From 9f87bacc0316ed46ca1df575b180441e7802de9d Mon Sep 17 00:00:00 2001
From: Joshua Richardson <jricjr4@gmail.com>
Date: Wed, 22 Jun 2011 20:04:52 -0700
Subject: [PATCH 3/3] Changed to HTML5/XHTML5.  Now rotation should work across all HTML5 compliant
 browsers.

---
 utils/HtmlFonts.cc     |   17 +++-------
 utils/HtmlFonts.h      |    2 +-
 utils/HtmlOutputDev.cc |   87 +++++++++++++++++++++++++++---------------------
 utils/HtmlOutputDev.h  |    9 ++++-
 4 files changed, 62 insertions(+), 53 deletions(-)
 mode change 100644 => 100755 utils/HtmlFonts.cc
 mode change 100644 => 100755 utils/HtmlFonts.h
 mode change 100644 => 100755 utils/HtmlOutputDev.cc
 mode change 100644 => 100755 utils/HtmlOutputDev.h

diff --git a/utils/HtmlFonts.cc b/utils/HtmlFonts.cc
old mode 100644
new mode 100755
index e4af947..de9130d
--- a/utils/HtmlFonts.cc
+++ b/utils/HtmlFonts.cc
@@ -22,6 +22,7 @@
 // Copyright (C) 2008 Tomas Are Haavet <tomasare@gmail.com>
 // Copyright (C) 2010 OSSD CDAC Mumbai by Leena Chourey (leenac@cdacmumbai.in) and Onkar Potdar (onkar@cdacmumbai.in)
 // Copyright (C) 2011 Joshua Richardson <jric@chegg.com>
+// Copyright (C) 2011 Stephen Reichling <sreichling@chegg.com>
 //
 // To see a description of the changes please see the Changelog file that
 // came with your tarball or type make ChangeLog if you are building from git
@@ -239,11 +240,11 @@ GooString* HtmlFont::HtmlFilter(Unicode* u, int uLen) {
   for (int i = 0; i < uLen; ++i) {
     switch (u[i])
       { 
-	case '"': tmp->append("&quot;");  break;
+	case '"': tmp->append("&#34;");  break;
 	case '&': tmp->append("&amp;");  break;
 	case '<': tmp->append("&lt;");  break;
 	case '>': tmp->append("&gt;");  break;
-	case ' ': tmp->append( !xml && ( i+1 >= uLen || !tmp->getLength() || tmp->getChar( tmp->getLength()-1 ) == ' ' ) ? "&nbsp;" : " " );
+	case ' ': tmp->append( !xml && ( i+1 >= uLen || !tmp->getLength() || tmp->getChar( tmp->getLength()-1 ) == ' ' ) ? "&#160;" : " " );
 	          break;
 	default:  
 	  {
@@ -297,25 +298,17 @@ int HtmlFontAccu::AddFont(const HtmlFont& font){
 }
 
 // get CSS font name for font #i 
-GooString* HtmlFontAccu::getCSStyle(int i, GooString* content, int j){
+GooString* HtmlFontAccu::getCSStyle(GooString* content){
   GooString *tmp;
-  GooString *iStr=GooString::fromInt(i);
-  GooString *jStr=GooString::fromInt(j);
   
   if (!xml) {
-    tmp = new GooString("<span class=\"ft");
-    tmp->append(jStr);
-    tmp->append(iStr);
-    tmp->append("\">");
+    tmp = new GooString("<span>");
     tmp->append(content);
     tmp->append("</span>");
   } else {
     tmp = new GooString("");
     tmp->append(content);
   }
-
-  delete jStr;
-  delete iStr;
   return tmp;
 }
 
