Operator | "" (quotation marks) |
Purpose | You can get more results than expected is you enter a search phrase (that is, two or more search terms meant to appear together), but do not enclose the phrase with quotation marks. To search for documents on 2018 sales, typically people would enter: 2018 sales In this case, the search results would include any document that contains both 2015 and sales, but not necessarily next to each other. |
Example | To search for documents containing the exact phrase 2018 sales, use quotation marks: "2018 sales" |
Operator | + |
Purpose | You can override the matching behavior using the + (plus) operator to search for exact words only. It is typically useful to search for: • link words (the, a, of, or, and) that are ignored by default, • the plural of a word. This operator is useful for building very specific queries. You can also prepend words by + in your query to search for the exact forms of these words only. For example, with the query foo +bar, foo has the standard semantic expansion (like lemmatization if activated) but not bar, which returns the exact form only (that is, bar). |
Operators | OR, AND, NOT, XOR, BOR |
Purpose | Searches for documents containing: • OR: either one search term OR another • AND: one search term AND another search term • NOT: one search term BUT NOT another search term • XOR: either one search term OR another BUT NOT both • BOR: either one search term OR another. Only use it for a fast OR on many documents (no expansion, no ranking). |
Example | Use OR to specify a list of similar terms that may occur in the document you are looking for. (movie star) OR (celebrities) searches for documents containing either movie star or celebrities. |
Operators | NOT, -XX, BUTNOT |
Purpose | Excludes documents containing a specific word or phrase from the search with a - (minus sign) or a NOT operator before the word to exclude. |
Example | new -york OR new NOT york searches for documents containing new but not york. Note: NOT and - are unary operators and depend on the implicit default operator AND. The expressions new -york OR new NOT york are therefore interpreted as new AND NOT york. You can also use the BUTNOT operator: "Martin Luther" BUTNOT "Martin Luther King" matches if there is at least an instance of Martin Luther not followed by King. |
Operators | Use prefix handlers: see the list of prefix handlers defined in Search logics > Query Language. |
Purpose | Refine your queries by targeting specific index fields with default prefix handlers like text:, title:, etc. You can also • specify aliases for these prefix handlers. For a list of aliases, see that prefix handler’s Alias field, in Search logics > Query Language. • search by category values • search numerical fields by a range of values • define custom prefix handlers to go further than the index field level, and trigger very specific search. For more details, see Using Prefix Handlers. |
Example | Search with a default prefix handler: title:foo searches for foo in document titles. Search with an alias: for the prefix handler document_file_size, you have the following aliases by default: file_size, imap_mail_file_size, nntp_post_file_size, ldap_record_file_size Search by category values: categories:fileattributes/extension/PDF Search a numerical range of values: NumericalPrefixHandler:[100 TO 200] Custom prefix handler: for a similarity search, we could enter a query like: similar: (ID1, ID2, ID3) where ID1, ID2, ID3 are the IDs of related terms, to search for all the documents having a part or all of these related terms. |
Operators | Prefix handler soundslike: You must create this prefix handler beforehand. For more details, see Using Prefix Handlers. |
Purpose | Finds documents using the phonetic spelling of search terms. Important: The language used for the query is important and must match the language specified in your Mashup UI configuration. If none is specified, Exalead CloudView uses the web browser’s preferred language. |
Example | To find a coworker with a name that sounds like Brona, enter: soundslike:brona to return results such as Bronagh and Branagh. |
