pdf2json 0.1.0

data/pdf2json-0.52-source/xpdf/PDFCore.h

@@ -0,0 +1,321 @@
+//========================================================================
+//
+// PDFCore.h
+//
+// Copyright 2004 Glyph & Cog, LLC
+//
+//========================================================================
+
+#ifndef PDFCORE_H
+#define PDFCORE_H
+
+#include <aconf.h>
+
+#ifdef USE_GCC_PRAGMAS
+#pragma interface
+#endif
+
+#include <stdlib.h>
+#include "SplashTypes.h"
+#include "CharTypes.h"
+
+class GString;
+class GList;
+class SplashBitmap;
+class SplashPattern;
+class BaseStream;
+class PDFDoc;
+class Links;
+class LinkDest;
+class LinkAction;
+class TextPage;
+class HighlightFile;
+class CoreOutputDev;
+class PDFCore;
+
+//------------------------------------------------------------------------
+// zoom factor
+//------------------------------------------------------------------------
+
+#define zoomPage  -1
+#define zoomWidth -2
+#define defZoom   125
+
+//------------------------------------------------------------------------
+
+
+//------------------------------------------------------------------------
+
+// Number of pixels of matte color between pages in continuous mode.
+#define continuousModePageSpacing 3
+
+//------------------------------------------------------------------------
+// PDFCorePage
+//------------------------------------------------------------------------
+
+class PDFCorePage {
+public:
+
+  PDFCorePage(int pageA, int wA, int hA, int tileWA, int tileHA);
+  ~PDFCorePage();
+
+  int page;
+  GList *tiles;			// cached tiles [PDFCoreTile]
+  int xDest, yDest;		// position of upper-left corner
+				//   in the drawing area
+  int w, h;			// size of whole page bitmap
+  int tileW, tileH;		// size of tiles
+  Links *links;			// hyperlinks for this page
+  TextPage *text;		// extracted text
+};
+
+//------------------------------------------------------------------------
+// PDFCoreTile
+//------------------------------------------------------------------------
+
+class PDFCoreTile {
+public:
+
+  PDFCoreTile(int xDestA, int yDestA);
+  virtual ~PDFCoreTile();
+
+  int xMin, yMin, xMax, yMax;
+  int xDest, yDest;
+  Guint edges;
+  SplashBitmap *bitmap;
+  double ctm[6];		// coordinate transform matrix:
+				//   default user space -> device space
+  double ictm[6];		// inverse CTM
+};
+
+#define pdfCoreTileTopEdge      0x01
+#define pdfCoreTileBottomEdge   0x02
+#define pdfCoreTileLeftEdge     0x04
+#define pdfCoreTileRightEdge    0x08
+#define pdfCoreTileTopSpace     0x10
+#define pdfCoreTileBottomSpace  0x20
+
+//------------------------------------------------------------------------
+// PDFHistory
+//------------------------------------------------------------------------
+
+struct PDFHistory {
+  GString *fileName;
+  int page;
+};
+
+#define pdfHistorySize 50
+
+
+//------------------------------------------------------------------------
+// PDFCore
+//------------------------------------------------------------------------
+
+class PDFCore {
+public:
+
+  PDFCore(SplashColorMode colorModeA, int bitmapRowPadA,
+	  GBool reverseVideoA, SplashColorPtr paperColorA,
+	  GBool incrementalUpdate);
+  virtual ~PDFCore();
+
+  //----- loadFile / displayPage / displayDest
+
+  // Load a new file.  Returns pdfOk or error code.
+  virtual int loadFile(GString *fileName, GString *ownerPassword = NULL,
+		       GString *userPassword = NULL);
+
+#ifdef WIN32
+  // Load a new file.  Returns pdfOk or error code.
+  virtual int loadFile(wchar_t *fileName, int fileNameLen,
+		       GString *ownerPassword = NULL,
+		       GString *userPassword = NULL);
+#endif
+
+  // Load a new file, via a Stream instead of a file name.  Returns
+  // pdfOk or error code.
+  virtual int loadFile(BaseStream *stream, GString *ownerPassword = NULL,
+		       GString *userPassword = NULL);
+
+  // Load an already-created PDFDoc object.
