ParserStorage: extract data from RSS/OPML.

Thanks to Alain Takoudjou <alain.takoudjou@nexedi.com> for his original work on this topic.

Gruntfile.js

@@ -169,6 +169,7 @@ module.exports = function (grunt) {
           'src/jio.storage/uuidstorage.js',
           'src/jio.storage/memorystorage.js',
           'src/jio.storage/zipstorage.js',
+          'src/jio.storage/parserstorage.js',
           'src/jio.storage/httpstorage.js',
           'src/jio.storage/dropboxstorage.js',
           'src/jio.storage/davstorage.js',

src/jio.storage/parserstorage.js

+/*jslint nomen: true*/
+/*global jIO, DOMParser */
+(function (jIO, DOMParser) {
+  "use strict";
+
+  /////////////////////////////////////////////////////////////
+  // OPML Parser
+  /////////////////////////////////////////////////////////////
+  function OPMLParser(txt) {
+    this._dom_parser = new DOMParser().parseFromString(txt, 'text/xml');
+  }
+
+  OPMLParser.prototype.parseHead = function () {
+    // fetch all children instead
+    var channel_element = this._dom_parser.querySelector("opml > head"),
+      tag_element,
+      i,
+      result = {};
+
+    for (i = channel_element.childNodes.length - 1; i >= 0; i -= 1) {
+      tag_element = channel_element.childNodes[i];
+      result[tag_element.tagName] = tag_element.textContent;
+    }
+    return result;
+  };
+
+  OPMLParser.prototype.parseOutline = function (result_list, outline_element,
+                                                prefix, include, id) {
+    var attribute,
+      i,
+      child,
+      result = {};
+
+    if ((id === prefix) || (id === undefined)) {
+      result_list.push({
+        id: prefix,
+        value: {}
+      });
+      if (include) {
+        for (i = outline_element.attributes.length - 1; i >= 0; i -= 1) {
+          attribute = outline_element.attributes[i];
+          if (attribute.value) {
+            result[attribute.name] = attribute.value;
+          }
+        }
+        result_list[result_list.length - 1].doc = result;
+      }
+    }
+
+    for (i = outline_element.childNodes.length - 1; i >= 0; i -= 1) {
+      child = outline_element.childNodes[i];
+      if (child.tagName === 'outline') {
+        this.parseOutline(result_list, child, prefix + '/' + i, include, id);
+      }
+    }
+  };
+
+  OPMLParser.prototype.getDocumentList = function (include, id) {
+    var result_list,
+      item_list = this._dom_parser.querySelectorAll("body > outline"),
+      i;
+
+    if ((id === '/0') || (id === undefined)) {
+      result_list = [{
+        id: '/0',
+        value: {}
+      }];
+      if (include) {
+        result_list[0].doc = this.parseHead();
+      }
+    } else {
+      result_list = [];
+    }
+
+    for (i = 0; i < item_list.length; i += 1) {
+      this.parseOutline(result_list, item_list[i], '/1/' + i, include, id);
+    }
+    return result_list;
+  };
+
+  /////////////////////////////////////////////////////////////
+  // RSS Parser
+  /////////////////////////////////////////////////////////////
+  function RSSParser(txt) {
+    this._dom_parser = new DOMParser().parseFromString(txt, 'text/xml');
+  }
+
+  RSSParser.prototype.parseElement = function (element) {
+    var tag_element,
+      i,
+      j,
+      attribute,
+      result = {};
+
+    for (i = element.childNodes.length - 1; i >= 0; i -= 1) {
+      tag_element = element.childNodes[i];
+      if (tag_element.tagName !== 'item') {
+        result[tag_element.tagName] = tag_element.textContent;
+
+        for (j = tag_element.attributes.length - 1; j >= 0; j -= 1) {
+          attribute = tag_element.attributes[j];
+          if (attribute.value) {
+            result[tag_element.tagName + '_' + attribute.name] =
+              attribute.value;
+          }
+        }
+      }
+    }
+    return result;
+  };
+
+  RSSParser.prototype.getDocumentList = function (include, id) {
+    var result_list,
+      item_list = this._dom_parser.querySelectorAll("rss > channel > item"),
+      i;
+
+    if ((id === '/0') || (id === undefined)) {
+      result_list = [{
+        id: '/0',
+        value: {}
+      }];
+      if (include) {
+        result_list[0].doc = this.parseElement(
+          this._dom_parser.querySelector("rss > channel")
+        );
+      }
+    } else {
+      result_list = [];
+    }
+
+    for (i = 0; i < item_list.length; i += 1) {
+      if ((id === '/0/' + i) || (id === undefined)) {
+        result_list.push({
+          id: '/0/' + i,
+          value: {}
+        });
+        if (include) {
+          result_list[result_list.length - 1].doc =
+            this.parseElement(item_list[i]);
+        }
+      }
+    }
+    return result_list;
+  };
+
+  /////////////////////////////////////////////////////////////
+  // Helpers
+  /////////////////////////////////////////////////////////////
+  var parser_dict = {
+    'rss': RSSParser,
+    'opml': OPMLParser
+  };
+
+  function getParser(storage) {
+    return storage._sub_storage.getAttachment(storage._document_id,
+                                              storage._attachment_id,
+                                              {format: 'text'})
+      .push(function (txt) {
+        return new parser_dict[storage._parser_name](txt);
+      });
+  }
+
+  /////////////////////////////////////////////////////////////
+  // Storage
+  /////////////////////////////////////////////////////////////
+  function ParserStorage(spec) {
+    this._attachment_id = spec.attachment_id;
+    this._document_id = spec.document_id;
+    this._parser_name = spec.parser;
+    this._sub_storage = jIO.createJIO(spec.sub_storage);
+  }
+
+  ParserStorage.prototype.hasCapacity = function (capacity) {
+    return (capacity === "list") || (capacity === 'include');
+  };
+
+  ParserStorage.prototype.buildQuery = function (options) {
+    if (options === undefined) {
+      options = {};
+    }
+    return getParser(this)
+      .push(function (parser) {
+        return parser.getDocumentList((options.include_docs || false));
+      });
+  };
+
+  ParserStorage.prototype.get = function (id) {
+    return getParser(this)
+      .push(function (parser) {
+        var result_list = parser.getDocumentList(true, id);
+        if (result_list.length) {
+          return result_list[0].doc;
+        }
+        throw new jIO.util.jIOError(
+          "Cannot find parsed document: " + id,
+          404
+        );
+      });
+  };
+
+  jIO.addStorage('parser', ParserStorage);
+
+}(jIO, DOMParser));
\ No newline at end of file

test/tests.html

@@ -45,6 +45,7 @@
   <script src="jio.storage/replicatestorage_fastrepair.tests.js"></script>
   <script src="jio.storage/replicatestorage_fastrepairattachment.tests.js"></script>
   <script src="jio.storage/shastorage.tests.js"></script>
+  <script src="jio.storage/parserstorage.tests.js"></script>
 
   <!--script src="jio.storage/qiniustorage.tests.js"></script-->
   <!--script src="jio.storage/indexstorage.tests.js"></script-->