Add emcc option "--split <size>" to split javascript file.

Splits the resulting javascript file into pieces to ease debugging.
This option only works if Javascript is generated (target -o <name>.js).

Files with function declarations must be loaded before main file upon execution.

Without "-g" option:
Creates files with function declarations up to the given size with the suffix "_functions.partxxx.js" and a main file with the suffix ".js".

With "-g" option:
Recreates the directory structure of the C source files and stores function declarations in their respective C files with the suffix ".js". If such a file exceeds the given size, files with the suffix ".partxxx.js" are created. The main file resides in the base directory and has the suffix ".js".

7 files changed, 338 insertions, 2 deletions

diff --git a/AUTHORS b/AUTHORS
index a46d90d9..e30a10ea 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -37,3 +37,4 @@ a license to everyone to use it as detailed in LICENSE.)
 * Benjamin Stover <benjamin.stover@gmail.com>
 * Riccardo Magliocchetti <riccardo.magliocchetti@gmail.com>
 * Janus Troelsen <janus.troelsen@stud.tu-darmstadt.de>
+* Lars Schneider <lars.schneider@autodesk.com> (copyright owned by Autodesk, Inc.)
\ No newline at end of file
diff --git a/emcc b/emcc
index 97bc2f68..1dc9b756 100755
--- a/emcc
+++ b/emcc
@@ -268,6 +268,27 @@ Options that are modified or new in %s include:
                               will not minify the code (closure does
                               that)
 
+  --split <size>           Splits the resulting javascript file into pieces
+                           to ease debugging. This option only works if
+                           Javascript is generated (target -o <name>.js).
+                           Files with function declarations must be loaded
+                           before main file upon execution.
+
+                           Without "-g" option:
+                              Creates files with function declarations up
+                              to the given size with the suffix
+                              "_functions.partxxx.js" and a main file with
+                              the suffix ".js".
+
+                           With "-g" option:
+                              Recreates the directory structure of the C
+                              source files and stores function declarations
+                              in their respective C files with the suffix
+                              ".js". If such a file exceeds the given size,
+                              files with the suffix ".partxxx.js" are created.
+                              The main file resides in the base directory and
+                              has the suffix ".js".
+
   --ignore-dynamic-linking Normally emcc will treat dynamic linking like
                            static linking, by linking in the code from
                            the dynamic library. This fails if the same
@@ -471,6 +492,7 @@ try:
   pre_js = ''
   post_js = ''
   minify_whitespace = None
+  split_js_file = None
   preload_files = []
   embed_files = []
   compression = None
@@ -527,6 +549,11 @@ try:
       minify_whitespace = int(newargs[i+1])
       newargs[i] = ''
       newargs[i+1] = ''
+    elif newargs[i].startswith('--split'):
+      check_bad_eq(newargs[i])
+      split_js_file = int(newargs[i+1])
+      newargs[i] = ''
+      newargs[i+1] = ''      
     elif newargs[i].startswith('--embed-file'):
       check_bad_eq(newargs[i])
       embed_files.append(newargs[i+1])
@@ -601,6 +628,9 @@ try:
       newargs[i+1] = ''
   newargs = [ arg for arg in newargs if arg is not '' ]
 
+  if split_js_file:
+      settings_changes.append("PRINT_SPLIT_FILE_MARKER=1")
+
   # Find input files
 
   input_files = []
@@ -1071,8 +1101,12 @@ try:
 
     html.close()
   else:
-    # copy final JS to output
-    shutil.move(final, target)
+    if split_js_file:
+        from tools.split import split_javascript_file
+        split_javascript_file(final, unsuffixed(target), split_js_file)
+    else:
+        # copy final JS to output
+        shutil.move(final, target)
 
 finally:
   if not TEMP_DIR:
diff --git a/src/jsifier.js b/src/jsifier.js
index 92b39b7d..8021f8a1 100644
--- a/src/jsifier.js
+++ b/src/jsifier.js
@@ -522,6 +522,11 @@ function JSify(data, functionsOnly, givenFunctions) {
         func.JS += ' */\n';
       }
 