+  virtual void loadDoc(PDFDoc *docA);
+
+  // Clear out the current document, if any.
+  virtual void clear();
+
+  // Same as clear(), but returns the PDFDoc object instead of
+  // deleting it.
+  virtual PDFDoc *takeDoc(GBool redraw);
+
+  // Display (or redisplay) the specified page.  If <scrollToTop> is
+  // set, the window is vertically scrolled to the top; otherwise, no
+  // scrolling is done.  If <addToHist> is set, this page change is
+  // added to the history list.
+  virtual void displayPage(int topPageA, double zoomA, int rotateA,
+			   GBool scrollToTop, GBool addToHist);
+
+  // Display a link destination.
+  virtual void displayDest(LinkDest *dest, double zoomA, int rotateA,
+			   GBool addToHist);
+
+  // Update the display, given the specified parameters.
+  virtual void update(int topPageA, int scrollXA, int scrollYA,
+		      double zoomA, int rotateA, GBool force, GBool addToHist);
+
+  //----- page/position changes
+
+  virtual GBool gotoNextPage(int inc, GBool top);
+  virtual GBool gotoPrevPage(int dec, GBool top, GBool bottom);
+  virtual GBool gotoNamedDestination(GString *dest);
+  virtual GBool goForward();
+  virtual GBool goBackward();
+  virtual void scrollLeft(int nCols = 16);
+  virtual void scrollRight(int nCols = 16);
+  virtual void scrollUp(int nLines = 16);
+  virtual void scrollUpPrevPage(int nLines = 16);
+  virtual void scrollDown(int nLines = 16);
+  virtual void scrollDownNextPage(int nLines = 16);
+  virtual void scrollPageUp();
+  virtual void scrollPageDown();
+  virtual void scrollTo(int x, int y);
+  virtual void scrollToLeftEdge();
+  virtual void scrollToRightEdge();
+  virtual void scrollToTopEdge();
+  virtual void scrollToBottomEdge();
+  virtual void scrollToTopLeft();
+  virtual void scrollToBottomRight();
+  virtual void zoomToRect(int pg, double ulx, double uly,
+			  double lrx, double lry);
+  virtual void zoomCentered(double zoomA);
+  virtual void zoomToCurrentWidth();
+  virtual void setContinuousMode(GBool cm);
+
+  //----- selection
+
+  // Current selected region.
+  void setSelection(int newSelectPage,
+		    int newSelectULX, int newSelectULY,
+		    int newSelectLRX, int newSelectLRY);
+  void moveSelection(int pg, int x, int y);
+  GBool getSelection(int *pg, double *ulx, double *uly,
+		     double *lrx, double *lry);
+
+  // Text extraction.
+  GString *extractText(int pg, double xMin, double yMin,
+		       double xMax, double yMax);
+
+  //----- find
+
+  virtual GBool find(char *s, GBool caseSensitive, GBool next, GBool backward,
+		     GBool onePageOnly);
+  virtual GBool findU(Unicode *u, int len, GBool caseSensitive,
+		      GBool next, GBool backward, GBool onePageOnly);
+
+
+  //----- coordinate conversion
+
+  // user space: per-pace, as defined by PDF file; unit = point
+  // device space: (0,0) is upper-left corner of a page; unit = pixel
+  // window space: (0,0) is upper-left corner of drawing area; unit = pixel
+
+  GBool cvtWindowToUser(int xw, int yw, int *pg, double *xu, double *yu);
+  GBool cvtWindowToDev(int xw, int yw, int *pg, int *xd, int *yd);
+  void cvtUserToWindow(int pg, double xy, double yu, int *xw, int *yw);
+  void cvtUserToDev(int pg, double xu, double yu, int *xd, int *yd);
+  void cvtDevToWindow(int pg, int xd, int yd, int *xw, int *yw);
+  void cvtDevToUser(int pg, int xd, int yd, double *xu, double *yu);
+
+  //----- misc access
