I have a category with keywords which in their tern have metadata schema. That schema consist of two fields and each of them is category. Very simple structure, but during publishing it resolves those metadata keyword fields into wrong tcm uris instead of title of the keyword, like the following:
2) Content of the deployer package
<tcmc:Topic rdf:about="tcm:10-11325-1024">
<rdfs:label>Analytics and optimization</rdfs:label>
<rdfs:comment>Analytics and optimization</rdfs:comment>
<tcmt:key>Analytics and optimization</tcmt:key>
<tcmt:isAbstract>false</tcmt:isAbstract>
<tcmt:isRoot>true</tcmt:isRoot>
<tcmt:metadata rdf:parseType="Literal">
<Metadata xmlns="uuid:a30b06d3-b6c5-4c2e-a53b-2b88771370ed">
<Divisions xlink:title="cma" xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="tcm:0-17737-1024">cma</Divisions>
<InterestProfile xlink:title="CMAAnalytics" xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="tcm:0-11175-1024">CMAAnalytics</InterestProfile>
</Metadata>
</tcmt:metadata>
</tcmc:Topic>
3) In code where I query Tridion it returns these uris:
TaxonomyFactory taxonomyFactory = new TaxonomyFactory();
TKeyword taxonomy = taxonomyFactory.GetTaxonomyKeywords(“tcm_of_the_category”);
if (taxonomy != null && taxonomy.KeywordChildren != null)
{
foreach (var item in taxonomy.KeywordChildren) //keyword metadata contains tcm uri with zero instead of title
{
Keyword keywordChildren = item as Keyword;
if (keywordChildren != null)
{
. . .
}
}
}
Does anyone have any ideas what might cause such an issue?
At a glance, my guess is that the internal template used to transform the categories is reading the metadata field data directly from the DB (or near enough in the BL layer) and not applying any blueprinting rules to it (likely for performance).
If you look at TCM Uris in content, when stored in the database, they all use 0 as their publication ID, and this ID is modified at "read" time.
Your call: You can call this a defect, ask Tridion to fix it, and it will degrade the performance of publishing a category, or you can deal with it in the delivery side - you know the publication Uri is 0, and you know you need to replace it with the current publication ID if you intend to use it for any purpose.
EDIT
So I went back and did some quick hacking. Indeed you can't load the keyword's content because, according to Tridion, the "Value" of field "Divisions" is the keyword URI. No way around that.
Quick way around this: load the keyword in question:
TaxonomyFactory tf = new TaxonomyFactory();
Keyword taxonomy = tf.GetTaxonomyKeywords("tcm:5-369-512");
if(taxonomy != null && taxonomy.KeywordChildren != null)
{
foreach (Keyword item in taxonomy.KeywordChildren)
{
NameValuePair key = (NameValuePair) item.KeywordMeta.NameValues["Field1"];
string correctUri = key.Value.ToString().Replace("tcm:0-", "tcm:5-");
Keyword theOtherKeyword = tf.GetTaxonomyKeyword(correctUri);
string title = theOtherKeyword.KeywordName;
}
}
Now... you probably want to be a bit smarter than me on that creative publication ID rewrite :)
You can see the field as a Component Link, you link to a specific Keyword item (object). Therefor you primarily get the URI, and I don't think that it resolves automatically to the Value property.
So the next step would be to obtain the Keyword object using the URI, and possibly construct the URI to include the right publication context.
Related
See Datasource Paging Issue (Revised)
for the original question.
Markus, you were kind enough to help with out with the issue of incorporating a record count into a query using a calculated datasource. I have a search form with 15 widgets - a mix of date ranges, dropdowns, text values and ._contains, ._equals, ._greaterThanOrEquals, ._lessThanOrEquals, etc.
I have tested this extensively against mySQL SQL code and it works fine.
I have now added a 16th parameter PropertyNames, which is a list with binding #datasource.query.filters.Property.PropertyName._in and Options blank. The widget on the form is hidden because it is only used for additional filtering.
Logic such as the following is used, such that a particular logged-in user can only view their own properties. So if they perform a search and the Property is not specified we do:-
if (params.param_Property === null && canViewAllRecords === false) {
console.log(params.param_PropertyNames); // correct output
ds.filters.Property.PropertyName._in = params.param_PropertyNames;
}
The record count (records.length) is correct, and if I for loop through the array of records the record set is correct.
However, on the results page the table displays a larger resultset which omits the PropertyNames filter. So if I was to search on Status 'Open' (mySQL results 50) and then I add a single value ['Property Name London SW45'] for params.param_PropertyNames the record count is 6, the records array is 6 but the datasource display is 50. So the datasource is not filtering on the property array.
Initially I tried without adding the additional parameter and form widget and just using code such as
if (params.param_Property === null && canViewAllRecords === false) {
console.log(params.param_PropertyNames); // correct output
ds.filters.Property.PropertyName._in = properties; // an array of
properties to filter out
}
But this didn't work, hence the idea of adding a form widget and an additional parameter to the calculated recordcount datasource.
