); bool isClip() const; - void ImportPdfObject(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex); - void ImportForm(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex); + void ImportPdfObject(FPDF_PAGEOBJECT pPageObject, FPDF_TEXTPAGE pTextPage, + int nPageObjectIndex); + void ImportForm(FPDF_PAGEOBJECT pPageObject, FPDF_TEXTPAGE pTextPage, int nPageObjectIndex); void ImportImage(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex); void ImportPath(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex); - void ImportText(FPDF_PAGEOBJECT pPageObject, int nPageObjectIndex); + void ImportText(FPDF_PAGEOBJECT pPageObject, FPDF_TEXTPAGE pTextPage, int nPageObjectIndex); void ImportText(const Point& rPos, const Size& rSize, const OUString& rStr); void SetupPageScale(const double dPageWidth, const double dPageHeight); @@ -193,7 +193,6 @@ public: ~ImpSdrPdfImport(); int GetPageCount() const { return mnPageCount; } - size_t DoImport(SdrObjList& rDestList, size_t nInsPos, int nPageNumber, SvdProgressInfo* pProgrInfo = nullptr); }; |