combined the sanitize and validate funcitons into a single function; limited the format characters supported to avoid validation complexity; updated documentation; feature implemented by Ryan Morlok (ryan.morlok@morlok.com) on behalf of Docalytics (http://www.docalytics.com/)

pdf2htmlEX.1.in

@@ -67,7 +67,7 @@ You need to modify the manifest if you do not want outline embedded.
 .B --split-pages <0|1> (Default: 0)
 If turned on, the pages, css, and outline will be stored into separated files and no consolidated <output-filename>.html will be generated. 
 
-<output-filename> may be used to specify the format for the filenames for individual pages. <output-filename> may contain a %d placeholder to indicate where the page number should be placed. 
+<output-filename> may be used to specify the format for the filenames for individual pages. <output-filename> may contain a %d placeholder to indicate where the page number should be placed. The placeholder supports a limited subset of normal numerical placeholders, including specified width and zero padding.
 
 If <output-filename> does not contain a placeholder for the page number, the page number will be inserted directly before the file extension. If the filename does not have an extension, the page number will be placed at the end of the file name.
 

src/pdf2htmlEX.cc

@@ -215,33 +215,39 @@ int main(int argc, char **argv)
             if(get_suffix(param.input_filename) == ".pdf")
             {
                 if(param.split_pages)
-                    param.output_filename = sanitize_filename(s.substr(0, s.size() - 4) + "%d.page");
+                {
+                    param.output_filename = s.substr(0, s.size() - 4) + "%d.page";
+                    sanitize_filename(param.output_filename);
+                }
                 else
+                {
                     param.output_filename = s.substr(0, s.size() - 4) + ".html";
+                }
 
             }
             else
             {
                 if(param.split_pages)
-                    param.output_filename = sanitize_filename(s + "%d.page");
+                {
+                    param.output_filename = s + "%d.page";
+                    sanitize_filename(param.output_filename);
+                }
                 else
+                {
                     param.output_filename = s + ".html";
+                }
                 
             }
         }
 		else if(param.split_pages)
         {
             // Need to make sure we have a page number placeholder in the filename
-            if(!contains_integer_placeholder(param.output_filename))
+            if(!sanitize_filename(param.output_filename))
             {
                 // Inject the placeholder just before the file extension
                 const string suffix = get_suffix(param.output_filename);
-                param.output_filename = sanitize_filename(param.output_filename.substr(0, param.output_filename.size() - suffix.size()) + "%d" + suffix);
-            }
-            else
-            {
-                // Already have the placeholder, just make sure the name is safe.
-                param.output_filename = sanitize_filename(param.output_filename);
+                param.output_filename = param.output_filename.substr(0, param.output_filename.size() - suffix.size()) + "%d" + suffix;
+                sanitize_filename(param.output_filename);
             }
         }
         if(param.css_filename.empty())

src/util/path.cc

@@ -40,7 +40,7 @@ void create_directories(const string & path)
     }
 }
 
-string sanitize_filename(const string & filename)
+bool sanitize_filename(string & filename)
 {
     string sanitized;
     bool format_specifier_found = false;
@@ -65,13 +65,13 @@ string sanitize_filename(const string & filename)
                     tmp.push_back(filename[i]);
                     
                     // If we aren't still in option specifiers, stop looking
-                    if(!strchr("+-#0123456789.", filename[i]))
+                    if(!strchr("0123456789", filename[i]))
                     {
                         break;
                     }
                 }
                 
-                // Check to see if we yielded a valid format speifier
+                // Check to see if we yielded a valid format specifier
                 if('d' == tmp.back())
                 {
                     // Found a valid integer format
@@ -94,46 +94,15 @@ string sanitize_filename(const string & filename)
         }
     }
 
-    return sanitized;
-}
-
-bool contains_integer_placeholder(const string & filename)
-{
-    for(size_t i = 0; i < filename.size(); i++) 
+    // Only sanitize if it is a valid format.
+    if(format_specifier_found)
     {
-        if('%' == filename[i])
-        {
-            size_t original_i = i;
-            char last_char = '%';
-            while(++i < filename.size())
-            {
-                last_char = filename[i];
-                
-                // If we aren't still in option specifiers, stop looking
-                if(!strchr("+-#0123456789.", last_char))
-                {
-                    break;
-                }
-            }
-            
-            // Check to see if we yielded a valid format speifier
-            if('d' == last_char)
-            {
-                // Yep.
-                return true;
-            }
-            else
-            {
-                // Nope. Resume looking where we left off.
-                i = original_i;
-            }
-        }
+        filename.assign(sanitized);   
     }
 
-    return false;
+    return format_specifier_found;
 }
 
-
 bool is_truetype_suffix(const string & suffix)
 {
     return (suffix == ".ttf") || (suffix == ".ttc") || (suffix == ".otf");

src/util/path.h

@@ -20,24 +20,14 @@ std::string get_filename(const std::string & path);
 std::string get_suffix(const std::string & path);
 
 /**
- * Function to sanitize a filename so that it can be eventually safely used in a printf 
- * statement. Allows a single %d placeholder, but no other format specifiers.
+ * Sanitize all occurrences of '%' except for the first valid format specifier. Filename
+ * is only sanitized if a formatter is found, and the function returns true.
  *
- * @param filename the filename to be sanitized.
+ * @param filename the filename to be sanitized. Value will be modified.
  *
- * @return the sanitized filename.
+ * @return true if a format specifier was found, false otherwise.
  */ 
-std::string sanitize_filename(const std::string & filename);
-
-/**
- * Function to check if a filename contains at least one %d integer placeholder
- * for use in a printf statement.
- *
- * @param filename the filename to check
- *
- * @return true if the filename contains an integer placeholder, false otherwise.
- */
-bool contains_integer_placeholder(const std::string & filename);
+bool sanitize_filename(std::string & filename);
 
 } //namespace pdf2htmlEX 
 #endif //PATH_H__

test/test_naming.py

@@ -190,6 +190,22 @@ class OutputNamingTests(unittest.TestCase):
         ])
         self.assertEquals(files, sorted(['3-pages.css', '3-pages.outline', 'f%%oo1.xyz', 'f%%oo2.xyz', 'f%%oo3.xyz']))
 
+    def test_generate_split_pages_specify_name_only_percent_d_is_used_percent_percent_with_actual_placeholder(self):
+        files = execute_pdf2htmlex_and_get_files([
+            '--split-pages', 1,
+            path_to_test_file('3-pages.pdf'),
+            'f%%o%do.xyz'
+        ])
+        self.assertEquals(files, sorted(['3-pages.css', '3-pages.outline', 'f%%o1o.xyz', 'f%%o2o.xyz', 'f%%o3o.xyz']))
+
+    def test_generate_split_pages_specify_name_only_percent_d_is_used_percent_percent_with_actual_placeholder(self):
+        files = execute_pdf2htmlex_and_get_files([
+            '--split-pages', 1,
+            path_to_test_file('3-pages.pdf'),
+            'fo%do%%.xyz'
+        ])
+        self.assertEquals(files, sorted(['3-pages.css', '3-pages.outline', 'fo1o%%.xyz', 'fo2o%%.xyz', 'fo3o%%.xyz']))
+
     def test_generate_split_pages_specify_name_only_formatter_starts_part_way_through_invalid_formatter(self):
         files = execute_pdf2htmlex_and_get_files([
             '--split-pages', 1,