Merge remote-tracking branch 'upstream/master'

CMakeLists.txt

@@ -7,7 +7,7 @@ cmake_minimum_required(VERSION 2.6.0 FATAL_ERROR)
 
 include_directories(${CMAKE_SOURCE_DIR}/src)
 
-set(PDF2HTMLEX_VERSION "0.7")
+set(PDF2HTMLEX_VERSION "0.8")
 set(ARCHIVE_NAME pdf2htmlex-${PDF2HTMLEX_VERSION})
 add_custom_target(dist
     COMMAND git archive --prefix=${ARCHIVE_NAME}/ HEAD
@@ -43,7 +43,7 @@ if(FONTFORGE_FOUND)
     link_directories(${FONTFORGE_LIBRARY_DIRS})
     set(PDF2HTMLEX_LIBS ${PDF2HTMLEX_LIBS} ${FONTFORGE_LIBRARIES})
 else()
-    message("Trying to locate fontforge...")
+    message("Trying to locate old versions of fontforge...")
     find_path(FF_INCLUDE_PATH fontforge/fontforge.h)
     if(FF_INCLUDE_PATH)
         message("Found fontforge.h: ${FF_INCLUDE_PATH}/fontforge/fontforge.h")
@@ -61,6 +61,14 @@ else()
     else()
         message(FATAL_ERROR "Error: cannot locate fontforge.h")
     endif()
+    find_path(FF_CONFIG_INCLUDE_PATH config.h PATHS
+      ${FONTFORGE_INCLUDE_DIRS} NO_DEFAULT_PATH)
+    if(FF_CONFIG_INCLUDE_PATH)
+    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -include ${FF_CONFIG_INCLUDE_PATH}/config.h")
+    message("Found config.h: ${FF_CONFIG_INCLUDE_PATH}/config.h")
+    else()
+    message("Cannot locate config.h for fontforge")
+    endif()
 
     macro(wl_find_library LIB_NAME RESULT)
         unset(${RESULT})
@@ -98,14 +106,6 @@ else()
     set(PDF2HTMLEX_LIBS ${PDF2HTMLEX_LIBS} ${PYTHON_LIBRARIES})
 endif()
 
-find_path(FF_CONFIG_INCLUDE_PATH config.h PATHS
-  ${FONTFORGE_INCLUDE_DIRS} NO_DEFAULT_PATH)
-if(FF_CONFIG_INCLUDE_PATH)
-set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -include ${FF_CONFIG_INCLUDE_PATH}/config.h")
-message("Found config.h: ${FF_CONFIG_INCLUDE_PATH}/config.h")
-else()
-message("Cannot locate config.h for fontforge")
-endif()
 
 # debug build flags (overwrite default cmake debug flags)
 set(CMAKE_C_FLAGS_DEBUG "-ggdb")

ChangeLog

@@ -1,8 +1,13 @@
-Latest v0.7
+Latest v0.8
+
+v0.7
+2013.03.01
 
 * Process outline
 * Fix build with poppler
 * Many code cleaning jobs [John Hewson]
+* Experimental printing support
+* Lots of code refinements
 
 v0.6
 2013.01.26

README.md

@@ -9,8 +9,8 @@ A beautiful demo is worth a thousand words:
 - **Scientific Paper**: [Default](http://coolwanglu.github.com/pdf2htmlEX/demo/demo.html) / [MediaFire](http://www.mediafire.com/view/?6po429kz9czcga2) / [Original](http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.148.349&rep=rep1&type=pdf)
 - **Full Circle Magazine**: [Default](http://coolwanglu.github.com/pdf2htmlEX/demo/issue65_en.html) / [MediaFire](http://www.mediafire.com/view/?6hxmt94k2vppnpb) / [Original](http://dl.fullcirclemagazine.org/issue65_en.pdf)   <sub>The 1st link might be slow</sub>
 - **Chinese**: [Default](http://coolwanglu.github.com/pdf2htmlEX/demo/chn.html) / [MediaFire](http://www.mediafire.com/view/?6550ldag9w0uuq3) / [Original](http://files.cnblogs.com/phphuaibei/git%E6%90%AD%E5%BB%BA.pdf)
-- Try your own files on [MediaFire](http://www.mediafire.com), which uses pdf2htmlEX for its PDF preview feature.
+- [Try your own files](https://github.com/coolwanglu/pdf2htmlEX/wiki/UploadDemo)
-  
+ 
 ## Introduction
 
