Quantcast
Channel: Optimizely Search & Navigation
Viewing all 6894 articles
Browse latest View live

Problems with synonyms

$
0
0

Hello!

We have just began rolling out Find for one of our customers. While working locally in our dev environments (using dev indecies) synonyms have been working without flaw. However, once we started rolling out the changes to the test and subsequentially the pre-production environment at our customers hosting, the synonyms seemingly stopped working. They can be added in the Find UI without any apparent errors, but they do not affect the search results in any way.

Since the code is the same, we suspect either:

  • Something in the configuration or type of index running in test and preproduction or;
  • Something in the environment that prevents the synonyms API from functioning properly

We are not able to run Fiddler or anything similiar in the hosting environments. 

We are running:

  • EPiServer           9.10.2
  • EPiServer.Find    12.2.4

Any solution or idea for debugging this is greatly appreciated!


'Preview the boosting effect' shows only undefined values

$
0
0

Hello!

We used Episerver Find to implement search functionalities in our project. The problem is that if we go to Find -> Configure, in 'Preview the boosting effect' we see only records with undefined,undefined. Please see the following image: https://www.screencast.com/t/9xxzYiEK.

Does anyone know what might be the problem? The version of Episerver Find we used is 12.5.2.

Thank you!

How Does Find Work WIth Non-Standard Locales

$
0
0

I'm working on a  project with Find and Commerce.  The client is suggesting for us to use non-standard locales within their website to help them manage content, does anyone know if find can cope with non-standard locales?

My recommendation would always go with an ISO standard, why reinvent the wheel right?  However, what I think is good and what I end up sometimes writing can be different.  The client has a third-party PIM where this approach makes sense if you come at it with a PIM first approach.

Let's say we have a website with three languages enabled:

EN - global English

Iran - ir-en (This is a non-standard locale, they have made up). 

Iran - fa-ir

This is so they can have a global English site, an Iran site in Iranian a specific Iran tailored English site, in this approach we use locale as a market identifier, rather than an actual locale, e.g. it's a way of not having to create subsites for different languages.

In commerce is I try and add a market with 'ir-en' an invalid locale, if the locale is not set on the server, an exception is thrown.  I then have to run some powershell script to regsiter it.

Will find be able to cope with this?  Obviously, find won't be able to run this powershell script?  In the background does find create a locale object on my content, or, does all the locale look-ups work based on a string identifier?

If I create content in find, 

Text in PDF body indexed but not found when searching

$
0
0

I have the following problem when searching for a term which only occurs in the text of a PDF file.

The text in PDF files is indexed, i.e. "SearchAttachmentText$$string" is filled with the text of the PDF. When I search for a term which only occurs in the SearchAttachmentText from the Find > Overview interface the item with the term is found.

When I search for the same term from the Find > Configure > Boosting interface the item is not found. Also, when searching from the website front-end, the item is not found.

If I search for the item based on words in the title, the item is found, and I can also see the term on which I initially searched in the excerpt.

Can anybody indicate what might be the cause of this problem?

Click-through rate resetting percentage to 0%

$
0
0

"Searches without relevant hits. Searches that generated hits but had a low click-through rate. Phrases that are most frequently searched for and for which search results are clicked the least appear at the top."
http://webhelp.episerver.com/17-2/find/statistics-input.htm

A customer is wondering why the search statistics displayed at "Searches without relevant hits" does not seem to update correctly compared with what they see in Google Analytics for search queries. I have tried to understand how it works and in that solution as well as in another customers Find implementation the percentage does not seem update correctly.

I tried this again today and made a search for phrase that I saw had 20% the last month. After doing the search the number of searches was updated from 5 to 6 but the percentage was updated to 0% from 20%. The search did not find any hits at all. Does that affect the percentage or is it a bug in Find? I made another attempt at a phrase that had 25% and 4 searches. After making a search and clicking a result the number of searches was updated to 5 and the percentage again updated to 0% instead of affecting the percentage positively (at least that's how I understand how it should work).

From my search on world.episerver.com and Google in trying to understand this issue, what I can see is that people are referring to a script that should be added to the page and values that should be added to the querystring. It might be that this has worked earlier but have for some reason stopped working in this customer's site.

On https://world.episerver.com/documentation/developer-guides/find/search-statistics/tracking/ it says that "When using Unified Search, Track() enables tracking of a query and its hits. When using Track() on a non-Unified Search, only query tracking is enabled. Use custom tracking (explained below) to track hits with a non-Unified Search." We have not changed the implementation to use TrackHit or TrackQuery so I am wondering if this has changed in a new version and the click-through rate has stopped working because of that?

Have anyone else found this issue? What could be the cause of this?

Indexing content greater than 50MB in size

$
0
0

Hello,

We have a requirement in our application to index some files which are greater than 50MB in size.

Please let me know if there is any workaround to do this.

Trying to optimize search query.