diff --git a/utils/HtmlFonts.h b/utils/HtmlFonts.h
old mode 100644
new mode 100755
index 607979c..032dbe8
--- a/utils/HtmlFonts.h
+++ b/utils/HtmlFonts.h
@@ -110,7 +110,7 @@ public:
   HtmlFont *Get(int i){
     return &(*accu)[i];
   } 
-  GooString* getCSStyle (int i,GooString* content, int j = 0);
+  GooString* getCSStyle (GooString* content);
   GooString* CSStyle(int i, int j = 0);
   int size() const {return accu->size();}
   
diff --git a/utils/HtmlOutputDev.cc b/utils/HtmlOutputDev.cc
old mode 100644
new mode 100755
index 3c87a5c..4e7d59c
--- a/utils/HtmlOutputDev.cc
+++ b/utils/HtmlOutputDev.cc
@@ -29,6 +29,7 @@
 // Copyright (C) 2010 Hib Eris <hib@hiberis.nl>
 // Copyright (C) 2010 OSSD CDAC Mumbai by Leena Chourey (leenac@cdacmumbai.in) and Onkar Potdar (onkar@cdacmumbai.in)
 // Copyright (C) 2011 Joshua Richardson <jric@chegg.com>
+// Copyright (C) 2011 Stephen Reichling <sreichling@chegg.com>
 //
 // To see a description of the changes please see the Changelog file that
 // came with your tarball or type make ChangeLog if you are building from git
@@ -630,13 +631,13 @@ void HtmlPage::coalesce() {
 					str1->size * sizeof(double));
       if (addSpace) {
 		  str1->text[str1->len] = 0x20;
-		  str1->htext->append(xml?" ":"&nbsp;");
+		  str1->htext->append(xml?" ":"&#160;");
 		  str1->xRight[str1->len] = str2->xMin;
 		  ++str1->len;
       }
       if (addLineBreak) {
 	  str1->text[str1->len] = '\n';
-	  str1->htext->append("<br>");
+	  str1->htext->append("<br/>");
 	  str1->xRight[str1->len] = str2->xMin;
 	  ++str1->len;
 	  str1->yMin = str2->yMin;
@@ -751,7 +752,7 @@ void HtmlPage::dumpAsXML(FILE* f,int page){
       fprintf(f,"<text top=\"%d\" left=\"%d\" ",xoutRound(tmp->yMin),xoutRound(tmp->xMin));
       fprintf(f,"width=\"%d\" height=\"%d\" ",xoutRound(tmp->xMax-tmp->xMin),xoutRound(tmp->yMax-tmp->yMin));
       fprintf(f,"font=\"%d\">", tmp->fontpos);
-      str1=fonts->getCSStyle(tmp->fontpos, str);
+      str1=fonts->getCSStyle(str);
       fputs(str1->getCString(),f);
       delete str;
       delete str1;
@@ -761,16 +762,11 @@ void HtmlPage::dumpAsXML(FILE* f,int page){
   fputs("</page>\n",f);
 }
 
-
-void HtmlPage::dumpComplex(FILE *file, int page){
-  FILE* pageFile;
+int HtmlPage::dumpComplexHeaders(FILE * const file, FILE *& pageFile, int page) {
   GooString* tmp;
   char* htmlEncoding;
 
-  if( firstPage == -1 ) firstPage = page; 
-  
-  if( !noframes )
-  {
+  if( !noframes ) {
       GooString* pgNum=GooString::fromInt(page);
       tmp = new GooString(DocName);
       if (!singleHtml){
@@ -784,36 +780,45 @@ void HtmlPage::dumpComplex(FILE *file, int page){
       if (!pageFile) {
 	  error(-1, "Couldn't open html file '%s'", tmp->getCString());
 	  delete tmp;
-	  return;
+      return 1;
       } 
 
