How to specify FULL reindex locale ?

As we transition to the new Hyland Community, some post links may be temporarily disrupted as we finalize redirects. We apologize for the inconvenience and are working to resolve this as quickly as possible. Details.

pcuvecle

Hi all,

I'm running Alfreco on a Linux Suse and I face a new problem after reindexing all the content (bean indexRecoveryComponent in recoveryMode=FULL) because data has been indexed by tokenising according to an english locale although all users use a french locale. As a result I'm unable to search anything in the web client (but it works if I switch the GUI to english)

Is there a way to specify what locale must be used during a full reindexation ?

WIKI extract
As of version 2.0 all uploaded content is tokensied according to the users locale. (It is not yet possible to specify locale on upload). At search time, the users locale is used for tokenisation. This is not currently picked up from the search parameters or cna be specified in the query.

A search is performed in the users locale except for MLText attributes which can be found in other locales as specified on the serach parameters. Content will also be found in the locales specified on the search paremeters if the tokens generated match thoses in other languages. The content part of the query only generate tokens in one language but thoses tokens are looked for in all the languages specified on the search parameters.

andy

Hi

So I will correct the typos …

The locale is set for content as it is loaded by the user. Each bit of content has an associated locale. For old data this may not have been set. At index time the locale is used to select tokenisation. It will use the locale set on content, falling back to the locale set on the node and then the Java/JVM default locale. You can set the default locale as options to the JVM.

When the text is tokenised the config can affect what tokens are generated in two ways, First, the tokeniser controls what tokens are generated, what are skipped as stop words, what are stemmed etc. Secondly, the locale support can generate tokens that can be used for different langauage/locale/varient matches. By default a locale specific token and a cross language token are generated for each token. Have you changed the configuration for this?

At search time the tokens generated from the search string are generated in a similar way. By default, the search generates only cross language tokens (it ignores the locale) and tries to match across all content. Again the default is in the config and can be set via the Java search API.

So, have you changed any of the tokenisation config?

If you are only working in French, you can probably solve your probelm by setting the JVM locale.

The index recovery should rebuild the index exactly as before (with the same config) - its tokenisation should be consistent.

Andy