pdf2json 0.1.0

data/pdf2json-0.52-source/xpdf/about-text.h

@@ -0,0 +1,48 @@
+//========================================================================
+//
+// about-text.h
+//
+// Copyright 2002-2007 Glyph & Cog, LLC
+//
+//========================================================================
+
+static char *aboutWinText[] = {
+  "http://www.foolabs.com/xpdf/",
+  "derekn@foolabs.com",
+  " ",
+  "Licensed under the GNU General Public License (GPL).",
+  "See the 'COPYING' file for details.",
+  " ",
+  "Supports PDF version " supportedPDFVersionStr ".",
+  " ",
+  "The PDF data structures, operators, and specification",
+  "are copyright 1985-2006 Adobe Systems Inc.",
+  " ",
+  "Mouse bindings:",
+  "  button 1: select text / follow link",
+  "  button 2: pan window",
+  "  button 3: menu",
+  " ",
+  "Key bindings:",
+  "  o              = open file",
+  "  r              = reload",
+  "  f / ctrl-F     = find text",
+  "  ctrl-G         = find next",
+  "  ctrl-P         = print",
+  "  n              = next page",
+  "  p              = previous page",
+  "  <PgDn> = <space>                = scroll down",
+  "  <PgUp> = <backspace> = <delete> = scroll up",
+  "  v              = forward (history path)",
+  "  b              = backward (history path)",
+  "  0 / + / -      = zoom zero / in / out",
+  "  z / w          = zoom page / page width",
+  "  alt-F          = toggle full-screen mode",
+  "  ctrl-L         = redraw",
+  "  q              = quit",
+  "  <home> / <end> = top / bottom of page",
+  "  <arrows>       = scroll",
+  " ",
+  "For more information, please read the xpdf(1) man page.",
+  NULL
+};

data/pdf2json-0.52-source/xpdf/about.xbm

@@ -0,0 +1,6 @@
+#define about_width 10
+#define about_height 15
+static unsigned char about_bits[] = {
+   0x78, 0x00, 0xfc, 0x00, 0xce, 0x01, 0x86, 0x01, 0x80, 0x01, 0x80, 0x01,
+   0xc0, 0x01, 0xe0, 0x00, 0x70, 0x00, 0x30, 0x00, 0x30, 0x00, 0x30, 0x00,
+   0x00, 0x00, 0x30, 0x00, 0x30, 0x00};

data/pdf2json-0.52-source/xpdf/backArrow.xbm

@@ -0,0 +1,6 @@
+#define backArrow_width 16
+#define backArrow_height 15
+static unsigned char backArrow_bits[] = {
+   0x80, 0x00, 0xc0, 0x00, 0xe0, 0x00, 0xf0, 0x00, 0xf8, 0x00, 0xfc, 0xcc,
+   0xfe, 0xcc, 0xff, 0xcc, 0xfe, 0xcc, 0xfc, 0xcc, 0xf8, 0xcc, 0xf0, 0x00,
+   0xe0, 0x00, 0xc0, 0x00, 0x80, 0x00};

data/pdf2json-0.52-source/xpdf/backArrowDis.xbm

@@ -0,0 +1,6 @@
+#define backArrowDis_width 16
+#define backArrowDis_height 15
+static unsigned char backArrowDis_bits[] = {
+   0x80, 0x00, 0x40, 0x00, 0xa0, 0x00, 0x50, 0x00, 0xa8, 0x00, 0x54, 0x44,
+   0xaa, 0x88, 0x55, 0x44, 0xaa, 0x88, 0x54, 0x44, 0xa8, 0x88, 0x50, 0x00,
+   0xa0, 0x00, 0x40, 0x00, 0x80, 0x00};

