ParserStorage: add Atom parser

/reviewed-on nexedi/jio!83

src/jio.storage/parserstorage.js

@@ -79,6 +79,77 @@
     }
     return result_list;
   };
+  /////////////////////////////////////////////////////////////
+  // ATOM Parser
+  /////////////////////////////////////////////////////////////
+  function ATOMParser(txt) {
+    this._dom_parser = new DOMParser().parseFromString(txt, 'text/xml');
+  }
+  ATOMParser.prototype.parseElement = function (element) {
+    var tag_element,
+      i,
+      j,
+      tag_name,
+      attribute,
+      result = {};
+
+    for (i = element.childNodes.length - 1; i >= 0; i -= 1) {
+      tag_element = element.childNodes[i];
+      if ((tag_element.nodeType === Node.ELEMENT_NODE) &&
+          (tag_element.tagName !== 'entry')) {
+        tag_name = tag_element.tagName;
+        // may have several links, with different rel value
+        // default is alternate
+        if (tag_name === 'link') {
+          tag_name += '_' + (tag_element.getAttribute('rel') || 'alternate');
+        } else {
+          result[tag_name] = tag_element.textContent;
+        }
+        for (j = tag_element.attributes.length - 1; j >= 0; j -= 1) {
+          attribute = tag_element.attributes[j];
+          if (attribute.value) {
+            result[tag_name + '_' + attribute.name] =
+              attribute.value;
+          }
+        }
+
+      }
+    }
+    return result;
+  };
+  ATOMParser.prototype.getDocumentList = function (include, id) {
+    var result_list,
+      item_list = this._dom_parser.querySelectorAll("feed > entry"),
+      i;
+
+    if ((id === '/0') || (id === undefined)) {
+      result_list = [{
+        id: '/0',
+        value: {}
+      }];
+      if (include) {
+        result_list[0].doc = this.parseElement(
+          this._dom_parser.querySelector("feed")
+        );
+      }
+    } 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;
+  };
 
   /////////////////////////////////////////////////////////////
   // RSS Parser
@@ -151,7 +222,8 @@
   /////////////////////////////////////////////////////////////
   var parser_dict = {
     'rss': RSSParser,
-    'opml': OPMLParser
+    'opml': OPMLParser,
+    'atom': ATOMParser
   };
 
   function getParser(storage) {
@@ -203,4 +275,4 @@
 
   jIO.addStorage('parser', ParserStorage);
 
-}(jIO, DOMParser, Node));
+}(jIO, DOMParser, Node));
\ No newline at end of file

test/jio.storage/parserstorage.tests.js

@@ -51,7 +51,57 @@
   };
 
   jIO.addStorage('rssstorage200', RSSStorage200);
+  /////////////////////////////////////////////////////////////////
+  // Custom atom test substorage definition
+  /////////////////////////////////////////////////////////////////
+  function ATOMStorage200() {
+    return this;
+  }
 