+
+  PDFDoc *getDoc() { return doc; }
+  int getPageNum() { return topPage; }
+  double getZoom() { return zoom; }
+  double getZoomDPI() { return dpi; }
+  int getRotate() { return rotate; }
+  GBool getContinuousMode() { return continuousMode; }
+  virtual void setReverseVideo(GBool reverseVideoA);
+  GBool canGoBack() { return historyBLen > 1; }
+  GBool canGoForward() { return historyFLen > 0; }
+  int getScrollX() { return scrollX; }
+  int getScrollY() { return scrollY; }
+  int getDrawAreaWidth() { return drawAreaWidth; }
+  int getDrawAreaHeight() { return drawAreaHeight; }
+  virtual void setBusyCursor(GBool busy) = 0;
+  LinkAction *findLink(int pg, double x, double y);
+
+protected:
+
+  int loadFile2(PDFDoc *newDoc);
+  void addPage(int pg, int rot);
+  void needTile(PDFCorePage *page, int x, int y);
+  void xorRectangle(int pg, int x0, int y0, int x1, int y1,
+		    SplashPattern *pattern, PDFCoreTile *oneTile = NULL);
+  int loadHighlightFile(HighlightFile *hf, SplashColorPtr color,
+			SplashColorPtr selectColor, GBool selectable);
+  PDFCorePage *findPage(int pg);
+  static void redrawCbk(void *data, int x0, int y0, int x1, int y1,
+			GBool composited);
+  void redrawWindow(int x, int y, int width, int height,
+		    GBool needUpdate);
+  virtual PDFCoreTile *newTile(int xDestA, int yDestA);
+  virtual void updateTileData(PDFCoreTile *tileA, int xSrc, int ySrc,
+			      int width, int height, GBool composited);
+  virtual void redrawRect(PDFCoreTile *tileA, int xSrc, int ySrc,
+			  int xDest, int yDest, int width, int height,
+			  GBool composited) = 0;
+  void clippedRedrawRect(PDFCoreTile *tile, int xSrc, int ySrc,
+			 int xDest, int yDest, int width, int height,
+			 int xClip, int yClip, int wClip, int hClip,
+			 GBool needUpdate, GBool composited = gTrue);
+  virtual void updateScrollbars() = 0;
+  virtual GBool checkForNewFile() { return gFalse; }
+
+  PDFDoc *doc;			// current PDF file
+  GBool continuousMode;		// false for single-page mode, true for
+				//   continuous mode
+  int drawAreaWidth,		// size of the PDF display area
+      drawAreaHeight;
+  double maxUnscaledPageW,	// maximum unscaled page size
+         maxUnscaledPageH;
+  int maxPageW;			// maximum page width (only used in
+				//   continuous mode)
+  int totalDocH;		// total document height (only used in
+				//   continuous mode)
+  int *pageY;			// top coordinates for each page (only used
+				//   in continuous mode)
+  int topPage;			// page at top of window
+  int scrollX, scrollY;		// offset from top left corner of topPage
+				//   to top left corner of window
+  double zoom;			// current zoom level, in percent of 72 dpi
+  double dpi;			// current zoom level, in DPI
+  int rotate;			// current page rotation
+
+  int selectPage;		// page number of current selection
+  int selectULX,		// coordinates of current selection,
+      selectULY,		//   in device space -- (ULX==LRX || ULY==LRY)
+      selectLRX,		//   means there is no selection
+      selectLRY;
+  GBool dragging;		// set while selection is being dragged
+  GBool lastDragLeft;		// last dragged selection edge was left/right
+  GBool lastDragTop;		// last dragged selection edge was top/bottom
+
+  PDFHistory			// page history queue
+    history[pdfHistorySize];
+  int historyCur;               // currently displayed page
+  int historyBLen;              // number of valid entries backward from