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

@@ -0,0 +1,112 @@
+//========================================================================
+//
+// config.h
+//
+// Copyright 1996-2007 Glyph & Cog, LLC
+//
+//========================================================================
+
+#ifndef CONFIG_H
+#define CONFIG_H
+
+//------------------------------------------------------------------------
+// version
+//------------------------------------------------------------------------
+
+// xpdf version
+#define xpdfVersion          "3.02"
+#define xpdfVersionNum       3.02
+#define xpdfMajorVersion     3
+#define xpdfMinorVersion     2
+#define xpdfUpdateVersion    0
+#define xpdfMajorVersionStr  "3"
+#define xpdfMinorVersionStr  "2"
+#define xpdfUpdateVersionStr "0"
+
+// supported PDF version
+#define supportedPDFVersionStr "1.7"
+#define supportedPDFVersionNum 1.7
+
+// copyright notice
+#define xpdfCopyright "Copyright 1996-2007 Glyph & Cog, LLC"
+
+// Windows resource file stuff
+#define winxpdfVersion "WinXpdf 3.02"
+#define xpdfCopyrightAmp "Copyright 1996-2007 Glyph && Cog, LLC"
+
+//------------------------------------------------------------------------
+// paper size
+//------------------------------------------------------------------------
+
+// default paper size (in points) for PostScript output
+#ifdef A4_PAPER
+#define defPaperWidth  595    // ISO A4 (210x297 mm)
+#define defPaperHeight 842
+#else
+#define defPaperWidth  612    // American letter (8.5x11")
+#define defPaperHeight 792
+#endif
+
+//------------------------------------------------------------------------
+// config file (xpdfrc) path
+//------------------------------------------------------------------------
+
+// user config file name, relative to the user's home directory
+#if defined(VMS) || (defined(WIN32) && !defined(__CYGWIN32__))
+#define xpdfUserConfigFile "xpdfrc"
+#else
+#define xpdfUserConfigFile ".xpdfrc"
+#endif
+
+// system config file name (set via the configure script)
+#ifdef SYSTEM_XPDFRC
+#define xpdfSysConfigFile SYSTEM_XPDFRC
+#else
+// under Windows, we get the directory with the executable and then
+// append this file name
+#define xpdfSysConfigFile "xpdfrc"
+#endif
+
+//------------------------------------------------------------------------
+// X-related constants
+//------------------------------------------------------------------------
+
+// default maximum size of color cube to allocate
+#define defaultRGBCube 5
+
+// number of fonts (combined t1lib, FreeType, X server) to cache
+#define xOutFontCacheSize 64
+
+// number of Type 3 fonts to cache
+#define xOutT3FontCacheSize 8
+
+//------------------------------------------------------------------------
+// popen
+//------------------------------------------------------------------------
+
+#if defined(_MSC_VER) || defined(__BORLANDC__)
+#define popen _popen
+#define pclose _pclose
+#endif
+
+#if defined(VMS) || defined(VMCMS) || defined(DOS) || defined(OS2) || defined(__EMX__) || defined(WIN32) || defined(__DJGPP__) || defined(MACOS)
+#define POPEN_READ_MODE "rb"
+#else
+#define POPEN_READ_MODE "r"
+#endif
+
+//------------------------------------------------------------------------
+// Win32 stuff
+//------------------------------------------------------------------------
+
+#ifdef CDECL
+#undef CDECL
+#endif
+
+#if defined(_MSC_VER) || defined(__BORLANDC__)
+#define CDECL __cdecl
+#else
+#define CDECL
+#endif
+
+#endif

data/pdf2json-0.52-source/xpdf/dblLeftArrow.xbm

@@ -0,0 +1,6 @@
+#define dblLeftArrow_width 16
+#define dblLeftArrow_height 15
+static unsigned char dblLeftArrow_bits[] = {
+   0x80, 0x80, 0xc0, 0xc0, 0xe0, 0xe0, 0xf0, 0xf0, 0xf8, 0xf8, 0xfc, 0xfc,
+   0xfe, 0xfe, 0xff, 0xff, 0xfe, 0xfe, 0xfc, 0xfc, 0xf8, 0xf8, 0xf0, 0xf0,
+   0xe0, 0xe0, 0xc0, 0xc0, 0x80, 0x80};