+      if (PRINT_SPLIT_FILE_MARKER) {
+          func.JS += '\n//FUNCTION_BEGIN_MARKER\n'
+          var associatedSourceFile = "NO_SOURCE";
+      }
+      
       func.JS += 'function ' + func.ident + '(' + paramIdents.join(', ') + ') {\n';
 
       if (PROFILE) {
@@ -572,6 +577,13 @@ function JSify(data, functionsOnly, givenFunctions) {
           if (EXECUTION_TIMEOUT > 0) {
             ret += indent + 'if (Date.now() - START_TIME >= ' + (EXECUTION_TIMEOUT*1000) + ') throw "Timed out!" + (new Error().stack);\n';
           }
+          
+          if (PRINT_SPLIT_FILE_MARKER && Debugging.on && Debugging.getAssociatedSourceFile(line.lineNum)) {
+            // Overwrite the associated source file for every line. The last line should contain the source file associated to
+            // the return value/address of outer most block (the marked function).
+            associatedSourceFile = Debugging.getAssociatedSourceFile(line.lineNum);
+          }
+          
           // for special labels we care about (for phi), mark that we visited them
           return ret + label.lines.map(function(line) { return line.JS + (Debugging.on ? Debugging.getComment(line.lineNum) : '') })
                                   .join('\n')
@@ -653,6 +665,11 @@ function JSify(data, functionsOnly, givenFunctions) {
         func.JS += '  return' + (func.returnType !== 'void' ? ' null' : '') + ';\n';
       }
       func.JS += '}\n';
+      
+      if (PRINT_SPLIT_FILE_MARKER) {
+          func.JS += '\n//FUNCTION_END_MARKER_OF_SOURCE_FILE_' + associatedSourceFile + '\n';
+      }
+      
       if (func.ident in EXPORTED_FUNCTIONS) {
         func.JS += 'Module["' + func.ident + '"] = ' + func.ident + ';';
       }
diff --git a/src/modules.js b/src/modules.js
index 0f3b483b..cf1b072e 100644
--- a/src/modules.js
+++ b/src/modules.js
@@ -127,6 +127,11 @@ var Debugging = {
                                                                 this.llvmLineToSourceFile[lineNum] + '"' : '';
   },
 
+  getAssociatedSourceFile: function(lineNum) {
+    if (!this.on) return null;
+    return lineNum in this.llvmLineToSourceLine ? this.llvmLineToSourceFile[lineNum] : null;
+  },
+  
   getIdentifier: function(lineNum) {
     if (!this.on) return null;
     if (lineNum === undefined) {
diff --git a/src/settings.js b/src/settings.js
index f8a81711..9a1afd69 100644
--- a/src/settings.js
+++ b/src/settings.js
@@ -202,6 +202,8 @@ var SHELL_FILE = 0; // set this to a string to override the shell file used
 
 var SHOW_LABELS = 0; // Show labels in the generated code
 
+var PRINT_SPLIT_FILE_MARKER = 0; // Prints markers in Javascript generation to split the file later on. See emcc --split option.
+
 var BUILD_AS_SHARED_LIB = 0; // Whether to build the code as a shared library, which
                              // must be loaded dynamically using dlopen().
                              // 0 here means this is not a shared lib: It is a main file.
diff --git a/tests/runner.py b/tests/runner.py
index 8d1f0674..a55b5ff5 100755
--- a/tests/runner.py
+++ b/tests/runner.py
@@ -7883,6 +7883,189 @@ elif 'browser' in str(sys.argv):
       finally:
         os.chdir(cwd)
 