+  ATOMStorage200.prototype.getAttachment = function (id, name) {
+    equal(id, 'foo');
+    equal(name, 'bar');
+    var txt = '<?xml version="1.0" encoding="utf-8"?>' +
+      '<feed xmlns="http://www.w3.org/2005/Atom">' +
+      '<title>Example Feed</title>' +
+      '<subtitle>A subtitle.</subtitle>' +
+      '<link href="http://example.org/feed/" rel="self" />' +
+      '<link href="http://example.org/" />' +
+      '<id>urn:uuid:60a76c80-d399-11d9-b91C-0003939e0af6</id>' +
+      '<updated>2003-12-13T18:30:02Z</updated>' +
+      '<entry>' +
+      '<id>tag:example.org,2018-03-29:81102</id>' +
+      '<title>Atom-Powered Robots Run Amok</title>' +
+      '<link href="/2003/12/13/atom03" />' +
+      '<link rel="alternate" type="text/html" ' +
+      'href="http://example.org/2003/12/13/atom03.html"/>' +
+      '<link rel="edit" href="http://example.org/2003/12/13/atom03/edit"/>' +
+      '<author>' +
+      '<name>John Doe</name>' +
+      '<email>johndoe@example.com</email>' +
+      '</author>' +
+      '<updated>2003-12-13T18:30:02Z</updated>' +
+      '<summary>Some text.</summary>' +
+      '</entry>' +
+      '<entry>' +
+      '<title>Atom-Powered Robots Run Amok</title>' +
+      '<link href="http://example.org/2003/12/13/atom03" />' +
+      '<link rel="alternate" type="text/html" ' +
+      'href="http://example.org/2003/12/13/atom03.html"/>' +
+      '<link rel="edit" href="http://example.org/2003/12/13/atom03/edit"/>' +
+      '<id>urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a</id>' +
+      '<author>' +
+      '<name>John Doe</name>' +
+      '<email>johndoe@example.com</email>' +
+      '</author>' +
+      '<updated>2004-12-13T18:30:02Z</updated>' +
+      '<summary>Some text.</summary>' +
+      '</entry>' +
+      '</feed>';
+    return new Blob([txt]);
+  };
+
+  jIO.addStorage('atomstorage200', ATOMStorage200);
   /////////////////////////////////////////////////////////////////
   // Custom OPML test substorage definition
   /////////////////////////////////////////////////////////////////
@@ -244,6 +294,131 @@
       });
   });
 