data/pdf2json-0.52-source/xpdf/dblLeftArrowDis.xbm

@@ -0,0 +1,6 @@
+#define dblLeftArrowDis_width 16
+#define dblLeftArrowDis_height 15
+static unsigned char dblLeftArrowDis_bits[] = {
+   0x80, 0x80, 0x40, 0x40, 0xa0, 0xa0, 0x50, 0x50, 0xa8, 0xa8, 0x54, 0x54,
+   0xaa, 0xaa, 0x55, 0x55, 0xaa, 0xaa, 0x54, 0x54, 0xa8, 0xa8, 0x50, 0x50,
+   0xa0, 0xa0, 0x40, 0x40, 0x80, 0x80};

data/pdf2json-0.52-source/xpdf/dblRightArrow.xbm

@@ -0,0 +1,6 @@
+#define dblRightArrow_width 16
+#define dblRightArrow_height 15
+static unsigned char dblRightArrow_bits[] = {
+   0x01, 0x01, 0x03, 0x03, 0x07, 0x07, 0x0f, 0x0f, 0x1f, 0x1f, 0x3f, 0x3f,
+   0x7f, 0x7f, 0xff, 0xff, 0x7f, 0x7f, 0x3f, 0x3f, 0x1f, 0x1f, 0x0f, 0x0f,
+   0x07, 0x07, 0x03, 0x03, 0x01, 0x01};

data/pdf2json-0.52-source/xpdf/dblRightArrowDis.xbm

@@ -0,0 +1,6 @@
+#define dblRightArrowDis_width 16
+#define dblRightArrowDis_height 15
+static unsigned char dblRightArrowDis_bits[] = {
+   0x01, 0x01, 0x02, 0x02, 0x05, 0x05, 0x0a, 0x0a, 0x15, 0x15, 0x2a, 0x2a,
+   0x55, 0x55, 0xaa, 0xaa, 0x55, 0x55, 0x2a, 0x2a, 0x15, 0x15, 0x0a, 0x0a,
+   0x05, 0x05, 0x02, 0x02, 0x01, 0x01};

data/pdf2json-0.52-source/xpdf/find.xbm

@@ -0,0 +1,6 @@
+#define find_width 15
+#define find_height 15
+static unsigned char find_bits[] = {
+   0x38, 0x0e, 0x28, 0x0a, 0x2e, 0x3a, 0xfe, 0x3f, 0x7f, 0x7f, 0x61, 0x43,
+   0x61, 0x43, 0x61, 0x43, 0x61, 0x43, 0xe1, 0x43, 0x21, 0x42, 0x21, 0x42,
+   0x21, 0x42, 0x21, 0x42, 0x3f, 0x7e};

data/pdf2json-0.52-source/xpdf/findDis.xbm

@@ -0,0 +1,6 @@
+#define findDis_width 15
+#define findDis_height 15
+static unsigned char findDis_bits[] = {
+   0x10, 0x04, 0x28, 0x0a, 0x04, 0x10, 0xaa, 0x2a, 0x55, 0x55, 0x20, 0x02,
+   0x41, 0x41, 0x20, 0x02, 0x41, 0x41, 0xa0, 0x02, 0x01, 0x40, 0x20, 0x02,
+   0x01, 0x40, 0x20, 0x02, 0x15, 0x54};

data/pdf2json-0.52-source/xpdf/forwardArrow.xbm

@@ -0,0 +1,6 @@
+#define forwardArrow_width 16
+#define forwardArrow_height 15
+static unsigned char forwardArrow_bits[] = {
+   0x00, 0x01, 0x00, 0x03, 0x00, 0x07, 0x00, 0x0f, 0x00, 0x1f, 0x33, 0x3f,
+   0x33, 0x7f, 0x33, 0xff, 0x33, 0x7f, 0x33, 0x3f, 0x00, 0x1f, 0x00, 0x0f,
+   0x00, 0x07, 0x00, 0x03, 0x00, 0x01};