 pdf2htmlEX renders PDF files in HTML, utilizing modern Web technologies.
@@ -29,6 +29,7 @@ The generated HTML file is static, Javascript is not required.
   - Correct font & position & styles
   - Proper reencoding
   - Generated HTML file is of similar size as the original (uncompressed) PDF file
+  - Fallback (image + hidden text) - better accuracy and compatibility
 * Output modes
   - Normal HTML
   - All-in-one HTML - portable & easy to share
@@ -36,6 +37,7 @@ The generated HTML file is static, Javascript is not required.
 * More PDF stuffs that you love
   - Links
   - Outline
+  - Printing (experimental)
 
 [Full list](https://github.com/coolwanglu/pdf2htmlEX/wiki/Feature-List)   
 [Compare with others](https://github.com/coolwanglu/pdf2htmlEX/wiki/Comparison)
@@ -83,15 +85,15 @@ Thanks to all packagers!
 ## Usage
 
     pdf2htmlEX /path/to/foobar.pdf
-    pdf2htmlEX --help
+
-    man pdf2htmlEX
+[Quick Start](https://github.com/coolwanglu/pdf2htmlEX/wiki/QuickStart)
 
 ## FAQ
 
 * [Troubleshooting compilation errors](https://github.com/coolwanglu/pdf2htmlEX/wiki/FAQ#wiki-compile)
 * [How can I help](https://github.com/coolwanglu/pdf2htmlEX/wiki/FAQ#wiki-help)
 * [I want more features](https://github.com/coolwanglu/pdf2htmlEX/wiki/FAQ#wiki-feature_commission)
-* [More about pdf2htmlEX](https://github.com/coolwanglu/pdf2htmlEX/wiki/)
+* [More...](https://github.com/coolwanglu/pdf2htmlEX/wiki/FAQ)
 
 ## LICENSE
 
@@ -114,7 +116,7 @@ pdf2htmlEX is maintained by one person in spare time, and it needs your help!
 
 * Lu Wang <coolwanglu@gmail.com>
   * For personal enquiries only
-  * Accepting messages in **Chinese**, **English** or **Japanese**.
+  * Accepting messages in **中文**, **English** or **日本語**.
 
 ## Acknowledge
 

TODO

@@ -1,19 +1,15 @@
-clean css class names
-print css for draw/link/image...
-
 == Future: ==
 
 Too difficult/complicated to implement:
  - integrate splash/cairo
- - naive support for image/drawing (SVG?)
+ - naive image/drawing (SVG?)
  - type 3 fonts (convert to SVG fonts?)
  - reflowable text/combine lines/unwrapping
- - Printing
+ - multi-thread
 
 Not enough motivated/Lazy
  - argument auto-completion
  - use absolute positioning for long whitespace
- - color invert
  - detect duplicate base fonts when embedding
  - disable selection if we know unicode is wrong
  - check if we can add information to the font, and let browsers show ligatures automatically
@@ -23,6 +19,7 @@ Not enough motivated/Lazy
  - merge sub/sup into one line
  - precise link dest: zoom
  - multiple charcode mapped to a same glyph
- - don't dump image when there is nothing
+ - don't dump image when it is empty
  - minimum line width of css drawing
  - ajax in pdf2htmlEX for separated pages
+ - separate classes for annotations (such that we don't have to hide all css drawings for printing)