Operators | Prefix handler spellslike: |
Purpose | Finds documents that do not exactly match the search terms. This is useful if uncertain of the correct spelling, or there are several accepted spellings for a search term. |
Example | Searching for spellslike:organise also returns documents containing organize. |
Operators | Prefix handlers date:, document_lastmodifieddate:, document_before:, document_after: |
Purpose | Retrieves documents based on a given date, or date range. By default, the input format is detected automatically. If you need to define a custom format, update the Input format field for your prefix handler in Search Logics > Query Language. What you must know: • We support the date formats: RFC 822, RFC 850, asctime, ISO 8601, and date format YYYY/MM/DD-HH:MM:SS (DD/MM/YYYY is NOT supported) • Operators are =, ==, <=, <, >=, >, != and : • The default timezone is GMT. • Quotes are required in search queries when there is at least a blank space in the date. For example, myDatePrefixHandler="12/15/2018 15:23:22 GMT+02" |
Supported formats | • Sun, 06 Nov 1994 08:49:37 GMT RFC 850: • Sunday, 06-Nov-94 08:49:37 GMT • Fri Nov 21 11:18:47 CET 2014 asctime: • Sun Nov 6 08:49:37 1994 RFC 822: • Fri, 21 Nov 2014 16:59:27 MET DST • Fri, 21 Nov 2014 17:59:14 EET • Fri, 21 Nov 2014 15:59:16 +0000 (UTC) • Fri, 21 Nov 2014 16:59:42 MET • Fri, 21 Nov 2014 15:58:04 +0000 (UTC) • Fri, 21 Nov 2014 07:58:28 -0800 American date format: • 12/23/2014 15:23:22 • 12/23/2014 15:23:22 GMT+02 • 09/23/2014 08:52:59 [+00:00] • 2014/12/23 15:23:22 • 2014/01/23-22:11:37 • 2014/12/23 • 2014/12 ISO 8601 samples (ISO works with / or - separators): • 2014-03-12 15:23:22 • 2014-03-12 • 2014-03 • 2014 • 2014-12-06T15:31Z • 2014-12-06T15:31:17+00:00 • Week numbers like 2016-W18-1T09:49:38Z are NOT supported |
Example | Let’s say that we give the modidied alias to the document_lastmodifieddate prefix handler. We could have: • modified="11/23/2018 10:18:02 GMT+01" for a fully explicit date query • modified="2018/11/23 10:18:02+00:00" for a fully explicit date query • modified="2018/11/23 10:18:02" for a date query with the default GMT time zone interpreted implicitly. • modified="11/23/2018 10:18" for a query with an implicit range of 1 minute. • modified=2018/11/23 for a query with an implicit range of one day. • modified=2018/11 for a query with an implicit range of 1 month. • modified=2018 for a query with an implicit range of 1 year. • modified<"11/23/2018 10:18:02 GMT+01" for all documents before the explicit date. • modified<"2018/11/23T10:18:02+01:00" for all documents before the explicit date. • modified<=11/23/2018 for all documents until the end of the 11/23/2014 day. • modified<=2018 for all documents until the end of the 31/12/2018 day. • modified:[2014/12/23 TO "2018/01/21-22:11:37 GMT+01"] to search documents in a specific date range. This range notation is inclusive, and works with numerical values too. We can also restrict a search query according to a document’s last modification or creation date: • "movie star" AND date >= 2018/05/21 finds documents containing movie star modified after May 21, 2014. • and "movie star" AND date <= 2018/05/21 finds documents on movie star modified before May 21, 2018. |
Operator | Prefix handler file_size: |
Purpose | Searches based on file size in bytes. |
Example | • file_size:1024 returns documents with a file size of 1 KB. • file_size>=1024 returns documents with a file size larger than 1 KB. |
Operator | Prefix handler language:XX |
Purpose | Limits your search to the documents of a specific language using the language:XX prefix handler (where XX can be EN, FR, DE, etc.). This is useful when you need to search using a term that you can find in many languages, but has different meanings from one language to another. |
Example | "Tour de France" language:en searches for English-language documents about the Tour de France. |
Operator | Prefix handler inurl: |
Purpose | Includes all web pages with URLs containing the search keywords. Unlike site:, this is a full text search of the URL text. |
