Data Catalog Search syntax reference
Azure Data Catalog is a fully managed service hosted in Microsoft Azure that serves as a system of registration and system of discovery for enterprise data sources. Azure Data Catalog has capabilities that enable technical and non-technical users to discover, understand, and consume data sources.
A key aspect of data discovery is the ability to search for data sources that have been registered in Azure Data Catalog. Azure Data Catalog has a powerful search syntax that enables users to easily build queries that return the data the users need.
Search Syntax Overview
Azure Data Catalog searches is similar to that used by Microsoft Windows and Microsoft Outlook, and which should be familiar to users of these tools.
|Basic Search||Basic search using one or more search terms. Results are any assets that match on any property with one or more of the terms specified.||
|Property Scoping||Only return data sources where the search term is matched with the specified property||
|Boolean Operators||Broaden or narrow a search using Boolean operations||
|Grouping with Parenthesis||Use parentheses to group parts of the query to achieve logical isolation, especially in conjunction with Boolean operators||
|Comparison Operators||Use comparisons other than equality for properties that have numeric and date data types||
Matching, Comparison, and Boolean Operators
||Use property scoping and return only those assets where a given property contains the text being searched. The semantics for the query are "prefix match".|
||Allows the user to specify an exact match. Only those assets that contain the property with exactly the value of the search term will be returned.|
||"Not equal to" operator. Will return only those assets that don't have the value indicated in the search query.|
||Finds items that contain the exact phrase social security. There is one special case to using quotes. If quotes are used with property scoping the semantics are grouping but not exact phrasing. In this case the behavior is the same as specifying the named property twice. Example: name:"social security" finds any assets that have a name property with the word social in it or a name property with the word security in it.|
||Finds items that contain tag1 and tag2 or have the name sales in database salesfy15. Typically used in conjunction with boolean operators|
||Finds items with a modified date after 11/05/2014.|
||Finds items with a date before 11/05/2014.|
||Finds items that contain social, but not security.|
||Finds items that contain social and security.|
||Finds items that contain social or security.|
||Allow filtering and return only those assets where a given property is set (or if the property represents a collection - it holds at least one element).|
By default, all of the searches in Azure Data Catalog are done using a technique called Prefix Match Semantics. This means that any search term starts a match at the beginning of the asset's properties.
As an example, consider two fictional assets registered in Azure Data Catalog with the following names:
- Salesman Quotes
A search for "sales" returns both of these assets, since their names both start with the word "sales". Future releases of Azure Data Catalog include support for exact match operators.
Property Scoped Searches
Azure Data Catalog query grammar supports property scoping. In the current preview, the property scopes are case-sensitive. That means that in order for the query to work, the actual casing of the property in the search query must match what is in the index.
Searches on invalid properties (properties that don't exist) result in an error.
Quotes behave in a special way when using property scoping. Quotes in any other context indicate exact phrasing. However, when quotes are used in property scoping the semantics are grouping. For example, name:”Sales Products” does a free text search looking on the content of the name property looking for “Sales” or “Products”. So the semantics of:
name:”Sales Products” is exactly the same as name:Sales name:Products
The general principle for property names in Searchable Properties is camel-case, which means that first letter is lower-case, and then each of the word first letters are uppercase.
The most useful properties are listed below.
|name||Finds items where the search term appears in the data source name||
|description||Finds items where the search term appears in the data source description||
|objectType||Finds items of a specific object type, such as table, view, or KPI||
|sourceType||Finds items of a specific data source type, such as SQL Server or SQL Server Analysis Services Multidimensional||
|tags||Finds items where the search term appears in the data source tags||
|timestamp||Finds items based on the date and time their metadata was most recently modified||
|lastRegisteredTime||Finds items based on the date and time their metadata was registered||
|friendlyName||Finds items where the search term appears in the data source friendly name||
|experts||Finds items where the data source expert matches the search term||
|termName||Finds items where the search term appears in business glossary terms applied to the data asset||
You can also use the following property names along with "has:" filter to check where assets have specific properties set.
|previews||Finds items that contain preview||
|documentation||Finds items that contain documentation||
|tableDataProfiles||Finds items that have a table profile (size, number of rows, etc.)||
|columnsDataProfiles||Finds items that have a column data profile (number of distinct values, min, max, etc.)||
The following sections show a few Search examples.
Return all assets with "sales" in the name
Return all assets registered after 4/20/2015 that include "sales" in any property
sales AND lastRegisteredTime>"4/20/2015"
Return all assets that include sales in any property, and which do not have the Q1FY2013 tag
sales AND tags<>"Q1FY2013"
Return all assets that don't have experts nor documentation assigned
not has:experts and not has:documentation