debian/changelog

@@ -1,3 +1,16 @@
+pdf2htmlex (0.8-1~git201303011406r3bc73-0ubuntu1) quantal; urgency=low
+
+  * Experimental printing support
+  * New version
+
+ -- WANG Lu <coolwanglu@gmail.com>  Fri, 01 Mar 2013 14:06:42 +0800
+
+pdf2htmlex (0.7-1~git201302282259r3bc73-0ubuntu1) quantal; urgency=low
+
+  * suggests ttfautohint
+
+ -- WANG Lu <coolwanglu@gmail.com>  Thu, 28 Feb 2013 22:59:45 +0800
+
 pdf2htmlex (0.7-1~git201302271054r3bc73-0ubuntu1) precise; urgency=low
 
   * Packaging for 12.04

debian/control

@@ -9,5 +9,6 @@ Homepage: http://github.com/coolwanglu/pdf2htmlEX
 Package: pdf2htmlex
 Architecture: any
 Depends: ${shlibs:Depends}, ${misc:Depends}, libpoppler27 (>= 0.20.3) | libpoppler28, libpng12-0, libfontforge1
+Suggests: ttfautohint
 Description: Converts PDF to HTML without losing format
  pdf2htmlEX converts PDF to HTML while retaining text, format & style as much as possible

pdf2htmlEX.1.in

@@ -87,6 +87,10 @@ Specify the filename of the generated outline file, if not embedded.
 
 If it's empty, the file name will be determined automatically.
 
+.TP
+.B --fallback <0|1> (Deafult: 0)
+Output in fallback mode, for better accuracy and browser compatibility, but the size becomes larger.
+
 .TP
 .B --process-nontext <0|1> (Default: 1)
 Whether to process non-text objects (as images)

share/base.css.in

@@ -72,6 +72,9 @@
     overflow:visible;
     background-color:transparent;
   }
+  .@CSS_CSS_DRAW_CN@ {
+    display:none;
+  }
 }
 /* Part 2: Page Elements: Modify with caution
  * The followings are base classes, which are meant to be override by PDF specific classes
@@ -113,6 +116,17 @@
 .@CSS_PAGE_CONTENT_BOX_CN@.opened { /* used by pdf2htmlEX.js, to show/hide pages */
   display:block;
 }
+.@CSS_BACKGROUND_IMAGE_CN@ {
+  position:absolute;
+  left:0;
+  top:0;
+  width:100%;
+  height:100%;
+  -ms-user-select:none;
+  -moz-user-select:none;
+  -webkit-user-select:none;
+  user-select:none;
+}
 @media print {
   .@CSS_PAGE_DECORATION_CN@ {
     margin:0;
@@ -150,11 +164,12 @@ span { /* text blocks within a line */
   color:transparent;
   z-index:-1;
 }
+/* selection background should not be opaque, for fallback mode */
 ::selection{
-  background: rgba(127,255,255,1);
+  background: rgba(127,255,255,0.4);
 }
 ::-moz-selection{
-  background: rgba(127,255,255,1);
+  background: rgba(127,255,255,0.4);
 }
 .@CSS_PAGE_DATA_CN@ { /* info for Javascript */
   display:none;

share/pdf2htmlEX.js.in

@@ -15,6 +15,7 @@ var pdf2htmlEX = (function(){
     page_decoration  : '@CSS_PAGE_DECORATION_CN@',
     page_content_box : '@CSS_PAGE_CONTENT_BOX_CN@',
     page_data        : '@CSS_PAGE_DATA_CN@',
+    background_image : '@CSS_BACKGROUND_IMAGE_CN@',
     link             : '@CSS_LINK_CN@',
     __dummy__        : 'no comma'
   };
@@ -126,11 +127,12 @@ var pdf2htmlEX = (function(){
       this.outline = $('#'+this.outline_id);
       this.container = $('#'+this.container_id);
 