Example | inurl:example returns: • http://www.example.com/ • http://www.exalead.com/blog/another_cloudview_example/ |
Operator | Prefix handler url: |
Purpose | Searches for pages with the same normalized URLs. You do not need to include the leading http://, https://, www., and trailing slashes in the query. |
Example | url:example returns: • http://www.example.com/ • http://www.exalead.com/blog/another_cloudview_example/ |
Operator | Prefix handler site: |
Purpose | Returns all documents on a site. Only expect results for documents with a publicurl meta, such as those pushed by the Crawler and the Feed Fetcher connectors. The leading "http://" or "https://" and "www.", and trailing slashes are optional in the query. |
Example | site:example.com always returns the same documents as site:http://www.example.com/ |
Operator | OPT |
Purpose | Specifies an optional word to include in the search. Use it to specify several terms without limiting the scope of the search. |
Example | cow OPT mad searches for documents containing cow that preferably also include mad. |
Operators | NEAR, NEXT, AFTER, BEFORE |
Purpose | Find documents where search terms are in proximity of one another. By default the maximum distance between terms is 16 words. Edit this value using the Search > Search Logics > Query Language > Default distance for proximity operators property. |
Example | "movie star" AFTER hollywood searches for documents where movie star appears soon after hollywood. Note: "movie star" is equivalent to movie NEXT star, the NEXT operator having a distance of 1 with the following word. You can also specify the maximum distance of the words by using NEAR/x, AFTER/x, and BEFORE/x. For example: • "movie star" NEAR/5 hollywood searches for documents where movie star appears within 5 words of hollywood, • and "movie star" BEFORE/5 hollywood searches for documents where movie star appears within 5 words before hollywood. Important: • You cannot use proximity operators with expressions whose "position" cannot be computed. For example, the query music NEAR (Madonna AND mp3) does not work, because the expression Madonna AND mp3 cannot be associated with a single word position value. • Some queries using proximity operators may fail with a No occurrence for query message when you want to open the preview of Office documents. This issue is linked to a format conversion limitation. |
Operator | * |
Purpose | Searches using the beginning of a word to find a proper noun using its short form, or its linguistic root. |
Example | Jenn* searches for documents containing words starting with Jenn, such as Jennifer, Jennie, Jenni, and Jenna. |
Operator | Regular expression patterns based on Perl 5. You must open and close patterns with a / (slash) character. |
Purpose | Searches using the beginning of a word to find a proper noun using its short form, or its linguistic root. |
Example | • /s.ren..pi.y/ searches for documents with words that match the pattern S . R EN .. PI . Y and would find documents with the word serendipity. • /mpg(1|2|3)?/ searches for documents containing any of the following: mpg, mpg1, mpg2, or mpg3. |
Operator | Prefix handler geo: |
Purpose | |
Example | To search within a radius or polygon using UQL, see Search with a Radius or Polygon (UQL). |
Operator | INNERJOIN |
Purpose | Combine records from two documents whenever there are matching values in a common field. See More About INNERJOIN. |
Operator | SPLIT |
Purpose | Searches for words in specific sections of a document. |
Name | Character |
|---|---|
Slash | / Use it for: • passing options, for example, NEAR/12 • pattern search as a regexp operator, for example, text:/bug.*/ |
Tab | \t |
Line feed | \n |
Carriage return | \r |
Round brackets | ( or ) |
Square brackets | [ or ] |
Curly brackets | { or } |
Colon | : |
Equal sign | = |
Greater than or less than | < or > |
Comma | , |
Operand | Description | Predicate value |
|---|---|---|
(e1) | Parenthesized sub expression, used to modify priority. For example: ((fast OR speed) AND NOT light) | e1 |
"e1" | Quoted expression, used to escape all special characters. Inside a double quoted group, words are handled in a tight (NEXT) sequence. All operators are ignored. | e1 |
