How the Search Syntax Works

The search works, in general, based on lemmatization; that is, a search for a word form finds all other word forms of the respective lemma:

 lex => finds all passages with "lex", "legum", "leges", etc. 

If a specific word form is to be found, an equality sign must be prepended to the search term:

 =lex => finds passages with "lex", but not those in which only "legum", "leges", etc. occur. 

(This operator can also be applied within phrases and other operators described further below.) Furthermore, the search allows the use of "?" and "*" as placeholders for one or infinitely many character(s):

 l?b?r => finds liber and labor lab* => finds labor and labello leg* => finds leges, legislator, legitimus, legatus, etc. 

How the Multiple-Term Search Works

The input of multiple search terms implies a search for text passages in which all the stated search terms must be present (implicit AND operator). If this is not desired, the terms must be separated by means of a "|" pipe (as OR operator):

 lex | ius => finds texts in which either lex or ius, or both, occur. 

At this, the precedence of the OR operator is higher than the precedence of the AND operator, such that

 Francisco Vitoria | Suárez => searches for Francisco and (Vitoria or Suárez), thus finds no document that only contains the string "Suárez". 

In case of doubt, the grouping of terms and operators may be explicitely stated in brackets:

 (Francisco Vitoria) | Suárez => searches for documents in which Francisco and Vitoria occur, or in which Suárez occurs. 

In case a term is to be excluded, it must be prepended by a "-" hyphen:

 lex -ius => finds all texts in which lex occurs but not ius. lex -(ius | naturalis) => finds all texts containing lex but neither ius nor naturalis. 

If an exact phrase is to be searched for, the respective terms must be embedded within quotation signs:

 "lex naturalis" => searches for lex naturalis, with both terms being required to occur in this very sequence. 

If the search terms are to occur close to each other, the phrase can either be relativized:

 "lex aeterna naturalis"~10 => searches for the occurrence of all three search terms within a scope of ten words. 

... or a tolerance for distance must be stated by means of the "NEAR/XY" operator between each search term:

 lex NEAR/10 aeterna NEAR/20 => searches for a passage of maximally 9 words between lex and aeterna and maximally 19 words between the found aeterna and naturalis. 

In case that merely some of a set of search terms shall be found, a phrase can be relativized through "/":

 lex "divina aeterna naturalis positiva civilis"/2 => searches for the term "lex" and 2 out of the 5 stated additional words. 

If the order but not the immediate succession of search terms is relevant, the order operator "<<" can be used:

 lex << ius => searches for lex ahead of ius 

(This operator has, along with the NEAR/XY operator, the lowest precedence of all operators.) The "SENTENCE" operator assigns two search terms to the condition of both thse terms having to occur within the same sentence.

 lex SENTENCE naturalis => searches for a sentence in which lex as well as naturalis occur. 

Search within Single Works

When searching globally or within a specific category of "data sources" ("Works: full text", "Works: headings", "Working Papers: ..."), all texts (of the selected category) are targeted through the search. A more selective search for single works (instead of the whole corpus) is not possible by means of the graphical user interface at the moment. However, the technically adventurous may limit their search queries already through the "@sphinx_work" field operator.

For this purpose, it is necessary to know the specific "work id" of the respective work. This five-place id consists of the initial "W" (for "work") followed by four digits and can be found in URLs of work-specific views (such as the reading view or the catalogue entry for a work) in the "wid=" parameter; alternatively, the citation proposal for a work, to be found in its catalogue entry, contains the work id in its URL: "https://id.{{ $domain }}/texts/[work id]".

Once the respective work's id has been identified, the search can be limited to this work through the field operator in the following way:

[query term] @sphinx_work ^[work id]

For example, searching for the term "lex" only in Diego de Castillo's "Tratado de Cuentas" (which has the work id "W0004") is possible by means of the following query (to be entered in the search field):

lex @sphinx_work ^W0004

Apart from that, the search syntax follows the guidelines stated above.

Where to Find more Information

The search is based, in the background, on the Sphinx search engine in version 2.1.3. The largest part of the documentation of search syntax displayed there also applies to the search function at hand. ^* * An important exception is the stating of field operators "@"…
An important exception is the stating of field operators "@", which are not available on our search function. The source text is always searched after (and in the search for everything, the lemma, author, and working paper search also the author name); with the search within works, the diplomatic as well as the constituted version are searched after.