+                                //   current entry
+  int historyFLen;              // number of valid entries forward from
+                                //   current entry
+
+
+  GList *pages;			// cached pages [PDFCorePage]
+  PDFCoreTile *curTile;		// tile currently being rasterized
+  PDFCorePage *curPage;		// page to which curTile belongs
+
+  SplashColor paperColor;
+  CoreOutputDev *out;
+
+  friend class PDFCoreTile;
+};
+
+#endif

data/pdf2json-0.52-source/xpdf/PDFDoc.cc

@@ -0,0 +1,404 @@
+//========================================================================
+//
+// PDFDoc.cc
+//
+// Copyright 1996-2003 Glyph & Cog, LLC
+//
+//========================================================================
+
+#include <aconf.h>
+
+#ifdef USE_GCC_PRAGMAS
+#pragma implementation
+#endif
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stddef.h>
+#include <string.h>
+#ifdef WIN32
+#  include <windows.h>
+#endif
+#include "GString.h"
+#include "config.h"
+#include "GlobalParams.h"
+#include "Page.h"
+#include "Catalog.h"
+#include "Stream.h"
+#include "XRef.h"
+#include "Link.h"
+#include "OutputDev.h"
+#include "Error.h"
+#include "ErrorCodes.h"
+#include "Lexer.h"
+#include "Parser.h"
+#include "SecurityHandler.h"
+#ifndef DISABLE_OUTLINE
+#include "Outline.h"
+#endif
+#include "PDFDoc.h"
+
+//------------------------------------------------------------------------
+
+#define headerSearchSize 1024	// read this many bytes at beginning of
+				//   file to look for '%PDF'
+
+//------------------------------------------------------------------------
+// PDFDoc
+//------------------------------------------------------------------------
+
+PDFDoc::PDFDoc(GString *fileNameA, GString *ownerPassword,
+	       GString *userPassword, void *guiDataA) {
+  Object obj;
+  GString *fileName1, *fileName2;
+
+  ok = gFalse;
+  errCode = errNone;
+
+  guiData = guiDataA;
+
+  file = NULL;
+  str = NULL;
+  xref = NULL;
+  catalog = NULL;
+#ifndef DISABLE_OUTLINE
+  outline = NULL;
+#endif
+
+  fileName = fileNameA;
+  fileName1 = fileName;
+
+
+  // try to open file
+  fileName2 = NULL;
+#ifdef VMS
+  if (!(file = fopen(fileName1->getCString(), "rb", "ctx=stm"))) {
+    error(-1, "Couldn't open file '%s'", fileName1->getCString());
+    errCode = errOpenFile;
+    return;
+  }
+#else
+  if (!(file = fopen(fileName1->getCString(), "rb"))) {
+    fileName2 = fileName->copy();
+    fileName2->lowerCase();
+    if (!(file = fopen(fileName2->getCString(), "rb"))) {
+      fileName2->upperCase();
+      if (!(file = fopen(fileName2->getCString(), "rb"))) {
+	error(-1, "Couldn't open file '%s'", fileName->getCString());
+	delete fileName2;
+	errCode = errOpenFile;
+	return;
+      }
+    }
+    delete fileName2;
+  }
+#endif
+
+  // create stream
+  obj.initNull();
+  str = new FileStream(file, 0, gFalse, 0, &obj);
+
+  ok = setup(ownerPassword, userPassword);
+}
+
+#ifdef WIN32
+PDFDoc::PDFDoc(wchar_t *fileNameA, int fileNameLen, GString *ownerPassword,
+	       GString *userPassword, void *guiDataA) {
+  OSVERSIONINFO version;
+  wchar_t fileName2[_MAX_PATH + 1];
+  Object obj;
+  int i;
+
+  ok = gFalse;
+  errCode = errNone;
+
+  guiData = guiDataA;
+
+  file = NULL;
+  str = NULL;
+  xref = NULL;
+  catalog = NULL;
+#ifndef DISABLE_OUTLINE
+  outline = NULL;
+#endif
+
+  //~ file name should be stored in Unicode (?)