data/pdf2json-0.52-source/xpdf/forwardArrowDis.xbm

@@ -0,0 +1,6 @@
+#define forwardArrowDis_width 16
+#define forwardArrowDis_height 15
+static unsigned char forwardArrowDis_bits[] = {
+   0x00, 0x01, 0x00, 0x02, 0x00, 0x05, 0x00, 0x0a, 0x00, 0x15, 0x22, 0x2a,
+   0x11, 0x55, 0x22, 0xaa, 0x11, 0x55, 0x22, 0x2a, 0x00, 0x15, 0x00, 0x0a,
+   0x00, 0x05, 0x00, 0x02, 0x00, 0x01};

data/pdf2json-0.52-source/xpdf/leftArrow.xbm

@@ -0,0 +1,5 @@
+#define leftArrow_width 8
+#define leftArrow_height 15
+static unsigned char leftArrow_bits[] = {
+   0x80, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc, 0xfe, 0xff, 0xfe, 0xfc, 0xf8, 0xf0,
+   0xe0, 0xc0, 0x80};

data/pdf2json-0.52-source/xpdf/leftArrowDis.xbm

@@ -0,0 +1,5 @@
+#define leftArrowDis_width 8
+#define leftArrowDis_height 15
+static unsigned char leftArrowDis_bits[] = {
+   0x80, 0x40, 0xa0, 0x50, 0xa8, 0x54, 0xaa, 0x55, 0xaa, 0x54, 0xa8, 0x50,
+   0xa0, 0x40, 0x80};