$
0
0

Hi,

I have a couple of documents that refer from other pages. Is possible to make one search query in Find that filters out documents that are not referred from other pages?

 

What we do today is search among all the documents we have on the web site. After we get the search result we then check each of the documents to see if it is referred in a property on all the pages on the web site.

Performance vice, this does not work well when we get many hits from the search engine at same time as the page tree is big.

How to index content from custom media provider

$
0
0

Hi,

I've created a custom media provider, but the media files are not being indexed. Is there something I need to implement for this to work?

I'm using EpiFind version 12.4.3.0 and CMS version 10.10.4.0


Exception: mapping for explicit nested path is not mapped as nested

$
0
0

Hi,

I’m having difficulties getting a nested value into my Find index.
I have a Commerce Variant with a property that exposes an IEnumerable<Price>:

        public IEnumerable<Price> AvailablePrices => this.Prices();

where this.Prices() is the default IPricingExtensions method. (I’ve also tried using Prices() directly but getting the same error)

In the conventions I have:
SearchClient.Instance.Conventions.NestedConventions.ForInstancesOf<DefaultVariation>().Add(x => x.AvailablePrices);

This gets indexed fine:

"AvailablePrices": [

        {

            "CustomerPricing": {

                "$type": "Mediachase.Commerce.Pricing.CustomerPricing, Mediachase.Commerce",

                "___types": [

                    "Mediachase.Commerce.Pricing.CustomerPricing",

                    "System.Object",

                    "System.IEquatable`1[[Mediachase.Commerce.Pricing.CustomerPricing, Mediachase.Commerce, Version=10.8.0.0, Culture=neutral, PublicKeyToken=6e58b501b34abce3]]"

               ]

            },

            "___types": [

                "EPiServer.Commerce.SpecializedProperties.Price",

                "System.Object",

                "System.ICloneable"

            ],

            "ValidFrom$$date": "2013-12-05T23:00:00Z",

            "$type": "EPiServer.Commerce.SpecializedProperties.Price, EPiServer.Business.Commerce",

            "UnitPrice": {

                "Amount$$number": 210,

Still, when querying I get:

nested: ElasticSearchIllegalArgumentException[mapping for explicit nested path is not mapped as nested: [AvailablePrices]];

 this is called in:

query = query.OrderBy(v => v.AvailablePrices, p => p.UnitPrice.Amount, price =>
     price.ValidFrom.LessThan(DateTime.Now)
     & (price.ValidUntil.Exists() | price.ValidUntil.GreaterThan(DateTime.Now))
     & (price.MarketId.Value.Match("BBNL")));

Episerver Find version: 12.5

Hope anyone can help! 

/Mark

Best way to index multiple attachments?

Unified Search - Statistics Tracking results in error

$
0
0

Hi,

I am using Unified Search and I want it to appear in the Statistics, so added the Track() method.

IEnumerable<UnifiedSearchHit> _SearchResults = searchClient.UnifiedSearch().For(queryString)
.InField(x => ((BasePageData)x).MetaKeywords)
.InField(x => ((BasePageData)x).MetaTitle)
.InField(x => ((BasePageData)x).MetaDescription)
.Track()
.Filter(x => !x.MatchTypeHierarchy(typeof(ImageData))).Filter(y => !y.MatchTypeHierarchy(typeof(ContainerPage)))
.Take(10).Skip((p - 1) * 10).GetResult(hitSpec, false);

The GetResult method in above statement throws exception "String reference not set to an instance of a String.\r\nParameter name: s"

Stacktrace: 

   at System.Text.Encoding.GetBytes(String s)
   at EPiServer.Find.TrackContext.HashString(String toHash)
   at EPiServer.Find.TrackContext..ctor()
   at EPiServer.Find.SearchExtensions.GetProjectedResult[TResult](ISearch`1 search, SearchContext context)
   at EPiServer.Find.SearchExtensions.GetResult(ITypeSearch`1 search, HitSpecification hitSpecification, Boolean filterForPublicSearch)

If I remove the Track() method, it works fine.

But I need to gather the statistics. Can't figure what I am doing wrong. Please help.

Thank you.

Having trouble with a filter

$
0
0

When searching I only want to get pages in specific branches of the page tree.

This works fine:

myQuery.Filter(x => x.Ancestors().Match("48"))

But shouldn't this give the same result?

var pageBranchFilter = SearchClient.Instance.BuildFilter<IContent>();
pageBranchFilter.Or(x => x.Ancestors().Match("48"));
myQuery.Filter(pageBranchFilter);

The second approach is completely ignored when doing the search.

How to display metadata from file in unified search?

$
0
0

I'm building a EPiServer 8.5 + Find 9.2 site and I wonder how I can get at the metadata embedded in the uploaded files (pdf, doc, docx etc). The site has public search function that uses unified search, and for the MediaData hits I would like to display the title metadata from the file itself in the serp. By default UnifiedSearchHit.Title contains the filename.