       if (!singleHtml)
-          fprintf(pageFile,"%s\n<HTML>\n<HEAD>\n<TITLE>Page %d</TITLE>\n\n", DOCTYPE, page);
+      fprintf(pageFile,"%s\n<HTML xmlns=\"http://www.w3.org/1999/xhtml\" lang=\"\" xml:lang=\"\">\n<HEAD>\n<TITLE>Page %d</TITLE>\n\n",
+          DOCTYPE, page);
       else
-          fprintf(pageFile,"%s\n<HTML>\n<HEAD>\n<TITLE>%s</TITLE>\n\n", DOCTYPE, tmp->getCString());
+      fprintf(pageFile,"%s\n<HTML xmlns=\"http://www.w3.org/1999/xhtml\" lang=\"\" xml:lang=\"\">\n<HEAD>\n<TITLE>%s</TITLE>\n\n", DOCTYPE, tmp->getCString());
 
       delete tmp;
 
       htmlEncoding = HtmlOutputDev::mapEncodingToHtml
 	  (globalParams->getTextEncodingName());
       if (!singleHtml)
-          fprintf(pageFile, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n", htmlEncoding);
+      fprintf(pageFile, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\"/>\n", htmlEncoding);
       else
-          fprintf(pageFile, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n <br>\n", htmlEncoding);
-  }
-  else 
-  {
+      fprintf(pageFile, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\"/>\n <br/>\n",
+          htmlEncoding);
+  } else {
       pageFile = file;
       fprintf(pageFile,"<!-- Page %d -->\n", page);
       fprintf(pageFile,"<a name=\"%d\"></a>\n", page);
   } 
   
-  fprintf(pageFile,"<DIV style=\"position:relative;width:%d;height:%d;\">\n",
-	pageWidth, pageHeight);
+  return 0;
+}
+
+void HtmlPage::dumpComplex(FILE *file, int page){
+  FILE* pageFile;
+  GooString* tmp;
+
+  if( firstPage == -1 ) firstPage = page; 
+  
+  if (dumpComplexHeaders(file, pageFile, page)) { error(-1, "Couldn't write headers."); return; }
 
   tmp=basename(DocName);
    
   fputs("<STYLE type=\"text/css\">\n<!--\n",pageFile);
+  fputs("\tp {margin: 0; padding: 0;}",pageFile);
   for(int i=fontsPageMarker;i!=fonts->size();i++) {
     GooString *fontCSStyle;
     if (!singleHtml)
@@ -831,10 +836,13 @@ void HtmlPage::dumpComplex(FILE *file, int page){
       fputs("</HEAD>\n<BODY bgcolor=\"#A0A0A0\" vlink=\"blue\" link=\"blue\">\n",pageFile); 
   }
   
+  fprintf(pageFile,"<DIV id=\"page%d-div\" style=\"position:relative;width:%dpx;height:%dpx;\">\n",
+      page, pageWidth, pageHeight);
+
   if( !ignore ) 
   {
     fprintf(pageFile,
-	    "<IMG width=\"%d\" height=\"%d\" src=\"%s%03d.%s\" alt=\"background image\">\n",
+	    "<IMG width=\"%d\" height=\"%d\" src=\"%s%03d.%s\" alt=\"background image\"/>\n",
 	    pageWidth, pageHeight, tmp->getCString(), 
 		(page-firstPage+1), imgExt->getCString());
   }
@@ -846,22 +854,23 @@ void HtmlPage::dumpComplex(FILE *file, int page){
     if (tmp1->htext){
       str=new GooString(tmp1->htext);
       fprintf(pageFile,
-	      "<DIV style=\"position:absolute;top:%d;left:%d\">",
+	      "<P style=\"position:absolute;top:%dpx;left:%dpx;white-space:nowrap\" class=\"ft",
 	      xoutRound(tmp1->yMin),
 	      xoutRound(tmp1->xMin));
-      fputs("<nobr>",pageFile); 
       if (!singleHtml)
-          str1=fonts->getCSStyle(tmp1->fontpos, str);
+          fputc('0', pageFile);
       else
-          str1=fonts->getCSStyle(tmp1->fontpos, str, page);
+          fprintf(pageFile, "%d", page);
+      fprintf(pageFile,"%d\">",tmp1->fontpos);
+      str1=fonts->getCSStyle(str);
       fputs(str1->getCString(),pageFile);
       delete str;      
       delete str1;
-      fputs("</nobr></DIV>\n",pageFile);
+      fputs("</P>\n",pageFile);
     }
   }
 
