added support for source_content_type in os_daemon config

README.md

@@ -60,6 +60,9 @@ timeout = 60000
 ; Ignore the following databases (again comma separated list)
 ; Regular expressions are again allowed:
 blacklist = /^_/
+; Only attachments which match any of the mime types below are processed (separate with comma).
+; Default is /^image\// Regular expressions are allowed:
+;source_content_types = /^image\//,application/postscript
 ```
 
 ## Imagemagick Configuration

lib/couchmagick-stream.js

@@ -11,6 +11,18 @@ var nano = require('nano');
 var strformat = require('strformat');
 var docuri = require('docuri');
 
+function compileRegexp(string) {
+  if (string instanceof RegExp) {
+    return string;
+  }
+  if (string.match(/^\/.*\/$/)) {
+    string = string.slice(1, string.length - 1);
+  } else {
+    string = '^' + string + '$';
+  }
+  return new RegExp(string);
+}
+
 
 // TODO: emit all documents, also filtered one, to enable checkpointing
 module.exports = function couchmagick(url, options) {
@@ -18,6 +30,15 @@ module.exports = function couchmagick(url, options) {
   options.concurrency = options.concurrency || 1;
   options.timeout = options.timeout || 60 * 1000; // 1 minute
 
+  options.source_content_types = options.source_content_types ?
+          options.source_content_types.split(/,\s*/) : [/^image\//];
+  if (options.source_content_types) {
+    if (!Array.isArray(options.source_content_types)) {
+      options.source_content_types = [options.source_content_types];
+    }
+    options.source_content_types = options.source_content_types.map(compileRegexp);
+  }
+
 
   var couch = nano(url);
   var configs = {};
@@ -224,7 +245,10 @@ module.exports = function couchmagick(url, options) {
         return false;
       }
 
-      return data.doc._attachments[data.name].content_type.match(/^image\//);
+      var content_type = data.doc._attachments[data.name].content_type;
+      return options.source_content_types.some(function(r) {
+              return content_type.match(r);
+      });
     }),
 
     // split stream into versions