-      // need a better design
+      // Open the outline if nonempty
       if(this.outline.children().length > 0) { 
         this.outline.addClass('opened');
       }
       
+      // collect pages
       var new_pages = new Array();
       var pl= $('.'+CSS_CLASS_NAMES['page_frame'], this.container);
       /* don't use for(..in..) */
@@ -140,14 +142,18 @@ var pdf2htmlEX = (function(){
       }
       this.pages = new_pages;
 
+      // register schedule rendering
       var _ = this;
       this.container.scroll(function(){ _.schedule_render(); });
 
       //this.zoom_fixer();
       
-      // used by outline/annot_link etc
+      // handle links
       this.container.add(this.outline).on('click', '.'+CSS_CLASS_NAMES['link'], this, this.link_handler);
 
+      // disable background image draging
+      $('.'+CSS_CLASS_NAMES['background_image'], this.container).on('dragstart', function(e){return false;});
+
       this.render();
     },
     pre_hide_pages : function() {

src/BackgroundRenderer/SplashBackgroundRenderer.cc

@@ -20,17 +20,19 @@ void SplashBackgroundRenderer::drawChar(GfxState *state, double x, double y,
   CharCode code, int nBytes, Unicode *u, int uLen)
 {
     // draw characters as image when
-    // - there is special filling method
+    // - in fallback mode
+    // - OR there is special filling method
     // - OR using a writing mode font
     // - OR using a Type 3 font
-    if(( (state->getFont()) 
+    if((param->fallback)
-         && ( (state->getFont()->getWMode())
+       || ( (state->getFont()) 
-              || (state->getFont()->getType() == fontType3)
+            && ( (state->getFont()->getWMode())
-            )
+                 || (state->getFont()->getType() == fontType3)
-       )
+               )
+          )
       )
     {
-        SplashOutputDev::drawChar(state,x,y,dx,dy,originX,originY,code, nBytes, u, uLen);
+        SplashOutputDev::drawChar(state,x,y,dx,dy,originX,originY,code,nBytes,u,uLen);
     }
 }
 
@@ -42,7 +44,7 @@ void SplashBackgroundRenderer::render_page(PDFDoc * doc, int pageno, const strin
 {
     doc->displayPage(this, pageno, param->h_dpi, param->v_dpi,
             0, 
-            (param->use_cropbox == 0), 
+            (!(param->use_cropbox)),
             false, false,
             nullptr, nullptr, &annot_cb, nullptr);
 

src/HTMLRenderer/HTMLRenderer.h

@@ -241,6 +241,9 @@ protected:
     double text_scale_factor1;
     double text_scale_factor2;
 
+    // 1px on screen should be printed as print_scale()pt
+    double print_scale (void) const { return 96.0 / DEFAULT_DPI / text_zoom_factor(); }
+
 
     ////////////////////////////////////////////////////
     // states
@@ -294,6 +297,7 @@ protected:
     RiseManager                        rise_manager;
     LeftManager                        left_manager;
     ////////////////////////////////////////////////
+    BGImageSizeManager         bgimage_size_manager;
 
     // optimize for web
     // we try to render the final font size directly

src/HTMLRenderer/TextLineBuffer.cc

@@ -275,7 +275,7 @@ int HTMLRenderer::TextLineBuffer::State::diff(const State & s) const
 
 // the order should be the same as in the enum
 const char * const HTMLRenderer::TextLineBuffer::State::css_class_names [] = {
-    CSS::FONT_NAME_CN,
+    CSS::FONT_FAMILY_CN,
     CSS::FONT_SIZE_CN,
     CSS::FILL_COLOR_CN,
     CSS::STROKE_COLOR_CN,

src/HTMLRenderer/font.cc

@@ -727,7 +727,7 @@ void HTMLRenderer::export_remote_font(const FontInfo & info, const string & suff
     }
 
     f_css.fs << "@font-face{"
-             << "font-family:" << CSS::FONT_NAME_CN << info.id << ";"
+             << "font-family:" << CSS::FONT_FAMILY_CN << info.id << ";"
              << "src:url(";
 