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

@@ -0,0 +1,298 @@
+//========================================================================
+//
+// pdffonts.cc
+//
+// Copyright 2001-2007 Glyph & Cog, LLC
+//
+//========================================================================
+
+#include <aconf.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stddef.h>
+#include <string.h>
+#include <math.h>
+#include "parseargs.h"
+#include "GString.h"
+#include "gmem.h"
+#include "GlobalParams.h"
+#include "Error.h"
+#include "Object.h"
+#include "Dict.h"
+#include "GfxFont.h"
+#include "Annot.h"
+#include "PDFDoc.h"
+#include "config.h"
+
+// NB: this must match the definition of GfxFontType in GfxFont.h.
+static char *fontTypeNames[] = {
+  "unknown",
+  "Type 1",
+  "Type 1C",
+  "Type 1C (OT)",
+  "Type 3",
+  "TrueType",
+  "TrueType (OT)",
+  "CID Type 0",
+  "CID Type 0C",
+  "CID Type 0C (OT)",
+  "CID TrueType",
+  "CID TrueType (OT)"
+};
+
+static void scanFonts(Dict *resDict, PDFDoc *doc);
+static void scanFont(GfxFont *font, PDFDoc *doc);
+
+static int firstPage = 1;
+static int lastPage = 0;
+static char ownerPassword[33] = "\001";
+static char userPassword[33] = "\001";
+static char cfgFileName[256] = "";
+static GBool printVersion = gFalse;
+static GBool printHelp = gFalse;
+
+static ArgDesc argDesc[] = {
+  {"-f",      argInt,      &firstPage,     0,
+   "first page to examine"},
+  {"-l",      argInt,      &lastPage,      0,
+   "last page to examine"},
+  {"-opw",    argString,   ownerPassword,  sizeof(ownerPassword),
+   "owner password (for encrypted files)"},
+  {"-upw",    argString,   userPassword,   sizeof(userPassword),
+   "user password (for encrypted files)"},
+  {"-cfg",        argString,      cfgFileName,    sizeof(cfgFileName),
+   "configuration file to use in place of .xpdfrc"},
+  {"-v",      argFlag,     &printVersion,  0,
+   "print copyright and version info"},
+  {"-h",      argFlag,     &printHelp,     0,
+   "print usage information"},
+  {"-help",   argFlag,     &printHelp,     0,
+   "print usage information"},
+  {"--help",  argFlag,     &printHelp,     0,
+   "print usage information"},
+  {"-?",      argFlag,     &printHelp,     0,
+   "print usage information"},
+  {NULL}
+};
+
+static Ref *fonts;
+static int fontsLen;
+static int fontsSize;
+
+int main(int argc, char *argv[]) {
+  PDFDoc *doc;
+  GString *fileName;
+  GString *ownerPW, *userPW;
+  GBool ok;
+  Page *page;
+  Dict *resDict;
+  Annots *annots;
+  Object obj1, obj2;
+  int pg, i;
+  int exitCode;
+
+  exitCode = 99;
+
+  // parse args
+  ok = parseArgs(argDesc, &argc, argv);
+  if (!ok || argc != 2 || printVersion || printHelp) {
+    fprintf(stderr, "pdffonts version %s\n", xpdfVersion);
+    fprintf(stderr, "%s\n", xpdfCopyright);
+    if (!printVersion) {
+      printUsage("pdffonts", "<PDF-file>", argDesc);
+    }
+    goto err0;
+  }
+  fileName = new GString(argv[1]);
+
+  // read config file
+  globalParams = new GlobalParams(cfgFileName);
+
+  // open PDF file
+  if (ownerPassword[0] != '\001') {
+    ownerPW = new GString(ownerPassword);
+  } else {
+    ownerPW = NULL;
+  }
+  if (userPassword[0] != '\001') {
+    userPW = new GString(userPassword);
+  } else {
+    userPW = NULL;
+  }
+  doc = new PDFDoc(fileName, ownerPW, userPW);
+  if (userPW) {
+    delete userPW;
+  }
+  if (ownerPW) {
+    delete ownerPW;
+  }
+  if (!doc->isOk()) {
+    exitCode = 1;
+    goto err1;
+  }
+
+  // get page range
+  if (firstPage < 1) {
+    firstPage = 1;
+  }
+  if (lastPage < 1 || lastPage > doc->getNumPages()) {
+    lastPage = doc->getNumPages();
+  }
+
+  // scan the fonts
+  printf("name                                 type              emb sub uni object ID\n");
+  printf("------------------------------------ ----------------- --- --- --- ---------\n");
+  fonts = NULL;