-  fputs("</DIV>\n", pageFile);
+  fputs("</DIV>\n", pageFile); // page%d-div
   
   if( !noframes )
   {
@@ -885,7 +894,7 @@ void HtmlPage::dump(FILE *f, int pageNum)
     int listlen=HtmlOutputDev::imgList->getLength();
     for (int i = 0; i < listlen; i++) {
       GooString *fName= (GooString *)HtmlOutputDev::imgList->del(0);
-      fprintf(f,"<IMG src=\"%s\"><br>\n",fName->getCString());
+      fprintf(f,"<IMG src=\"%s\"/><br/>\n",fName->getCString());
       delete fName;
     }
     HtmlOutputDev::imgNum=1;
@@ -896,7 +905,7 @@ void HtmlPage::dump(FILE *f, int pageNum)
 		str=new GooString(tmp->htext); 
 		fputs(str->getCString(),f);
 		delete str;      
-		fputs("<br>\n",f);  
+		fputs("<br/>\n",f);
       }
     }
 	fputs("<hr>\n",f);  
@@ -963,7 +972,7 @@ GooString* HtmlMetaVar::toString()
     result->append(name);
     result->append("\" content=\"");
     result->append(content);
-    result->append("\">"); 
+    result->append("\"/>");
     return result;
 }
 
@@ -1009,7 +1018,7 @@ void HtmlOutputDev::doFrame(int firstPage){
   fputs("\n<HEAD>",fContentsFrame);
   fprintf(fContentsFrame,"\n<TITLE>%s</TITLE>",docTitle->getCString());
   htmlEncoding = mapEncodingToHtml(globalParams->getTextEncodingName());
-  fprintf(fContentsFrame, "\n<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n", htmlEncoding);
+  fprintf(fContentsFrame, "\n<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\"/>\n", htmlEncoding);
   dumpMetaVars(fContentsFrame);
   fprintf(fContentsFrame, "</HEAD>\n");
   fputs("<FRAMESET cols=\"100,*\">\n",fContentsFrame);
@@ -1079,12 +1088,13 @@ HtmlOutputDev::HtmlOutputDev(char *fileName, char *title,
          }
          delete left;
          fputs(DOCTYPE, fContentsFrame);
-         fputs("<HTML>\n<HEAD>\n<TITLE></TITLE>\n</HEAD>\n<BODY>\n",fContentsFrame);
+         fputs("<HTML xmlns=\"http://www.w3.org/1999/xhtml\" lang=\"\" xml:lang=\"\">\n<HEAD>\n<TITLE></TITLE>\n</HEAD>\n<BODY>\n",
+             fContentsFrame);
 
          if (doOutline)
          {
              GooString *str = basename(Docname);
-             fprintf(fContentsFrame, "<A href=\"%s%s\" target=\"contents\">Outline</a><br>", str->getCString(), complexMode ? "-outline.html" : "s.html#outline");
+             fprintf(fContentsFrame, "<A href=\"%s%s\" target=\"contents\">Outline</a><br/>", str->getCString(), complexMode ? "-outline.html" : "s.html#outline");
              delete str;
          }
      }
@@ -1128,10 +1138,10 @@ HtmlOutputDev::HtmlOutputDev(char *fileName, char *title,
     } 
     else 
     {
-      fprintf(page,"%s\n<HTML>\n<HEAD>\n<TITLE>%s</TITLE>\n",
+      fprintf(page,"%s\n<HTML xmlns=\"http://www.w3.org/1999/xhtml\" lang=\"\" xml:lang=\"\">\n<HEAD>\n<TITLE>%s</TITLE>\n",
 	      DOCTYPE, docTitle->getCString());
       