     {
@@ -749,8 +749,8 @@ void HTMLRenderer::export_remote_font(const FontInfo & info, const string & suff
     f_css.fs << ")"
              << "format(\"" << format << "\");"
              << "}" // end of @font-face
-             << "." << CSS::FONT_NAME_CN << info.id << "{"
+             << "." << CSS::FONT_FAMILY_CN << info.id << "{"
-             << "font-family:" << CSS::FONT_NAME_CN << info.id << ";"
+             << "font-family:" << CSS::FONT_FAMILY_CN << info.id << ";"
              << "line-height:" << round(info.ascent - info.descent) << ";"
              << "font-style:normal;"
              << "font-weight:normal;"
@@ -772,12 +772,12 @@ static string general_font_family(GfxFont * font)
 // TODO: this function is called when some font is unable to process, may use the name there as a hint
 void HTMLRenderer::export_remote_default_font(long long fn_id) 
 {
-    f_css.fs << "." << CSS::FONT_NAME_CN << fn_id << "{font-family:sans-serif;visibility:hidden;}" << endl;
+    f_css.fs << "." << CSS::FONT_FAMILY_CN << fn_id << "{font-family:sans-serif;visibility:hidden;}" << endl;
 }
 
 void HTMLRenderer::export_local_font(const FontInfo & info, GfxFont * font, const string & original_font_name, const string & cssfont) 
 {
-    f_css.fs << "." << CSS::FONT_NAME_CN << info.id << "{";
+    f_css.fs << "." << CSS::FONT_FAMILY_CN << info.id << "{";
     f_css.fs << "font-family:" << ((cssfont == "") ? (original_font_name + "," + general_font_family(font)) : cssfont) << ";";
 
     string fn = original_font_name;

src/HTMLRenderer/general.cc

@@ -111,7 +111,7 @@ void HTMLRenderer::process(PDFDoc *doc)
 
         if(param->process_nontext)
         {
-            auto fn = str_fmt("%s/p%x.png", (param->single_html ? param->tmp_dir : param->dest_dir).c_str(), i);
+            auto fn = str_fmt("%s/bg%x.png", (param->single_html ? param->tmp_dir : param->dest_dir).c_str(), i);
             if(param->single_html)
                 tmp_files.add((char*)fn);
 
@@ -121,7 +121,7 @@ void HTMLRenderer::process(PDFDoc *doc)
         doc->displayPage(this, i, 
                 text_zoom_factor() * DEFAULT_DPI, text_zoom_factor() * DEFAULT_DPI,
                 0, 
-                (param->use_cropbox == 0), 
+                (!(param->use_cropbox)),
                 false, false,
                 nullptr, nullptr, nullptr, nullptr);
 
@@ -173,35 +173,29 @@ void HTMLRenderer::startPage(int pageNum, GfxState *state, XRef * xref)
         << "<div id=\"" << CSS::PAGE_FRAME_CN << pageNum 
             << "\" class=\"" << CSS::PAGE_FRAME_CN
             << "\" data-page-no=\"" << pageNum << "\">"
-        << "<div class=\"" << CSS::PAGE_CONTENT_BOX_CN << "\" style=\"";
+        << "<div class=\"" << CSS::PAGE_CONTENT_BOX_CN 
+            << " " << CSS::PAGE_CONTENT_BOX_CN << pageNum
+            << "\">";
 