+  test("get all IDs from ATOM", function () {
+    this.jio = jIO.createJIO({
+      type: 'parser',
+      document_id: 'foo',
+      attachment_id: 'bar',
+      parser: 'atom',
+      sub_storage: {
+        type: 'atomstorage200'
+      }
+    });
+
+    var expected_dict = {
+      "data": {
+        "rows": [
+          {
+            "id": "/0",
+            "value": {}
+          },
+          {
+            "id": "/0/0",
+            "value": {}
+          },
+          {
+            "id": "/0/1",
+            "value": {}
+          }
+        ],
+        "total_rows": 3
+      }
+    };
+
+    stop();
+    expect(3);
+
+    this.jio.allDocs()
+      .then(function (result) {
+        deepEqual(result, expected_dict, "Check documents");
+      })
+      .fail(function (error) {
+        ok(false, error);
+      })
+      .always(function () {
+        start();
+      });
+  });
+
+  test("get all documents from ATOM", function () {
+    this.jio = jIO.createJIO({
+      type: 'parser',
+      document_id: 'foo',
+      attachment_id: 'bar',
+      parser: 'atom',
+      sub_storage: {
+        type: 'atomstorage200'
+      }
+    });
+
+    var expected_dict = {
+      "data": {
+        "rows": [
+          {
+            "doc": {
+              "id": "urn:uuid:60a76c80-d399-11d9-b91C-0003939e0af6",
+              "link_alternate_href": "http://example.org/",
+              "link_self_href": "http://example.org/feed/",
+              "link_self_rel": "self",
+              "subtitle": "A subtitle.",
+              "title": "Example Feed",
+              "updated": "2003-12-13T18:30:02Z"
+            },
+            "id": "/0",
+            "value": {}
+          },
+          {
+            "doc": {
+              "author": "John Doejohndoe@example.com",
+              "id": "tag:example.org,2018-03-29:81102",
+              "link_alternate_href": "/2003/12/13/atom03",
+              "link_alternate_rel": "alternate",
+              "link_alternate_type": "text/html",
+              "link_edit_href": "http://example.org/2003/12/13/atom03/edit",
+              "link_edit_rel": "edit",
+              "summary": "Some text.",
+              "title": "Atom-Powered Robots Run Amok",
+              "updated": "2003-12-13T18:30:02Z"
+            },
+            "id": "/0/0",
+            "value": {}
+          },
+          {
+            "doc": {
+              "author": "John Doejohndoe@example.com",
+              "id": "urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a",
+              "link_alternate_href": "http://example.org/2003/12/13/atom03",
+              "link_alternate_rel": "alternate",
+              "link_alternate_type": "text/html",
+              "link_edit_href": "http://example.org/2003/12/13/atom03/edit",
+              "link_edit_rel": "edit",
+              "summary": "Some text.",
+              "title": "Atom-Powered Robots Run Amok",
+              "updated": "2004-12-13T18:30:02Z"
+            },
+            "id": "/0/1",
+            "value": {}
+          }
+        ],
+        "total_rows": 3
+      }
+    };
+
+    stop();
+    expect(3);
+
+    this.jio.allDocs({include_docs: true})
+      .then(function (result) {
+        deepEqual(result, expected_dict, "Check documents");
+      })
+      .fail(function (error) {
+        ok(false, error);
+      })
+      .always(function () {
+        start();
+      });
+  });
+
   test("get all IDs from OPML", function () {
     this.jio = jIO.createJIO({
       type: 'parser',
@@ -542,6 +717,104 @@
         start();
       });
   });
+  test("get Atom feed", function () {
+    this.jio = jIO.createJIO({
+      type: 'parser',
+      document_id: 'foo',
+      attachment_id: 'bar',
+      parser: 'atom',
+      sub_storage: {
+        type: 'atomstorage200'
+      }
+    });
+
+    stop();
+    expect(3);
+
+    this.jio.get('/0')
+      .then(function (result) {
+        deepEqual(result, {
+          "id": "urn:uuid:60a76c80-d399-11d9-b91C-0003939e0af6",
+          "link_alternate_href": "http://example.org/",
+          "link_self_href": "http://example.org/feed/",
+          "link_self_rel": "self",
+          "subtitle": "A subtitle.",
+          "title": "Example Feed",
+          "updated": "2003-12-13T18:30:02Z"
+        }, "Check document");
+      })
+      .fail(function (error) {
+        ok(false, error);
+      })
+      .always(function () {
+        start();
+      });
+  });
+
+  test("get ATOM entry", function () {
+    this.jio = jIO.createJIO({
+      type: 'parser',
+      document_id: 'foo',
+      attachment_id: 'bar',
+      parser: 'atom',
+      sub_storage: {
+        type: 'atomstorage200'
+      }
+    });
+
+    stop();
+    expect(3);
+
+    this.jio.get('/0/1')
+      .then(function (result) {
+        deepEqual(result, {
+          "author": "John Doejohndoe@example.com",
+          "id": "urn:uuid:1225c695-cfb8-4ebb-aaaa-80da344efa6a",
+          "link_alternate_href": "http://example.org/2003/12/13/atom03",
+          "link_alternate_rel": "alternate",
+          "link_alternate_type": "text/html",
+          "link_edit_href": "http://example.org/2003/12/13/atom03/edit",
+          "link_edit_rel": "edit",
+          "summary": "Some text.",
+          "title": "Atom-Powered Robots Run Amok",
+          "updated": "2004-12-13T18:30:02Z"
+        }, "Check document");
+      })
+      .fail(function (error) {
+        ok(false, error);
+      })
+      .always(function () {
+        start();
+      });
+  });
+
+  test("get unknown atom entry", function () {
+    this.jio = jIO.createJIO({
+      type: 'parser',
+      document_id: 'foo',
+      attachment_id: 'bar',
+      parser: 'atom',
+      sub_storage: {
+        type: 'atomstorage200'
+      }
+    });
+
+    stop();
+    expect(5);
+
+    this.jio.get('foo')
+      .then(function (result) {
+        ok(false, result);
+      })
+      .fail(function (error) {
+        ok(error instanceof jIO.util.jIOError, error);
+        equal(error.message, "Cannot find parsed document: foo");
+        equal(error.status_code, 404);
+      })
+      .always(function () {
+        start();
+      });
+  });
 
   test("get OPML head", function () {
     this.jio = jIO.createJIO({
@@ -635,4 +908,4 @@
       });
   });
 
-}(jIO, QUnit, Blob));
+}(jIO, QUnit, Blob));
\ No newline at end of file