From f9324b3f2bf64574bf20e37686624b8600c3ecc4 Mon Sep 17 00:00:00 2001
From: Corentin Allemand <corentin.allemand@aquafadas.com>
Date: Wed, 31 Oct 2012 17:18:37 +0100
Subject: [PATCH 1/2] Add with media params

---
 utils/HtmlOutputDev.cc |    4 +++-
 utils/pdftohtml.cc     |    7 +++++--
 2 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/utils/HtmlOutputDev.cc b/utils/HtmlOutputDev.cc
index e4bd0b1..7d60975 100644
--- a/utils/HtmlOutputDev.cc
+++ b/utils/HtmlOutputDev.cc
@@ -97,6 +97,7 @@ extern GBool stout;
 extern GBool xml;
 extern GBool showHidden;
 extern GBool noMerge;
+extern GBool withMedia;
 
 extern double wordBreakThreshold;
 
@@ -877,7 +878,8 @@ void HtmlPage::dumpComplex(FILE *file, int page){
   fprintf(pageFile,"<div id=\"page%d-div\" style=\"position:relative;width:%dpx;height:%dpx;\">\n",
       page, pageWidth, pageHeight);
 
-  if( !ignore ) 
+
+  if( !ignore && !withMedia)
   {
     fprintf(pageFile,
 	    "<img width=\"%d\" height=\"%d\" src=\"%s%03d.%s\" alt=\"background image\"/>\n",
diff --git a/utils/pdftohtml.cc b/utils/pdftohtml.cc
index 97372be..72e784f 100644
--- a/utils/pdftohtml.cc
+++ b/utils/pdftohtml.cc
@@ -84,6 +84,7 @@ double wordBreakThreshold=10;  // 10%, below converted into a coefficient - 0.1
 GBool showHidden = gFalse;
 GBool noMerge = gFalse;
 GBool fontFullName = gFalse;
+GBool withMedia = gFalse;
 static char ownerPassword[33] = "";
 static char userPassword[33] = "";
 static GBool printVersion = gFalse;
@@ -141,7 +142,9 @@ static const ArgDesc argDesc[] = {
   {"-wbt",    argFP,    &wordBreakThreshold, 0,
    "word break threshold (default 10 percent)"},
   {"-fontfullname", argFlag, &fontFullName, 0,
-   "outputs font full name"},   
+   "outputs font full name"},
+  {"-withmedia", argFlag, &withMedia, 0,
+   "extract all media"},
   {NULL}
 };
 
@@ -383,7 +386,7 @@ int main(int argc, char *argv[]) {
     htmlOut->dumpDocOutline(doc);
   }
   
-  if ((complexMode || singleHtml) && !xml && !ignore) {
+  if ((complexMode || singleHtml) && !xml && !ignore && !withMedia) {
 #ifdef HAVE_SPLASH
     GooString *imgFileName = NULL;
     // White paper color
-- 
1.7.10.2 (Apple Git-33)


From 1d677d4f2c4106bc73e25be5294161b02bbd3810 Mon Sep 17 00:00:00 2001
From: Corentin Allemand <corentin.allemand@aquafadas.com>
Date: Wed, 31 Oct 2012 18:16:19 +0100
Subject: [PATCH 2/2] Extract images with "-withMedia" parameter

---
 utils/HtmlOutputDev.cc |   27 +++++++++++++++++++++++++--
 1 file changed, 25 insertions(+), 2 deletions(-)

diff --git a/utils/HtmlOutputDev.cc b/utils/HtmlOutputDev.cc
index 7d60975..916ec91 100644
--- a/utils/HtmlOutputDev.cc
+++ b/utils/HtmlOutputDev.cc
@@ -889,6 +889,26 @@ void HtmlPage::dumpComplex(FILE *file, int page){
   
   delete tmp;
   
+  if(!ignore && withMedia){
+    int listlen=imgList->getLength();
+    GooString *imgfname;
+    for (int i = 0; i < listlen; i++) {
+      HtmlImage *img = (HtmlImage*)imgList->del(0);
+      imgfname=basename(img->fName);
+      fprintf(pageFile,
+     	      "<div style=\"position:absolute;top:%dpx;left:%dpx;width:%dpx;height:%dpx\">",
+       	      xoutRound(img->yMin),
+       	      xoutRound(img->xMin),
+       	      xoutRound(img->xMax-img->xMin),
+       	      xoutRound(img->yMax-img->yMin));
+
+      fprintf(pageFile,"<img width=\"%d\" height=\"%d\" ",xoutRound(img->xMax-img->xMin),xoutRound(img->yMax-img->yMin));
+      fprintf(pageFile,"src=\"%s\"/>",imgfname->getCString());
+      fputs("</div>\n", pageFile);
+      delete img;
+    }
+  }
+
   for(HtmlString *tmp1=yxStrings;tmp1;tmp1=tmp1->yxNext){
     if (tmp1->htext){
       fprintf(pageFile,
@@ -906,6 +926,9 @@ void HtmlPage::dumpComplex(FILE *file, int page){
     }
   }
 
+
+
+
   fputs("</div>\n", pageFile);
   
   if( !noframes )
@@ -1466,7 +1489,7 @@ void HtmlOutputDev::drawImageMask(GfxState *state, Object *ref, Stream *str,
 				  int width, int height, GBool invert,
 				  GBool interpolate, GBool inlineImg) {
 
-  if (ignore||(complexMode && !xml)) {
+  if (ignore|| !(complexMode && withMedia)) {
     OutputDev::drawImageMask(state, ref, str, width, height, invert, interpolate, inlineImg);
     return;
   }
@@ -1488,7 +1511,7 @@ void HtmlOutputDev::drawImage(GfxState *state, Object *ref, Stream *str,
 			      int width, int height, GfxImageColorMap *colorMap,
 			      GBool interpolate, int *maskColors, GBool inlineImg) {
 
-  if (ignore||(complexMode && !xml)) {
+  if (ignore || !(complexMode && withMedia)) {
     OutputDev::drawImage(state, ref, str, width, height, colorMap, interpolate,
 			 maskColors, inlineImg);
     return;
-- 
1.7.10.2 (Apple Git-33)