Vertec search

With Vertec 6.8, we have introduced a full-text search function that searches emails (incl. attachments) and documents stored in Vertec in addition to texts. You can find a detailed technical overview in Full-text search and indexing.

Clicking on the magnifying glass in the navigation menu (or F6 and F12) opens the Vertec search page:

The following search fields exist:

Click on Search or Enter to start the search. The results appear below:

Double-clicking on a search result opens the relevant object.  

Documents and email attachments (Word/PDF) in which the search term was found are shown as links and can be opened directly.

For activities, the project and the contact, if available, are also shown. These are also direct links to the relevant object:

The classes in which the search term was found are displayed above the results list:

With the checkboxes, you can show and hide individual classes, which can be useful for long lists of results or if you are looking for something specific.

Search results found in an email or document are shown separately.

• For emails, the system setting Index subject, body and attachments (Word/PDF) of emails must be activated.
• For documents, the file storage must be set to Internal and the system setting Index Word/PDF documents (only for internal document storage) must be activated.

If there are activities with different activity types in the results list, a button appears called Filter by activity type:

You can filter the activities in the results list by type. The relevant filter is then shown next to the button:

Relevance

The results are sorted in descending order of relevance, which is calculated as follows:

For each search result, a so-called score is provided. This is calculated by the indexer and refers, among other things, to the number of hits in relation to the length of the text.

The absolute score is no longer significant for the display of relevance on the search page, because it always refers to the current search results. The result with the highest score (regardless of how high) is ranked at the top with 4 relevance points. The remaining results are ranked proportionally according to their score below. 

The search syntax

Search for multiple terms

If more than one word is entered as a search term, the words are searched with AND, so only results containing a combination of the words are shown.

This can be adjusted as follows:

• Exact search with quotation marks: Search terms can be enclosed in quotation marks. In this case, the entire term is searched. For example, the search for "Christoph Keller" only finds entries in which this term occurs. However, if you search for Christoph Keller, all the entries containing Christoph and Keller are found.
• Or search with OR: The Or search can be used to search for several terms without having to find all terms in the same entry. The search for Christoph OR Keller
  finds all entries in which either Christoph or Keller (or both) appear.
• Exclude terms with NOT: Certain terms can be excluded with NOT. The search for Christoph NOT Keller finds all the entries that contain Christoph, but not Keller.

Umlauts and special characters

Umlauts and special characters are ignored, in both indexing and in the search. For example, the search for “Müller” searches for “Muller” and thus finds both “Müller” and “Muller,”. The search for “Sørenson” finds “Sørenson,” “Sörenson” and “Sorenson”. This achieves the best results across all languages.

Stop words

A set of language-specific stop words that are not searched for, such as “and,” “in,” etc.

If a search term consists only of stop words, no results are shown.

Typographical errors

For typographical errors (typos), the Levenshtein distance is used If Activate similar words search is enabled. This means that a certain deviation is tolerated.

In the Vertec full-text search, the following rules apply:

• The search term must have at least three characters
• A mutation is allowed.

In English, the search for abcense (two characters swapped), absense (one character changed) or absnce (one character omitted), finds the term absence. In German, the search for Rechnugn (two characters swapped), Rechnunn (one character changed) or Rechung (one character omitted), finds the term Rechnung.

Word stem search

The word stem search means that, for example, the search for house also find houses.

The word stem search is always active automatically. It is language-dependent and is based on the language selected on the search page.

For example, the search for bikes, only finds bike if EN is the language selected.  

In addition to the default language, the languages defined as additional search languages in the system settings are available for the word stem search.

Configuring search fields

In the full-text search, all indexed data and documents are searched.  

All objects for which at least one configuration is defined are considered. For this, there is a subfolder indexed fields in the
class settings on each class:

In addition, you can find an overview of all index field definitions in the folder Settings > Full-text search:

Here, you can define additional index fields and include them in the search index. The index configurations shown here are supplied by Vertec as the default.

Note: Indexing of documents and emails is not configured here. It is defined in the System settings for full-text search.

The index field definitions have the following attributes:

• Index type: Specifies the data source and is one of the following values:
  • Field: The value is in a persistent database field. The index field must be a persistent member of the class.
  • Key: The value is in a key on the data set. The index field must contain the name of the key.
  • Custom fields: The value is in a custom field. The internal name of the custom field must be specified in the index field. This must be of type “Character," "ANSI Text" or "Text".
• Index label: String. Optional. The label shows the name with which the field value is written to the index. Therefore, it must be unique per object type (class).

  The label is used later when the search results are displayed on the search page to indicate the field in which the search term was found. If no label is set, the value index field is shown instead.

• Index field: String. Specifies the field name, depending on the selected index type.

If new index fields are defined, the indexing must be refreshed before corresponding results appear. This can be done with one of the Python functions or with the scheduled tasks.