     if(param->process_nontext)
     {
-        f_pages.fs << "background-image:url(";
+        f_pages.fs << "<img class=\"" << CSS::BACKGROUND_IMAGE_CN 
-
+            << "\" alt=\"\" src=\"";
+        if(param->single_html)
         {
-            if(param->single_html)
+            auto path = str_fmt("%s/bg%x.png", param->tmp_dir.c_str(), pageNum);
-            {
+            ifstream fin((char*)path, ifstream::binary);
-                auto path = str_fmt("%s/p%x.png", param->tmp_dir.c_str(), pageNum);
+            if(!fin)
-                ifstream fin((char*)path, ifstream::binary);
+                throw string("Cannot read background image ") + (char*)path;
-                if(!fin)
+            f_pages.fs << "data:image/png;base64," << base64stream(fin);
-                    throw string("Cannot read background image ") + (char*)path;
-                f_pages.fs << "'data:image/png;base64," << base64stream(fin) << "'";
-            }
-            else
-            {
-                f_pages.fs << str_fmt("p%x.png", pageNum);
-            }
         }
-
+        else
-        // TODO print css
+        {
-        f_pages.fs << ");background-position:0 0;background-size:" 
+            f_pages.fs << str_fmt("bg%x.png", pageNum);
-            << state->getPageWidth() << "px " 
+        }
-            << state->getPageHeight() << "px;background-repeat:no-repeat;";
+        f_pages.fs << "\"/>";
     }
 
-    f_pages.fs << "\">";
-
     reset_state();
 }
 
@@ -462,22 +456,24 @@ void HTMLRenderer::dump_css (void)
     width_manager           .dump_css(f_css.fs);
     rise_manager            .dump_css(f_css.fs);
     left_manager            .dump_css(f_css.fs);
+    bgimage_size_manager.dump_css(f_css.fs);
     
     // print css
-    double print_scale = 96.0 / DEFAULT_DPI / text_zoom_factor();
+    double ps = print_scale();
     f_css.fs << CSS::PRINT_ONLY << "{" << endl;
-    transform_matrix_manager.dump_print_css(f_css.fs, print_scale);
+    transform_matrix_manager.dump_print_css(f_css.fs, ps);
-    letter_space_manager    .dump_print_css(f_css.fs, print_scale);
+    letter_space_manager    .dump_print_css(f_css.fs, ps);
-    stroke_color_manager    .dump_print_css(f_css.fs, print_scale);
+    stroke_color_manager    .dump_print_css(f_css.fs, ps);
-    word_space_manager      .dump_print_css(f_css.fs, print_scale);
+    word_space_manager      .dump_print_css(f_css.fs, ps);
-    whitespace_manager      .dump_print_css(f_css.fs, print_scale);
+    whitespace_manager      .dump_print_css(f_css.fs, ps);
-    fill_color_manager      .dump_print_css(f_css.fs, print_scale);
+    fill_color_manager      .dump_print_css(f_css.fs, ps);
-    font_size_manager       .dump_print_css(f_css.fs, print_scale);
+    font_size_manager       .dump_print_css(f_css.fs, ps);
-    bottom_manager          .dump_print_css(f_css.fs, print_scale);
+    bottom_manager          .dump_print_css(f_css.fs, ps);
-    height_manager          .dump_print_css(f_css.fs, print_scale);
+    height_manager          .dump_print_css(f_css.fs, ps);
-    width_manager           .dump_print_css(f_css.fs, print_scale);
+    width_manager           .dump_print_css(f_css.fs, ps);
-    rise_manager            .dump_print_css(f_css.fs, print_scale);
+    rise_manager            .dump_print_css(f_css.fs, ps);
-    left_manager            .dump_print_css(f_css.fs, print_scale);
+    left_manager            .dump_print_css(f_css.fs, ps);
+    bgimage_size_manager.dump_print_css(f_css.fs, ps);
     f_css.fs << "}" << endl;
 }
 

src/HTMLRenderer/state.cc

@@ -346,7 +346,7 @@ void HTMLRenderer::check_state_change(GfxState * state)
     }
 