+    def test_split(self):
+      # test HTML generation.
+      self.reftest(path_from_root('tests', 'htmltest.png'))
+      output = Popen(['python', EMCC, path_from_root('tests', 'hello_world_sdl.cpp'), '-o', 'something.js', '--split', '100']).communicate()
+      assert os.path.exists(os.path.join(self.get_dir(), 'something.js')), 'must be side js'
+      assert os.path.exists(os.path.join(self.get_dir(), 'something_functions.js')), 'must be side js'
+      assert os.path.exists(os.path.join(self.get_dir(), 'something.include.html')), 'must be side js'
+      
+      open(os.path.join(self.get_dir(), 'something.html'), 'w').write('''
+      
+      <!doctype html>
+      <html lang="en-us">
+        <head>
+          <meta charset="utf-8">
+          <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
+          <title>Emscripten-Generated Code</title>
+          <style>
+            .emscripten { padding-right: 0; margin-left: auto; margin-right: auto; display: block; }
+            canvas.emscripten { border: 1px solid black; }
+            textarea.emscripten { font-family: monospace; width: 80%; }
+            div.emscripten { text-align: center; }
+          </style>
+        </head>
+        <body>
+          <hr/>
+          <div class="emscripten" id="status">Downloading...</div>
+          <div class="emscripten">
+            <progress value="0" max="100" id="progress" hidden=1></progress>  
+          </div>
+          <canvas class="emscripten" id="canvas" oncontextmenu="event.preventDefault()"></canvas>
+          <hr/>
+          <div class="emscripten"><input type="button" value="fullscreen" onclick="Module.requestFullScreen()"></div>
+          <hr/>
+          <textarea class="emscripten" id="output" rows="8"></textarea>
+          <hr>
+          <script type='text/javascript'>
+            // connect to canvas
+            var Module = {
+              preRun: [],
+              postRun: [],
+              print: (function() {
+                var element = document.getElementById('output');
+                element.value = ''; // clear browser cache
+                return function(text) {
+                  // These replacements are necessary if you render to raw HTML
+                  //text = text.replace(/&/g, "&amp;");
+                  //text = text.replace(/</g, "&lt;");
+                  //text = text.replace(/>/g, "&gt;");
+                  //text = text.replace('\\n', '<br>', 'g');
+                  element.value += text + "\\n";
+                  element.scrollTop = 99999; // focus on bottom
+                };
+              })(),
+              printErr: function(text) {
+                if (0) { // XXX disabled for safety typeof dump == 'function') {
+                  dump(text + '\\n'); // fast, straight to the real console
+                } else {
+                  console.log(text);
+                }
+              },
+              canvas: document.getElementById('canvas'),
+              setStatus: function(text) {
+                if (Module.setStatus.interval) clearInterval(Module.setStatus.interval);
+                var m = text.match(/([^(]+)\((\d+(\.\d+)?)\/(\d+)\)/);
+                var statusElement = document.getElementById('status');
+                var progressElement = document.getElementById('progress');
+                if (m) {
+                  text = m[1];
+                  progressElement.value = parseInt(m[2])*100;
+                  progressElement.max = parseInt(m[4])*100;
+                  progressElement.hidden = false;
+                } else {
+                  progressElement.value = null;
+                  progressElement.max = null;
+                  progressElement.hidden = true;
+                }
+                statusElement.innerHTML = text;
+              },
+              totalDependencies: 0,
+              monitorRunDependencies: function(left) {
+                this.totalDependencies = Math.max(this.totalDependencies, left);
+                Module.setStatus(left ? 'Preparing... (' + (this.totalDependencies-left) + '/' + this.totalDependencies + ')' : 'All downloads complete.');
+              }
+            };
+            Module.setStatus('Downloading...');
+          </script>''' + open(os.path.join(self.get_dir(), 'something.include.html')).read() + '''
+        </body>
+      </html>
+      ''')
+      
+      self.run_browser('something.html', 'You should see "hello, world!" and a colored cube.', '/report_result?0')
+
+    def test_split_in_source_filenames(self):
+      self.reftest(path_from_root('tests', 'htmltest.png'))