-      fprintf(page, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\">\n", htmlEncoding);
+      fprintf(page, "<META http-equiv=\"Content-Type\" content=\"text/html; charset=%s\"/>\n", htmlEncoding);
       
       dumpMetaVars(page);
       fprintf(page,"</HEAD>\n");
@@ -1197,7 +1207,7 @@ void HtmlOutputDev::startPage(int pageNum, GfxState *state) {
 		fprintf(fContentsFrame,"<A href=\"%s-%d.html\"",str->getCString(),pageNum);
       else 
 		fprintf(fContentsFrame,"<A href=\"%ss.html#%d\"",str->getCString(),pageNum);
-      fprintf(fContentsFrame," target=\"contents\" >Page %d</a><br>\n",pageNum);
+      fprintf(fContentsFrame," target=\"contents\" >Page %d</a><br/>\n",pageNum);
     }
   }
 
@@ -1226,7 +1236,7 @@ void HtmlOutputDev::endPage() {
   maxPageWidth = pages->pageWidth;
   maxPageHeight = pages->pageHeight;
   
-  //if(!noframes&&!xml) fputs("<br>\n", fContentsFrame);
+  //if(!noframes&&!xml) fputs("<br/>\n", fContentsFrame);
   if(!stout && !globalParams->getErrQuiet()) printf("Page-%d\n",(pageNum));
 }
 
@@ -1604,7 +1614,8 @@ GBool HtmlOutputDev::dumpDocOutline(Catalog* catalog)
 				return gFalse;
 			delete str;
 			bClose = gTrue;
-     		fputs("<HTML>\n<HEAD>\n<TITLE>Document Outline</TITLE>\n</HEAD>\n<BODY>\n", output);
+     		fputs("<HTML xmlns=\"http://www.w3.org/1999/xhtml\" lang=\"\" xml:lang=\"\">\n<HEAD>\n<TITLE>Document Outline</TITLE>\n</HEAD>\n<BODY>\n",
+     		    output);
 		}
 	}
  
diff --git a/utils/HtmlOutputDev.h b/utils/HtmlOutputDev.h
old mode 100644
new mode 100755
index 47d274f..bc6ac80
--- a/utils/HtmlOutputDev.h
+++ b/utils/HtmlOutputDev.h
@@ -20,6 +20,7 @@
 // Copyright (C) 2009 Kovid Goyal <kovid@kovidgoyal.net>
 // Copyright (C) 2010 Hib Eris <hib@hiberis.nl>
 // Copyright (C) 2011 Joshua Richardson <jric@chegg.com>
+// Copyright (C) 2011 Stephen Reichling <sreichling@chegg.com>
 //
 // To see a description of the changes please see the Changelog file that
 // came with your tarball or type make ChangeLog if you are building from git
@@ -54,8 +55,11 @@
 
 #define xoutRound(x) ((int)(x + 0.5))
 
-#define DOCTYPE "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\">"
-#define DOCTYPE_FRAMES "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Frameset//EN\"\n\"http://www.w3.org/TR/html4/frameset.dtd\">"
+//#define DOCTYPE "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\">"
+//#define DOCTYPE_FRAMES "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Frameset//EN\"\n\"http://www.w3.org/TR/html4/frameset.dtd\">"
+#define DOCTYPE "<!DOCTYPE html>"
+#define DOCTYPE_FRAMES "<!DOCTYPE html>"
+
 
 class GfxState;
 class GooString;
@@ -178,6 +182,7 @@ private:
   void setDocName(char* fname);
   void dumpAsXML(FILE* f,int page);
   void dumpComplex(FILE* f, int page);
+  int dumpComplexHeaders(FILE * const file, FILE *& pageFile, int page);
 
   // marks the position of the fonts that belong to current page (for noframes)
   int fontsPageMarker; 
-- 
1.7.4.1