If I inspect at query.parameters then I see:-
"param_Status": "Open",
"param_PropertyNames": ["Property Name London SW45"],
If I inspect query.filters:-
name=param_Status, value=Open
name=param_PropertyNames, value=[]}]}
It looks as though the filter isn't set. Even hard coding
ds.filters.Property.PropertyName._in = ['Property Name London SW45'],
I get the same reuslt.
Have you got any idea what would be causing this issue and what I can do for a workaround ?
Using a server side solution I would suggest editing both your SQL datasource query script (server side) that is supposed to filter by this property list and including the same code in your server side script for your calculated Count datasource. The code would look something like this, not knowing your exact details:
var subquery = app.models.Directory.newQuery();
subquery.filters.PrimaryEmail._equals = Session.getActiveUser().getEmail();
subquery.prefetch.Property._add();
var results = subquery.run();
if(!results[0].CanViewAllRecords) {
query.filters.Property.PropertyName._in = results[0].Property.map(function(i) {return i.PropertyName;});
}
By adding this code you are filtering your directory by your current user and prefetching the Property relation table, then you set the filter only if your user canviewallRecords is false and use JS map function to create an array of the PropertyName field in the Property table. As I stated, your code may not be exactly the same depending on how you have to retrieve your user canviewallrecords property and then of course I don't know your relation between user and Property table either, is it one-to-many or other. But this should give you an idea how to implement this on server side.
Custom Search UI will be populated when user selects Complex asset in the Advance search screen drop down(apart from Folders,Contents) where 12 fields will be displayed .So when user clicks search button ,need to read those values and redirect to the alfresco repo files(org/alfresco/slingshot/search/search.get.js).We have already customized these files(search.get.js,search.lib.js) existed in the repository to suit out logic and working fine in 4.2.2;As we are migrating to 511,so we need to change this logic in customized faceted-search.get.js to read these values.How to write this logic in customized facted-search.get.js?
It's not actually possible to read those URL hash attributes in the faceted-search.get.js file because the JavaScript controller of the WebScript does not have access to that part of the URL (it only has information about the URL and the request parameters, not the hash parameters).
The hash parameters are actually handled on the client-side by the AlfSearchList widget.
Maybe you could explain what you're trying to achieve so that I can suggest an alternative - i.e. the end goal for the user, not the specifics of the coding you're trying to achieve.
We will be reading the querystring values something like below in the .get.js file.
function getNodeRef(){
var queryString = page.url.getQueryString();
var nodeRef = "NOT FOUND";
var stringArray = queryString.split("&");
for (var t = 0; t < stringArray.length; t++) {
if (stringArray[t].indexOf('nodeRef=') > -1) {
nodeRef = stringArray[t].split('=')[1];
break;
}
}
if (nodeRef !== "NOT FOUND") {
nodeRef = nodeRef.replace("://", "/");
return nodeRef;
}
else {
throw new Error("Node Reference is not found.");
}
}
It may be help you and we will wait for Dave Drapper suggestion also.
I'm writing a GUI extension and using the Anquilla framework to get a list of Keywords within a Category. I'm obtaining an XML document for the list of keywords then working with that document within my extension.
My problem is that the returned XML doesn't contain the Keyword's 'Description' value. I have the Title and Key etc.
My original code looks like this:
var category = $models.getItem("CATEGORYTCMID:);
var list = category.getListKeywords();
list.getXml();
A typical node returned is this:
<tcm:Item ID="tcm:4-1749-1024"
Type="1024" Title="rate_one" Lock="0" IsRoot="true"
Modified="2012-12-17T23:01:59" FromPub="010 Schema"
Key="rate_one_value" IsAbstract="false"
CategoryTitle="TagSelector"
CategoryID="tcm:4-469-512" Icon="T1024L0P0"
Allow="268560384" Deny="96" IsNew="false"
Managed="1024"/></tcm:ListKeywords>
So I've tried using a Filter to give me additional column information:
var filter = new Tridion.ContentManager.ListFilter();
filter.columns = Tridion.Constants.ColumnFilter.EXTENDED;
var list = category.getListKeywords(filter);
Unfortunately this only gives the additional XML attributes:
IsShared="true" IsLocalized="false"
I'd really like the description value to be part of this XML without having to create a Keyword object from the XML. Is such a thing possible?
cough any ideas? cough
I'm afraid you'll have to load the Keyword itself to get the Description.
It's not used in any lists, so it's not returned in the XML.
You could always create a List Extender to add this information to the list, but try to be smart about it since this extender will execute everytime a GetList is called.
Won't save you from having to open every keyword in the list, but you'll be doing it server-side (with Core Service/NetTcp for instance) which will probably be easier and faster than opening each keyword with Anguilla.
