Index categories

Proximity Operations (NEAR and WITHIN operators)

Search syntax (Basic searches, Boolean and proximity operations, combined searches)

Truncation, wildcards and literals

Pseudo Indexes table

Words Indexes table

Fields indexes table

Our STAR WEB interface to the ESTC gives you almost full access to a whole series of indexing capabilities, beyond the simple ones that are represented by the search page options. Those rich features can be exploited by checking the Expert search box by the Keyword search line. Once you have checked the Expert search box, you can formulate complex Boolean and proximity searches that call upon combinations of dozens of indexes.

While more flexible, expert searching is more complex and requires a detailed knowledge of both existing indexes and search syntax for our system. It is to be used as a last option, after the simple search strategies have failed to produce the result you expect. However, reading this page can give you useful insights and tips that can sometimes save a lot of time and effort. It is up to you to find the right balance for your needs.

Index categories

There are three main categories of indexes available:

• Pseudo indexes actually combine together several pre-existing indexes.  In other words, they allow you to trace data across multiple fields: title, author, subject, etc.
• Fields indexes are actual indexes that records strings from the beginning of a field, or the beginning of a subfield.
• Words indexes are indexes that record the location of words, regardless of the position of that word in a field or subfield

Real indexes (i.e. Fields and Words) are defined as a combination of input field and subfield source, basic indexing rule, and proximity options. In the case of subfields, it is possible to create an index with a selective range of subfields, for instance only subfield a, or subfields a and b, leaving out subfield c. Pseudo indexes are dependent upon the real indexes that compose them. For isntance, if you have a pseudo index called "Title", that index may depend upon two other indexes, one that indexes the subfield a of the 245 using the Fields basic rule, and one that indexes every word in the 245 regardless of subfield location.

I provide tables below (see ESTC Pseudo Indexes, ESTC Words Indexes, ESTC Fields Indexes) detailing each available index for our interface to the ESTC as a reference.

As a matter of convention, index names that end with a "W" (like 100W) are word indexes, and the ones that end with an F are field indexes (600F). Additionally, index names that end wth "ALL" often refer to a pseudo index that combines both word and field indexing. However, it is always best to refer to the tables to be sure.

Return to the top of the page.

Proximity Operations

In addition to the basic indexing rules, an index can be modified to provide the user proximity operations. This allows you to specify for instance that you are searching for two terms in the same occurence of a repeating field. The index can even be set to keep track of terms that are within the same sentence or paragraph, or even at a certain distance from each other. In the list of options given in the index tables below (see ESTC Pseudo Indexes, ESTC Words Indexes, ESTC Fields Indexes):

• ADJ indicates that you can specify how close or NEAR you want a term to be from another,
• S means that you can indicate that two terms should be in the same sentence,
• O means that you can chose to search two or more terms that are contained in the same occurence of the field.

Return to the top of the page.

Search syntax

Our STAR interface supports a rich and flexible range of searching strategies and options that allow you to use the full capability of the indexing options detailed above.

ALL OPERATORS ARE CASE INSENSITIVE: however I will use capital letters below to make the syntax clearer

Basic search

A basic search can done by giving the name of the index, followed by an equal sign, followed by the search term itself:

Example: 100ALL = Voltaire
Example: 100F=Milton, John

BOOLEAN syntax

Basic searches be combined according with the familiar BOOLean operators: AND, OR, AND NOT and OR NOT. These operators can be used either with the same index or with different indexes:

Example: 100ALL=Milton AND John
Example: 100F=Milton, John* AND PUBCITY=Paris
Example: 100ALL=Milton AND TI=(Paradise OR regained)

Proximity searches

The most common proximity operator is the NEAR operator. This is equivalent to the BOOLean AND opertor together with the restrictions included in the definition of the index itself. For instance, the 650 fields index is with the field "Occurence" options, that indexes every subfield starting at the beginning.

Example: 650=catholic* NEAR apologetic*

In this case, the search will retrieve all records that have at least one occurence of the 650 input field that contains two subfields one beginning with "Catholic", the other with "apologetic".

Example: 500=wing NEAR deprecated

The 500 is a words index, that has the options ADJacent, and Occurence: This will retrieve a record in which the two words occur in the same occurence of the 500 data field.

If you want to specify that the two words "wing" and "deprecated" should be not only contained within the same occurence, but also close to each other, in the same sentence, you would use the syntax:

Example: 500=wing W/S deprecated

The W/S operator means "within the same sentence".

If you want the two words to be not only in the same sentence but no further from each other than 4 other words, you would use:

Example: 500=wing W/4 deprecated

In this case, W/4 means "within four words". You can replace the 4 with any number: w/2, w/9, etc.

Proximity and BOOLean operators can be combined:

Example: 100ALL=(Milton NEAR John) AND TI=((paradise w/2 regained) OR (paradise W/S lost))
Example: TI=(paradise NEAR lost) AND NOT 100ALL=(Milton OR Voltaire)

As shown in the last example, parentheses can be used to resolve logical ambiguities.

Searching for blank fields

You can search for empty fields with the syntax: "".

Example: NOT 100ALL=""

This will retrieve all records that have an author's field.

Return to the top of the page.

Truncation, wildcards, and literals

Truncation

Truncation, symbolised by an asterisk, is most useful with Fields indexes, that require an exact match of a string of characters usually up to a maximum of 69.

The "600f" index for instance is built using the fields rule up to a maximum of 25 characters. An index key looks like this:

00 |A JOHN, |C KING OF EN

In order to find that key you must specify all of it and only all of it:

Example: 600F=00 |A JOHN, |C KING OF EN

Another index, the 600FF is built similarly but up to a maximum limit of 69 characters. An index key for the same item as above looks like:

600FF   00 |A JOHN, |C KING OF ENGLAND, |D 1167-1216 |V DRAMA.

If you want to limit your search to 600s that include the dates, but does not care about anything after that, you could use the syntax:

600FF=600FF   00 |A JOHN, |C KING OF ENGLAND, |D 1167-1216*

Truncation can also be used with words indexes:

Example: 100W=Volta*

Forward truncation

Forward truncation is possible, simply by placing the asterisk at the beginning of the search string:

100ALL=*ilton NEAR Paris

It can be however extremely slow, and is not recommended unless absolutely necessary.

Wildcards

Wildcards, represented by a question mark,  allow you to skip a character in a search term:

Example: TI=Merc?ri?s

This retrieves titles including the words: Mercurius, Mercvrius, Mercurivs, or Mercvrivs, as well as Merctrizs.

Literals

Certain words and characters have a syntactic meaning that sometimes clashes with the search string itself:

Example: TI=Tintin and Snow-White: An unfulfilled love story, or, "My life as a snow flake", by P**** Hilton.

In this case, the words AND and OR, and the quotation and asterisk characters would throw the search engine in utter confusion.

The best way to resolve word ambiguities is to place the search string in quotes:

Example: 245="Ladies AND gentlemen of the jury, I am NOT guilty"*

The truncation sign can then be placed at the end of the quotes.

To resolve character ambiguities you can also use the escape character: a single quote. For instance, if you want to search for a name that includes quotation marks:

Example:  100ALL=joe NEAR "'"killer shark'"" NEAR smith

A single quote can be used to escape a single quote:

Example: l''ancien testament

Return to the top of the page.

ESTC Pseudo Indexes

Return to the top of the page.

ESTC Words indexes

Return to the top of the page.

ESTC Fields Indexes

Return to the top of the page.