+      output = Popen(['python', EMCC, path_from_root('tests', 'hello_world_sdl.cpp'), '-o', 'something.js', '-g', '--split', '100']).communicate()
+      assert os.path.exists(os.path.join(self.get_dir(), 'something.js')), 'must be side js'
+      assert os.path.exists(self.get_dir() + '/something/' + path_from_root('tests', 'hello_world_sdl.cpp.js')), 'must be side js'
+      assert os.path.exists(os.path.join(self.get_dir(), 'something.include.html')), 'must be side js'
+
+      open(os.path.join(self.get_dir(), 'something.html'), 'w').write('''
+
+      <!doctype html>
+      <html lang="en-us">
+        <head>
+          <meta charset="utf-8">
+          <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
+          <title>Emscripten-Generated Code</title>
+          <style>
+            .emscripten { padding-right: 0; margin-left: auto; margin-right: auto; display: block; }
+            canvas.emscripten { border: 1px solid black; }
+            textarea.emscripten { font-family: monospace; width: 80%; }
+            div.emscripten { text-align: center; }
+          </style>
+        </head>
+        <body>
+          <hr/>
+          <div class="emscripten" id="status">Downloading...</div>
+          <div class="emscripten">
+            <progress value="0" max="100" id="progress" hidden=1></progress>  
+          </div>
+          <canvas class="emscripten" id="canvas" oncontextmenu="event.preventDefault()"></canvas>
+          <hr/>
+          <div class="emscripten"><input type="button" value="fullscreen" onclick="Module.requestFullScreen()"></div>
+          <hr/>
+          <textarea class="emscripten" id="output" rows="8"></textarea>
+          <hr>
+          <script type='text/javascript'>
+            // connect to canvas
+            var Module = {
+              preRun: [],
+              postRun: [],
+              print: (function() {
+                var element = document.getElementById('output');
+                element.value = ''; // clear browser cache
+                return function(text) {
+                  // These replacements are necessary if you render to raw HTML
+                  //text = text.replace(/&/g, "&amp;");
+                  //text = text.replace(/</g, "&lt;");
+                  //text = text.replace(/>/g, "&gt;");
+                  //text = text.replace('\\n', '<br>', 'g');
+                  element.value += text + "\\n";
+                  element.scrollTop = 99999; // focus on bottom
+                };
+              })(),
+              printErr: function(text) {
+                if (0) { // XXX disabled for safety typeof dump == 'function') {
+                  dump(text + '\\n'); // fast, straight to the real console
+                } else {
+                  console.log(text);
+                }
+              },
+              canvas: document.getElementById('canvas'),
+              setStatus: function(text) {
+                if (Module.setStatus.interval) clearInterval(Module.setStatus.interval);
+                var m = text.match(/([^(]+)\((\d+(\.\d+)?)\/(\d+)\)/);
+                var statusElement = document.getElementById('status');
+                var progressElement = document.getElementById('progress');
+                if (m) {
+                  text = m[1];
+                  progressElement.value = parseInt(m[2])*100;
+                  progressElement.max = parseInt(m[4])*100;
+                  progressElement.hidden = false;
+                } else {
+                  progressElement.value = null;
+                  progressElement.max = null;
+                  progressElement.hidden = true;
+                }
+                statusElement.innerHTML = text;
+              },
+              totalDependencies: 0,
+              monitorRunDependencies: function(left) {
+                this.totalDependencies = Math.max(this.totalDependencies, left);
+                Module.setStatus(left ? 'Preparing... (' + (this.totalDependencies-left) + '/' + this.totalDependencies + ')' : 'All downloads complete.');
+              }
+            };
+            Module.setStatus('Downloading...');
+          </script>''' + open(os.path.join(self.get_dir(), 'something.include.html')).read() + '''
+        </body>
+      </html>
+      ''')
+
+      self.run_browser('something.html', 'You should see "hello, world!" and a colored cube.', '/report_result?0')
+        
     def test_compression(self):
       open(os.path.join(self.get_dir(), 'main.cpp'), 'w').write(self.with_report_result(r'''
         #include <stdio.h>
diff --git a/tools/split.py b/tools/split.py
new file mode 100755
index 00000000..ab1e97ca