+  fontsLen = fontsSize = 0;
+  for (pg = firstPage; pg <= lastPage; ++pg) {
+    page = doc->getCatalog()->getPage(pg);
+    if ((resDict = page->getResourceDict())) {
+      scanFonts(resDict, doc);
+    }
+    annots = new Annots(doc->getXRef(), doc->getCatalog(),
+			page->getAnnots(&obj1));
+    obj1.free();
+    for (i = 0; i < annots->getNumAnnots(); ++i) {
+      if (annots->getAnnot(i)->getAppearance(&obj1)->isStream()) {
+	obj1.streamGetDict()->lookup("Resources", &obj2);
+	if (obj2.isDict()) {
+	  scanFonts(obj2.getDict(), doc);
+	}
+	obj2.free();
+      }
+      obj1.free();
+    }
+    delete annots;
+  }
+
+  exitCode = 0;
+
+  // clean up
+  gfree(fonts);
+ err1:
+  delete doc;
+  delete globalParams;
+ err0:
+
+  // check for memory leaks
+  Object::memCheck(stderr);
+  gMemReport(stderr);
+
+  return exitCode;
+}
+
+static void scanFonts(Dict *resDict, PDFDoc *doc) {
+  Object obj1, obj2, xObjDict, xObj, resObj;
+  Ref r;
+  GfxFontDict *gfxFontDict;
+  GfxFont *font;
+  int i;
+
+  // scan the fonts in this resource dictionary
+  gfxFontDict = NULL;
+  resDict->lookupNF("Font", &obj1);
+  if (obj1.isRef()) {
+    obj1.fetch(doc->getXRef(), &obj2);
+    if (obj2.isDict()) {
+      r = obj1.getRef();
+      gfxFontDict = new GfxFontDict(doc->getXRef(), &r, obj2.getDict());
+    }
+    obj2.free();
+  } else if (obj1.isDict()) {
+    gfxFontDict = new GfxFontDict(doc->getXRef(), NULL, obj1.getDict());
+  }
+  if (gfxFontDict) {
+    for (i = 0; i < gfxFontDict->getNumFonts(); ++i) {
+      if ((font = gfxFontDict->getFont(i))) {
+	scanFont(font, doc);
+      }
+    }
+    delete gfxFontDict;
+  }
+  obj1.free();
+
+  // recursively scan any resource dictionaries in objects in this
+  // resource dictionary
+  resDict->lookup("XObject", &xObjDict);
+  if (xObjDict.isDict()) {
+    for (i = 0; i < xObjDict.dictGetLength(); ++i) {
+      xObjDict.dictGetVal(i, &xObj);
+      if (xObj.isStream()) {
+	xObj.streamGetDict()->lookup("Resources", &resObj);
+	if (resObj.isDict()) {
+	  scanFonts(resObj.getDict(), doc);
+	}
+	resObj.free();
+      }
+      xObj.free();
+    }
+  }
+  xObjDict.free();
+}
+
+static void scanFont(GfxFont *font, PDFDoc *doc) {
+  Ref fontRef, embRef;
+  Object fontObj, toUnicodeObj;
+  GString *name;
+  GBool emb, subset, hasToUnicode;
+  int i;
+
+  fontRef = *font->getID();
+
+  // check for an already-seen font
+  for (i = 0; i < fontsLen; ++i) {
+    if (fontRef.num == fonts[i].num && fontRef.gen == fonts[i].gen) {
+      return;
+    }
+  }
+
+  // font name
+  name = font->getOrigName();
+
+  // check for an embedded font
+  if (font->getType() == fontType3) {
+    emb = gTrue;
+  } else {
+    emb = font->getEmbeddedFontID(&embRef);
+  }
+
+  // look for a ToUnicode map
+  hasToUnicode = gFalse;
+  if (doc->getXRef()->fetch(fontRef.num, fontRef.gen, &fontObj)->isDict()) {
+    hasToUnicode = fontObj.dictLookup("ToUnicode", &toUnicodeObj)->isStream();
+    toUnicodeObj.free();
+  }
+  fontObj.free();
+
+  // check for a font subset name: capital letters followed by a '+'
+  // sign
+  subset = gFalse;
+  if (name) {
+    for (i = 0; i < name->getLength(); ++i) {
+      if (name->getChar(i) < 'A' || name->getChar(i) > 'Z') {
+	break;
+      }
+    }
+    subset = i > 0 && i < name->getLength() && name->getChar(i) == '+';
+  }
+
+  // print the font info
+  printf("%-36s %-17s %-3s %-3s %-3s",
+	 name ? name->getCString() : "[none]",
+	 fontTypeNames[font->getType()],
+	 emb ? "yes" : "no",
+	 subset ? "yes" : "no",
+	 hasToUnicode ? "yes" : "no");
+  if (fontRef.gen >= 100000) {
+    printf(" [none]\n");
+  } else {
+    printf(" %6d %2d\n", fontRef.num, fontRef.gen);
+  }
+
+  // add this font to the list
+  if (fontsLen == fontsSize) {
+    fontsSize += 32;
+    fonts = (Ref *)greallocn(fonts, fontsSize, sizeof(Ref));
+  }
+  fonts[fontsLen++] = *font->getID();
+}