+  fileName = new GString();
+  for (i = 0; i < fileNameLen; ++i) {
+    fileName->append((char)fileNameA[i]);
+  }
+
+  // zero-terminate the file name string
+  for (i = 0; i < fileNameLen && i < _MAX_PATH; ++i) {
+    fileName2[i] = fileNameA[i];
+  }
+  fileName2[i] = 0;
+
+  // try to open file
+  // NB: _wfopen is only available in NT
+  version.dwOSVersionInfoSize = sizeof(version);
+  GetVersionEx(&version);
+  if (version.dwPlatformId == VER_PLATFORM_WIN32_NT) {
+    file = _wfopen(fileName2, L"rb");
+  } else {
+    file = fopen(fileName->getCString(), "rb");
+  }
+  if (!file) {
+    error(-1, "Couldn't open file '%s'", fileName->getCString());
+    errCode = errOpenFile;
+    return;
+  }
+
+  // create stream
+  obj.initNull();
+  str = new FileStream(file, 0, gFalse, 0, &obj);
+
+  ok = setup(ownerPassword, userPassword);
+}
+#endif
+
+PDFDoc::PDFDoc(BaseStream *strA, GString *ownerPassword,
+	       GString *userPassword, void *guiDataA) {
+  ok = gFalse;
+  errCode = errNone;
+  guiData = guiDataA;
+  if (strA->getFileName()) {
+    fileName = strA->getFileName()->copy();
+  } else {
+    fileName = NULL;
+  }
+  file = NULL;
+  str = strA;
+  xref = NULL;
+  catalog = NULL;
+#ifndef DISABLE_OUTLINE
+  outline = NULL;
+#endif
+  ok = setup(ownerPassword, userPassword);
+}
+
+GBool PDFDoc::setup(GString *ownerPassword, GString *userPassword) {
+  str->reset();
+
+  // check header
+  checkHeader();
+
+  // read xref table
+  xref = new XRef(str);
+  if (!xref->isOk()) {
+    error(-1, "Couldn't read xref table");
+    errCode = xref->getErrorCode();
+    return gFalse;
+  }
+
+  // check for encryption
+  if (!checkEncryption(ownerPassword, userPassword)) {
+    errCode = errEncrypted;
+    return gFalse;
+  }
+
+  // read catalog
+  catalog = new Catalog(xref);
+  if (!catalog->isOk()) {
+    error(-1, "Couldn't read page catalog");
+    errCode = errBadCatalog;
+    return gFalse;
+  }
+
+#ifndef DISABLE_OUTLINE
+  // read outline
+  outline = new Outline(catalog->getOutline(), xref);
+#endif
+
+  // done
+  return gTrue;
+}
+
+PDFDoc::~PDFDoc() {
+#ifndef DISABLE_OUTLINE
+  if (outline) {
+    delete outline;
+  }
+#endif
+  if (catalog) {
+    delete catalog;
+  }
+  if (xref) {
+    delete xref;
+  }
+  if (str) {
+    delete str;
+  }
+  if (file) {
+    fclose(file);
+  }
+  if (fileName) {
+    delete fileName;
+  }
+}
+
+// Check for a PDF header on this stream.  Skip past some garbage
+// if necessary.
+void PDFDoc::checkHeader() {
+  char hdrBuf[headerSearchSize+1];
+  char *p;
+  int i;
+
+  pdfVersion = 0;
+  for (i = 0; i < headerSearchSize; ++i) {
+    hdrBuf[i] = str->getChar();
+  }
+  hdrBuf[headerSearchSize] = '\0';
+  for (i = 0; i < headerSearchSize - 5; ++i) {
+    if (!strncmp(&hdrBuf[i], "%PDF-", 5)) {
+      break;
+    }
+  }
+  if (i >= headerSearchSize - 5) {
+    error(-1, "May not be a PDF file (continuing anyway)");
+    return;
+  }
+  str->moveStart(i);
+  if (!(p = strtok(&hdrBuf[i+5], " \t\n\r"))) {
+    error(-1, "May not be a PDF file (continuing anyway)");
+    return;
+  }
+  pdfVersion = atof(p);
+  if (!(hdrBuf[i+5] >= '0' && hdrBuf[i+5] <= '9') ||
+      pdfVersion > supportedPDFVersionNum + 0.0001) {
+    error(-1, "PDF version %s -- xpdf supports version %s"
+	  " (continuing anyway)", p, supportedPDFVersionStr);
+  }
+}
+
+GBool PDFDoc::checkEncryption(GString *ownerPassword, GString *userPassword) {
+  Object encrypt;
+  GBool encrypted;
+  SecurityHandler *secHdlr;
+  GBool ret;
+