--- /dev/null
+++ b/tools/split.py
@@ -0,0 +1,94 @@
+import sys
+import os
+
+def split_javascript_file(input_filename, output_filename_prefix, max_part_size_in_bytes):
+ 
+  try:
+    # Contains the entire Emscripten generated Javascript code
+    input_file = open(input_filename,'r')
+
+    # Javascript main file. On execution, this file needs to be loaded at last (!)
+    output_main_filename = output_filename_prefix + ".js"
+    output_main_file = open(output_main_filename,'w')
+
+    # File with HTML script tags to load the Javascript files in HTML later on
+    output_html_include_file = open(output_filename_prefix + ".include.html",'w')
+
+    # Variable will contain the source of a Javascript function if we find one during parsing
+    js_function = None
+    
+    # Dictionary with source file as key and an array of functions associated to that source file as value
+    function_buckets = {};
+    
+    output_part_file = None
+  
+    # Iterate over Javascript source; write main file; parse function declarations.
+    for line in input_file:
+      if line == "//FUNCTION_BEGIN_MARKER\n":
+        js_function = "//Func\n"
+      elif line.startswith("//FUNCTION_END_MARKER_OF_SOURCE_FILE_"):
+        # At the end of the function marker we get the source file that is associated to that function.
+        associated_source_file_base = line[len("//FUNCTION_END_MARKER_OF_SOURCE_FILE_"):len(line)-1]
+        
+        if associated_source_file_base == "NO_SOURCE":
+          # Functions without associated source file are stored in a file in the base directory
+          associated_source_file_base = output_filename_prefix + "_functions";
+        else:
+          # Functions with a known associated source file are stored in a file in the directory `output_filename_prefix`
+          associated_source_file_base = output_filename_prefix + os.path.realpath(associated_source_file_base)
+        
+        # Add the function to its respective file
+        if associated_source_file_base not in function_buckets:
+          function_buckets[associated_source_file_base] = []
+        function_buckets[associated_source_file_base] += [js_function]
+        
+        # Clear the function read cache
+        js_function = None
+      else:
+        if js_function is None:
+          output_main_file.write(line)
+        else:
+          js_function += line
+          
+    # Iterate over all function buckets and write their functions to the associated files
+    # An associated file is split into chunks of `max_part_size_in_bytes`
+    for associated_source_file_base in function_buckets:
+      # At first we try to name the Javascript source file to match the assoicated source file + `.js`
+      js_source_file = associated_source_file_base + ".js"
+    
+      # Check if the directory of the Javascript source file exists
+      js_source_dir = os.path.dirname(js_source_file)
+      if len(js_source_dir) > 0 and not os.path.exists(js_source_dir):
+        os.makedirs(js_source_dir)
+    
+      output_part_file_counter = 0
+      output_part_file = None
+      for js_function in function_buckets[associated_source_file_base]:
+        if output_part_file is None:
+          output_html_include_file.write("<script type=\"text/javascript\" src=\"" + js_source_file + "\"></script>")
+          output_part_file = open(js_source_file,'w')
+        
+        output_part_file.write(js_function)
+        
+        if output_part_file is not None and output_part_file.tell() > max_part_size_in_bytes:
+          output_part_file.close()
+          output_part_file = None
+          output_part_file_counter += 1
+          js_source_file = associated_source_file_base + ".part" + str(output_part_file_counter) + ".js"
+    
+      if output_part_file is not None:
+        output_part_file.close()
+        output_part_file = None
+      
+    # Write the main Javascript file at last to the HTML includes because this file contains the code to start
+    # the execution of the generated Emscripten application and requires all the extracted functions.
+    output_html_include_file.write("<script type=\"text/javascript\" src=\"" + output_main_filename + "\"></script>")
+
+  except Exception, e:
+    print >> sys.stderr, 'error: Splitting of Emscripten generated Javascript failed: %s' % str(e)
+  
+  finally:
+     if input_file is not None: input_file.close()
+     if output_main_file is not None: output_main_file.close()
+     if output_part_file is not None: output_part_file.close()
+     if output_html_include_file is not None: output_html_include_file.close()
\ No newline at end of file