     // fill color
-    if(all_changed || fill_color_changed)
+    if((!(param->fallback)) && (all_changed || fill_color_changed))
     {
         // * PDF Spec. Table 106 –  Text rendering modes
         static const char FILL[8]   = { true, false, true, false, true, false, true, false };
@@ -369,7 +369,7 @@ void HTMLRenderer::check_state_change(GfxState * state)
     }
 
     // stroke color
-    if(all_changed || stroke_color_changed)
+    if((!(param->fallback)) && (all_changed || stroke_color_changed))
     {
         // * PDF Spec. Table 106 –  Text rendering modes
         static const char STROKE[8] = { false, true, true, false, false, true, true, false };

src/Param.h

@@ -32,6 +32,7 @@ struct Param
     std::string outline_filename;
     int process_nontext;
     int process_outline;
+    int fallback;
     
     // fonts
     int embed_base_font;

src/css_class_names.cmakelists.txt

@@ -2,19 +2,21 @@
 
 # Note
 # don't use: (otherwise conflicted with others when there is an ID suffix)
-# p f
+# p f s
 
 set(CSS_INVALID_ID          "_")
 
-set(CSS_LINE_CN             "tl") # text line
+set(CSS_LINE_CN             "t") # text 
-set(CSS_TRANSFORM_MATRIX_CN "tm") # transform matrix
+set(CSS_TRANSFORM_MATRIX_CN "m") # matrix
 
 set(CSS_PAGE_DECORATION_CN  "pd") # page decoration
 set(CSS_PAGE_FRAME_CN       "pf") # page frame
 set(CSS_PAGE_CONTENT_BOX_CN "pc") # page content
 set(CSS_PAGE_DATA_CN        "pi") # page info
 
-set(CSS_FONT_NAME_CN        "fn") # font name
+set(CSS_BACKGROUND_IMAGE_CN "bi") # background image
+
+set(CSS_FONT_FAMILY_CN      "ff") # font family
 set(CSS_FONT_SIZE_CN        "fs") # font size
 
 set(CSS_FILL_COLOR_CN       "fc") # fill color

src/pdf2htmlEX.cc

@@ -64,7 +64,7 @@ void parse_options (int argc, char **argv)
         .add("zoom", &param.zoom, 0, "zoom ratio", nullptr, true)
         .add("fit-width", &param.fit_width, 0, "fit width to <fp> pixels", nullptr, true) 
         .add("fit-height", &param.fit_height, 0, "fit height to <fp> pixels", nullptr, true)
-        .add("use-cropbox", &param.use_cropbox, 0, "use CropBox instead of MediaBox")
+        .add("use-cropbox", &param.use_cropbox, 1, "use CropBox instead of MediaBox")
         .add("hdpi", &param.h_dpi, 144.0, "horizontal resolution for graphics in DPI")
         .add("vdpi", &param.v_dpi, 144.0, "vertical resolution for graphics in DPI")
         
@@ -76,6 +76,7 @@ void parse_options (int argc, char **argv)
         .add("outline-filename", &param.outline_filename, "", "filename of the generated outline file")
         .add("process-nontext", &param.process_nontext, 1, "render graphics in addition to text")
         .add("process-outline", &param.process_outline, 1, "show outline in HTML")
+        .add("fallback", &param.fallback, 0, "output in fallback mode")
         
         // fonts
         .add("embed-base-font", &param.embed_base_font, 0, "embed local match for standard 14 fonts")
@@ -104,7 +105,8 @@ void parse_options (int argc, char **argv)
         // misc.
         .add("clean-tmp", &param.clean_tmp, 1, "remove temporary files after conversion")
         .add("data-dir", &param.data_dir, PDF2HTMLEX_DATA_PATH, "specify data directory")
-        .add("css-draw", &param.css_draw, 0, "[experimental and unsupported] CSS drawing")
+        // TODO: css drawings are hidden on print, for annot links, need to fix it for other drawings
+//        .add("css-draw", &param.css_draw, 0, "[experimental and unsupported] CSS drawing")
         .add("debug", &param.debug, 0, "print debugging information")
         