"word1 word2" | Quoted expression. | word1 NEXT word2 |
"word1 OR word2" | Quoted expression | word1 NEXT "OR" NEXT word2 |
Kind | Syntax | Example |
|---|---|---|
Regular expressions | [field:]/pattern/{options} | title:/desi.*/{w=10000,#=100} |
Wildcard | [field:]word*{options} [field:]*word{options} | desi* title:*faces |
Operand | Description | Predicate value |
|---|---|---|
s=N | Replace the predicate's score by an explicit value. | price<500{s=1000} |
s+=N | Increase the predicate's score by a given value. | GUI{s+=100000} |
s-=N | Decrease the predicate's score by a given value. Note: The score of a given predicate can be negative, but the final score of the document can never be lower than 0. | corporate/tree:"Top/Attributes/XXX" {s-=100000} |
w=N | Replace the predicate's weight by an explicit value. | design{w=10000} |
w*=N | Multiply the predicate's weight by a given value. | design{w*=2} |
Option | Description | Example |
|---|---|---|
k=number | Set explicit matching level. | k=1 is the lowercase matching mode k=2 is the normalized matching mode. |
hl=0 | Deactivates search result highlighting and summary for a specific node of the query only. | word1 word2{hl=0} word3 |
Priority | Operator | Explanation | Example |
|---|---|---|---|
1 | prefix handlers | Prefix handlers are always processed first. | Obama before:2009/01/01, searches for all documents relating to Obama before January 2009. |
2 | FUZZYAND/option (expression) | Search for documents that match at least N queries, where N is determined by the fuzzyand option. This option can be either: • minimum success: at least N queries must match. N is a positive integer. • or maximum failures: up to N queries can fail. N is a negative integer. | For a document that contains: The quick brown fox. If min. success=2: FUZZYAND/2 (the quick brown foxes) matches, but FUZZYAND/2 (a brown foxx) does not. If max. failure=-1: FUZZYAND/-1 (a quick brown fox) matches, but FUZZYAND/-1 (a quick foxx) does not. |
3 | OPT e1 | Optional operator | OPT graphical |
4 | NOT e1 | Negation operator | NOT myword |
5 | e1 NEXT e2 | Explicit sequence operator for adjacency match. | user NEXT interface |
6 | e1 AFTER e2 | Proximity (mono-directional) match | interface AFTER user |
6 | e1 AFTER/distance e2 | AFTER with explicit word distance | interface AFTER/4 user |
6 | e1 BEFORE e2 | Proximity with mono-directional match | user BEFORE interface |
6 | e1 BEFORE/distance e2 | BEFORE with explicit word distance | interface BEFORE/4 user |
7 | e1 NEAR e2 | Proximity operator with bidirectional match | user NEAR interface |
7 | e1 NEAR/distance e2 | NEAR with explicit word distance | user NEAR/4 interface |
8 | e1 SPLIT e2 | A document is returned if e1 appears in at least one of the document sections delimited by the e2 delimiter. If the e2 delimiter is not present in a document, then the document is returned if e1 is valid at the document level. | user interface SPLIT Chapter Matches if user interface appears between two "Chapters" |
9 | e1 e2 | Implicit match operator on a sequence of words. It uses the implicit operator, which is AND by default. | search engine |
10 | (e1) INNERJOIN/key (e2) | Search for documents matching e1 where e2 appears in child documents. The relation between documents is contained in a key index field, which must be an unsigned integer. For performance reasons, it is best to enable the Stored in Memory option. | subject:exalead INNERJOIN/msgId fulltext:france We first select documents whose subject is exalead, then the join is made with documents containing the word france. |
11 | e1 BUTNOT e2 | The search matches if there is at least an instance of e1 is not also an instance of e2 at the same position. | York BUTNOT "New York" |
12 | e1 AND e2 | Explicit conjunction match. | user AND interface |
13 | e1 XOR e2 | Exclusive OR operation. It can be either e1 OR e2, but not e1 AND e2 | design XOR conception |
14 | e1 OR e2 | Disjunction match | design OR conception |
15 | e1 BOR e2 | Disjunction match To use only for a fast OR on many documents | design BOR conception |