I suppose I could implement SearchTitle and extract the metadata myself, but that seems silly since Find/Elastic is already doing that (I can find the file in the index explorer by searching for the metadata title).

I'm hoping there's a simple way to get the metadata!

Episerver Find in a separate project

$
0
0

Hi everyone,

I am wondering if it is possible to have Episerver Find service in a separate project and not in the same project where Episerver CMS is referenced.
I am trying to keep responsibilities apart and code in a modular way.
I tried to create a new project (Class library) and added Episerver.Find references, then I edited the Web project Web.config, but the Find tab doesn't show up in the Editor view. Probably it is because during Episerver.Find installation some UI files are modified.

Based on your experience is there a way to separate them or is it not possible right now?

Thank you in advance.

Indexing extra properties for Episerver Find and retrieving information from the index

$
0
0

Hello everyone,

I am currently using an extension method to index extra properties in the index for the website I am working on. I have verified that this is being indexed correctly in the index.

I am wondering how I would filter against this and return it to an API. Currently, the only suggestion I have seen is to call the extension method again to get the information needed, but ideally, I would like to return this from the index rather than calling the logic to get the required information again.

Is there any way to get this information from the index again rather than calling the same method to work out the information?

Thanks 


EpiServer Find Service security documentation and workflow

$
0
0

Hello,

Does anyone know of or have any documentation on the Find cloud service architecture and workflow? I know there are overall documentation on the DXC architecture and how it has WAH,IDS ect for the CMS itself, but when i look at the documenation and diagrams, Find shows as a black box outside of the Azure environment.

Is this a mistake, is there a more detailed diagram on how that functions in the cloud and the security around it?

Thanks.

Multi-words synonyms and WithAndAsDefaultOperator

$
0
0

I have an issue where multi-word synonyms have to work together with WithAndAsDefaultOperator, but I found that it doesn't:

https://support.episerver.com/hc/en-us/articles/115004105823?input_string=episerver+find+synonym+do+not+work

In the resolution section, it is mentioned that is going to be done in some future version and the link to the issue. Unfortunately, the issue is not available anymore.

Can anyone from Episerver take a look where the reported issue disappeared? And when this is going to be fixed?

Text in PDF body indexed but not found when searching

$
0
0

I have the following problem when searching for a term which only occurs in the text of a PDF file.

The text in PDF files is indexed, i.e. "SearchAttachmentText$$string" is filled with the text of the PDF. When I search for a term which only occurs in the SearchAttachmentText from the Find > Overview interface the item with the term is found.

When I search for the same term from the Find > Configure > Boosting interface the item is not found. Also, when searching from the website front-end, the item is not found.

If I search for the item based on words in the title, the item is found, and I can also see the term on which I initially searched in the excerpt.

Can anybody indicate what might be the cause of this problem?

NestedConventions Usage and performance advice

$
0
0


In our current find index, we have products at the root level, and variants under products. We do not have any Nested Conventions enabled. There are about 30K
products, each with a few variants (many with only one variant).

Current setup:


ProductIndexModel

List<VariantIndexModel>
[~70 basic properties]

VariantIndexModel


[~20 basic properties]

We now have a requirement to be able to filter by and facet on up to 10 "technical attributes" at the variant level.
We need to be able to filter on technical attribute combinations ... e.g. (TA1 = "A" and TA2 = "B"), and also combinations within a technical attribute, e.g. ((TA1 = "A" or "B") and TA2 = "C")

It seems like the best (only?) way to accomplish this is with Nested Conventions and the NestedTermsFacetExtensions extension.

Modified index models would look like:

ProductIndexModel


List<VariantIndexModel>
List<TechnicalAttributeIndexModel> [rollup of all TechnicalAttributeIndexModels from all of this product's variants]
[~70 basic properties]

VariantIndexModel

[~20 basic properties]
TechnicalAttributeIndexModel (10 string properties (e.g. TA1, TA2, TA3 ... TA10))

We would add a nested convention, like this:


Conventions.NestedConventions.ForType<ProductIndexModel>().Add(p => p.TechnicalAttributes);

("TechnicalAttributes" is the name of the List<TechnicalAttributeIndexModel> in ProductIndexModel)

I can see that this setup more than doubles the time to run a full index job (I guess due to all the new "hidden documents" that get created), but what should
I expect in terms of query performance ? Behind the scenes is this similar to a SQL join that will perform well, or should I be concerned because the index is now
much larger ?

Would love to hear from someone with knowledge of how Nested Conventions works under the hood, or with real world experience using it in a solution.

Thanks!!!

Issues with new developer indexes

$
0
0

Are there issues with newly created developer indexes? I receive "(401) Unauthorized" errors when indexing and also received nested "IOException[No space left on device]" while indexing last Friday. I have an old index that works just fine on the very same project. I'm confident I have the correct configuration using the correct private url.

Viewing all 6894 articles
Browse latest View live