In this instance I only need the one keyword, so I simply get it from the CMS. Getting an object in Anguilla is a bit weird, here's the code:
In your main code area:
var selectedKy = $models.getItem("TcmUriOfKeywordHere");
if (selectedKy.isLoaded()) {
p.selectedKy = selectedKy;
this.onselectedKyLoaded();
} else {
$evt.addEventHandler(selectedKy, "load", this.onselectedKyLoaded);
selectedKy.load();
}
It's worth noting how I store the keyword in the properties of the item, so I can obtain it in the onselectedKyLoaded function
The function called once the item is loaded
ContentBloom.ExampleGuiExtension.prototype.onselectedKyLoaded = function (event) {
var p = this.properties;
var selectedDescription = p.selectedKy.getDescription();
// do what you need to do with the description :)
};
I resolved this, thanks to the answer here: https://stackoverflow.com/a/12805939/1221032 - Cheers Nuno :)
Following on from my earlier question about creating Address Books (many thanks Peter!), I have a small throw-away console application doing just that and working great - but in addition I'm trying to update the metadata of a Keyword with the Item Id of the created Address Book.
Slightly shortened snippet ...
StaticAddressBook ab = new StaticAddressBook();
ab.Title = title;
ab.Key = key;
ab.Save();
// id is a correct Keyword TCM ID
Keyword k = tdse.GetObject(id, EnumOpenMode.OpenModeEdit);
if (k != null)
{
k.MetadataFields["addressbookid"].value[0] = ab.Id.ItemId;
k.Save(true);
}
I keep getting the following error on Save():
XML validation error. Reason: The element 'Metadata' in namespace
'uuid:2065d525-a365-4b45-b68e-bf45f0fba188' has invalid child element
'addressbookid' in namespace
'uuid:2065d525-a365-4b45-b68e-bf45f0fba188'. List of possible elements
expected: 'contact_us_email' in namespace
'uuid:2065d525-a365-4b45-b68e-bf45f0fba188'
But I know the Keyword has the correct Metadata assigned, (thats why I don't bother checking!). Shortened Tridion XML from a current keyword in question:
<tcm:Keyword>
<tcm:Data>
<tcm:MetadataSchemaxlink:type="simple"xlink:title="IP.Location.Metadata" xlink:href="tcm:49-2142-8" />
<tcm:Metadata>
<Metadata xmlns="uuid:2065d525-a365-4b45-b68e-bf45f0fba188">
<email>...</email>
<addressbookid>3</addressbookid>
<contact_us_email>...</contact_us_email>
<request_a_sample_email>...</request_a_sample_email>
<webinar_feedback_email>....</webinar_feedback_email>
</Metadata>
</tcm:Metadata>
<tcm:IsRoot>true</tcm:IsRoot>
</tcm:Data>
</tcm:Keyword>
Have I missed something can Keyword metadata not be updated in this way?
I guess I could look at the Core Service to update Keywords, but it seemed to to make sense to do everything within this application.
UPDATE
Order was key here, strangely!
The following code works:
ItemFields fields = k.MetadataFields;
System.Diagnostics.Debug.WriteLine(fields.Count);
string email = fields[1].value[1];
string contact = fields[3].value[1];
string request = fields[4].value[1];
string webinar = fields[5].value[1];
fields[1].value[1] = email;
fields[2].value[1] = ab.Id.ItemId;
fields[3].value[1] = contact;
fields[4].value[1] = request;
fields[5].value[1] = webinar;
k.Save(true);
Got caught out by the non-0-based index when getting/setting values and had to reassign existing fields back, in order.
Cheers
It seems that the order of the fields has changed in the Schema since that Component was created. At least the Schema expects contact_us_email in the position where you current have addressbookid.
There may be other changes, so I'd verify the order of fields in the Schema and make sure the Component(s) match, before you run your tool.
I'm working on a web application that renders a reporting services report as a PDF and one of my requirements is to allow for multiple locations to be included on the same report. To handle this, I have created a multi-value report parameter which seems to do the trick when using the RS UI, etc. But, I am using the webservice for reporting services and cannot for the life of me figure out how to set the value of the parameter to be identified as having multiple values.
I've tried simply setting it as "LOC1,LOC2", but that is being picked up as a single value. I have also tried "LOC1" + System.Environment.NewLine + "Loc2".
You can send it through as a comma-delimited string if you're willing to parse it on the other end. A lot of languages have a String.Split(",") style method you can use for that.
Either that, or you can construct an array (or list, or collection) and pass that through as the parameter, though this would involve changing the contract on the webservice method.
Figured it out, you have to each value separately under the same name, snippet:
//Register parameters
ArrayList<ParameterValue> parmValues;
for(Map.Entry<String,String> entry:reportParams.entrySet()) {
//is it multi-value?
if(entry.getValue().contains(",")) {
//yes, add multiple ParameterValues under the same name
// with each different value
for(String mval:entry.getValue().split(",")) {
ParameterValue pv = new ParameterValue();
pv.setName(entry.getKey());
pv.setValue(mval.trim());
parmValues.add(pv);
}
} else {
//no, just a single value
ParameterValue pv = new ParameterValue();
pv.setName(entry.getKey());
pv.setValue(entry.getValue());
parmValues.add(pv);
}
}