Skip to main content

Search Syntax

Apache Lucene​

The Document Search endpoint uses a subset of the Apache Lucene query syntax. mEditor does not support the full Lucene search syntax, so we've detailed the supported subset below.

Terms and Phrases​

A single word is a Term, and a series of words is a Phrase.

  • OMI is a Term.
  • OMI-Validation Requirements is a Phrase.

Terms do not require quotation marks, but phrases do. Any of these are valid:


curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Creator:"Carl%20Creator"'
curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Creator:"Carl"'
curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Creator:Carl'

curl requires that spaces be URL-encoded as %20, but only the quotation marks " are of significance for this example.

Fields​

{field name}:{term or phrase}

In the prior example we ran the following query: ?query=Dataset_Citation.Dataset_Creator:"Carl%20Creator". The colon : denotes the separator between field name and the term or phrase you are searching for.

Dot Notation​

Given this partial data structure:

{
  ...other properties
  "Metadata_Dates": {
    "Metadata_Creation": "2016-10-27",
    "Metadata_Last_Revision": "2020-03-16",
    "Data_Creation": "2020-01-30",
    "Data_Last_Revision": "2020-01-30"
  },
  "Dataset_Citation": [
    {
      "Dataset_Publisher": "Goddard Earth Sciences Data and Information Services Center (GES DISC)",
      "Data_Presentation_Form": "Digital Science Data",
      "Persistent_Identifier": {
        "HasDOI": true,
        "Type": "DOI",
        "Identifier": "10.5067/E7TYRXPJKWOQ"
      },
      "Dataset_Title": "GLDAS Noah Land Surface Model L4 3 hourly 0.25 x 0.25 degree V2.1",
      "Dataset_Series_Name": "GLDAS_NOAH025_3H",
      "Dataset_Release_Date": "2020-01-30",
      "Dataset_Release_Place": "Greenbelt, Maryland, USA",
      "Version": "2.1",
      "Online_Resource": "https://disc.gsfc.nasa.gov/datacollection/GLDAS_NOAH025_3H_2.1.html"
    }
  ],
  ...other properties
}

You could search the Metadata_Last_Revision field with the following query:


curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Metadata_Dates.Metadata_Last_Revision:2020-03-16'

Notice that we navigate from Metadata_Dates down into Metadata_Last_Revision via Metadata_Dates.Metadata_Last_Revision. It is the . that enables that traversal, so we call this "dot notation." This also works when the data structure's field value resolves to a list of values, as does Dataset_Citation. To search for datasets released in Greenbelt, Maryland, USA, you could use this query:


curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Release_Place:"Greenbelt,%20Maryland,%20USA"'

Notice the use of quotes " around the Phrase, since it contains spaces.

Wildcard Search Operator​

Often you will need to search for a term or phrase that is not an exact match, and wildcard searches enable that. Using the previous data structure, you could search the Dataset_Release_Place with any of the following:


# matches Greenbelt plus any text
curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Release_Place:Greenbelt*'

# matches Greenbelt, Maryland plus any text (like "USA")
curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Release_Place:Greenbelt*Maryland*'

Notice the lack of quotation marks surrounding the phrases. Phrases may not use quotation marks and the wildcard operator, so the wildcard operator is used to "fill in for" the space character (and in this case, a space and a comma).

Logical Boolean Operators and Grouping​

You can use logical operators (AND, OR) in your queries and even organize them into logical groups with parenthesis (). From our previous search, we find that sometimes Maryland is abbreviated MD in our data. Again using the previous data structure as a reference, we could find the release place of MD or Maryland with the following:


# matches MD OR Maryland
curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Release_Place:*MD*%20OR%20Dataset_Citation.Dataset_Release_Place:*Maryland*'

# matches MD OR Maryland, AND must also be last revised 2020-03-16
curl '$MEDITOR_BASE_URL/api/models/Collection%20Metadata/search?query=Dataset_Citation.Dataset_Release_Place:*MD*%20OR%20Dataset_Citation.Dataset_Release_Place:*Maryland*%20AND%20Metadata_Dates.Metadata_Last_Revision:"2020-03-16"'

Though hard to see when surrounded by %20, the logical OR is %20OR%20 and the logical and is %20AND%20. %20 is how URL parsers encode the space character.