         // meta

src/util/StateManager.h

@@ -453,6 +453,43 @@ public:
     }
 };
 
+/////////////////////////////////////
+/*
+ * Manage the background image sizes
+ * Kind of similar with StateManager, but not exactly the same
+ * anyway temporarly leave it here
+ */
+class BGImageSizeManager
+{
+public:
+    void install(int page_no, double width, double height){
+        value_map.insert(std::make_pair(page_no, std::make_pair(width, height)));
+    }
+
+    void dump_css(std::ostream & out) {
+        for(auto iter = value_map.begin(); iter != value_map.end(); ++iter)
+        {
+            const auto & s = iter->second;
+            out << "." << CSS::PAGE_CONTENT_BOX_CN << iter->first << "{";
+            out << "background-size:" << round(s.first) << "px " << round(s.second) << "px;";
+            out << "}" << std::endl;
+        }
+    }
+
+    void dump_print_css(std::ostream & out, double scale) {
+        for(auto iter = value_map.begin(); iter != value_map.end(); ++iter)
+        {
+            const auto & s = iter->second;
+            out << "." << CSS::PAGE_CONTENT_BOX_CN << iter->first << "{";
+            out << "background-size:" << round(s.first * scale) << "pt " << round(s.second * scale) << "pt;";
+            out << "}" << std::endl;
+        }
+    }
+
+private:
+    std::unordered_map<int, std::pair<double,double>> value_map; 
+};
+
 } // namespace pdf2htmlEX 
 
 #endif //STATEMANAGER_H__

src/util/css_const.h.in

@@ -35,7 +35,9 @@ const char * const PAGE_FRAME_CN       = "@CSS_PAGE_FRAME_CN@";
 const char * const PAGE_CONTENT_BOX_CN = "@CSS_PAGE_CONTENT_BOX_CN@";
 const char * const PAGE_DATA_CN        = "@CSS_PAGE_DATA_CN@";
 
-const char * const FONT_NAME_CN        = "@CSS_FONT_NAME_CN@";
+const char * const BACKGROUND_IMAGE_CN = "@CSS_BACKGROUND_IMAGE_CN@";
+
+const char * const FONT_FAMILY_CN      = "@CSS_FONT_FAMILY_CN@";
 const char * const FONT_SIZE_CN        = "@CSS_FONT_SIZE_CN@";
 const char * const FILL_COLOR_CN       = "@CSS_FILL_COLOR_CN@";
 const char * const STROKE_COLOR_CN     = "@CSS_STROKE_COLOR_CN@";

src/util/ffw.c

@@ -260,12 +260,14 @@ void ffw_cidflatten(void)
     SFFlatten(cur_fv->sf->cidmaster);
 }
 
+/*
+ * There is no check if a glyph with the same unicode exists!
+ */
 void ffw_add_empty_char(int32_t unicode, int width)
 {
-    // append the new char to Enc
     SplineChar * sc = SFMakeChar(cur_fv->sf, cur_fv->map, cur_fv->map->enccount);
-    sc->unicodeenc = unicode;
+    SCSetMetaData(sc, sc->name, unicode, sc->comment);
-    sc->width = width;
+    SCSynchronizeWidth(sc, width, sc->width, cur_fv);
 }
 
 int ffw_get_em_size(void)
@@ -319,6 +321,13 @@ void ffw_metric(double * ascent, double * descent)
     sf->descent = em - bb.maxy;
     */
 
+    /*
+     * The embedded fonts are likely to have inconsistent values for the 3 sets of ascent/descent
+     * PDF viewers don't care, since they don't even use these values
+     * But have to unify them, for different browsers on different platforms
+     * Things may become easier when there are CSS rules for baseline-based positioning.
+     */
+
     info->os2_winascent = a;
     info->os2_typoascent = a;
     info->hhead_ascent = a;