+  xref->getTrailerDict()->dictLookup("Encrypt", &encrypt);
+  if ((encrypted = encrypt.isDict())) {
+    if ((secHdlr = SecurityHandler::make(this, &encrypt))) {
+      if (secHdlr->checkEncryption(ownerPassword, userPassword)) {
+	// authorization succeeded
+       	xref->setEncryption(secHdlr->getPermissionFlags(),
+			    secHdlr->getOwnerPasswordOk(),
+			    secHdlr->getFileKey(),
+			    secHdlr->getFileKeyLength(),
+			    secHdlr->getEncVersion(),
+			    secHdlr->getEncAlgorithm());
+	ret = gTrue;
+      } else {
+	// authorization failed
+	ret = gFalse;
+      }
+      delete secHdlr;
+    } else {
+      // couldn't find the matching security handler
+      ret = gFalse;
+    }
+  } else {
+    // document is not encrypted
+    ret = gTrue;
+  }
+  encrypt.free();
+  return ret;
+}
+
+void PDFDoc::displayPage(OutputDev *out, int page,
+			 double hDPI, double vDPI, int rotate,
+			 GBool useMediaBox, GBool crop, GBool printing,
+			 GBool (*abortCheckCbk)(void *data),
+			 void *abortCheckCbkData) {
+  if (globalParams->getPrintCommands()) {
+    printf("***** page %d *****\n", page);
+  }
+    
+    
+  catalog->getPage(page)->display(out, hDPI, vDPI,
+				  rotate, useMediaBox, crop, getLinks(page), printing, catalog,
+				  abortCheckCbk, abortCheckCbkData);
+}
+
+void PDFDoc::displayPages(OutputDev *out, int firstPage, int lastPage,
+			  double hDPI, double vDPI, int rotate,
+			  GBool useMediaBox, GBool crop, GBool printing,
+			  GBool (*abortCheckCbk)(void *data),
+			  void *abortCheckCbkData) {
+  int page;
+
+  for (page = firstPage; page <= lastPage; ++page) {
+    displayPage(out, page, hDPI, vDPI, rotate, useMediaBox, crop, printing,
+		abortCheckCbk, abortCheckCbkData);
+  }
+}
+
+void PDFDoc::displayPageSlice(OutputDev *out, int page,
+			      double hDPI, double vDPI, int rotate,
+			      GBool useMediaBox, GBool crop, GBool printing,
+			      int sliceX, int sliceY, int sliceW, int sliceH,
+			      GBool (*abortCheckCbk)(void *data),
+			      void *abortCheckCbkData) {
+  catalog->getPage(page)->displaySlice(out, hDPI, vDPI,
+				       rotate, useMediaBox, crop,
+				       sliceX, sliceY, sliceW, sliceH,
+				       getLinks(page), printing, catalog,
+				       abortCheckCbk, abortCheckCbkData);
+}
+
+Links *PDFDoc::getLinks(int page) {
+  return catalog->getPage(page)->getLinks(catalog);
+}
+
+void PDFDoc::processLinks(OutputDev *out, int page) {
+  catalog->getPage(page)->processLinks(out, catalog);
+}
+
+GBool PDFDoc::isLinearized() {
+  Parser *parser;
+  Object obj1, obj2, obj3, obj4, obj5;
+  GBool lin;
+
+  lin = gFalse;
+  obj1.initNull();
+  parser = new Parser(xref,
+	     new Lexer(xref,
+	       str->makeSubStream(str->getStart(), gFalse, 0, &obj1)),
+	     gTrue);
+  parser->getObj(&obj1);
+  parser->getObj(&obj2);
+  parser->getObj(&obj3);
+  parser->getObj(&obj4);
+  if (obj1.isInt() && obj2.isInt() && obj3.isCmd("obj") &&
+      obj4.isDict()) {
+    obj4.dictLookup("Linearized", &obj5);
+    if (obj5.isNum() && obj5.getNum() > 0) {
+      lin = gTrue;
+    }
+    obj5.free();
+  }
+  obj4.free();
+  obj3.free();
+  obj2.free();
+  obj1.free();
+  delete parser;
+  return lin;
+}
+
+GBool PDFDoc::saveAs(GString *name) {
+  FILE *f;
+  int c;
+
+  if (!(f = fopen(name->getCString(), "wb"))) {
+    error(-1, "Couldn't open file '%s'", name->getCString());
+    return gFalse;
+  }
+  str->reset();
+  while ((c = str->getChar()) != EOF) {
+    fputc(c, f);
+  }
+  str->close();
+  fclose(f);
+  return gTrue;
+}