Search Queries
Search Modes in Brief
Sutta references are matched first. Typing mn8, sn 56.11, iti92 will list those suttas.
Fulltext Match: it searches the content for keywords using the query expressions, non-accented letters matching accented ones. (I.e. it makes a query in the tantivy fulltext index and assigns scores to the results.)
Fulltext search matches words in full, not in part, e.g. 'bodhi' will not match 'bodhisatta', but words are stemmed and will match declensions, e.g. 'bodhiṁ / bodhiyā'.
Searching with a single Pāli word will return results which include the inflected forms, the results ranked by the frequency of the term.
When searching for an exact text, use two or three words wrapped in double quotes, e.g. "amatogadhā sabbe dhammā" (including the quote marks).
Prefixing the word with the '+' sign means a term must be included, the '-' signs means it must be excluded.
bhikkhu +kamma -vipaka means should include 'bhikkhu', must include 'kamma', must exclude 'vipaka'.
Title / Headword Match: it searches only the titles of suttas or the headwords of dictionary words. (SQL queries)
Fulltext Match Queries
Powered by the tantivy fulltext search engine. Read more in the QueryParser docs.
The words in a query term are related as OR by default. kamma vipāka searches for entries which SHOULD include kamma OR vipāka, but not MUST include.
Prefixing the word with the '+' sign means a term must be included, the '-' signs means it must be excluded.
bhikkhu +kamma -vipaka means should include 'bhikkhu', must include 'kamma', must exclude 'vipaka'.
The texts are indexed with Pāli, English, etc. grammar stemmers, so declension forms will also match in the appropriate language.
- dukkha will match dukkhaṁ / dukkhā / dukkhāni / dukkhena etc.,
- bhikkhu kamma vipaka will match bhikkhave kammānaṁ vipāko,
- monk receives robes will match monks receiving robes.
Latin terms are expanded to include diacritics, patipada will match paṭipadā.
A pharse query is expressed with quote marks: "paṭhamena jhānena".
The query can match parts of the document:
- title:sticks cessation - match 'sticks' in the title, 'cessation' in the content
- word:kamma +work - match 'kamma' in the headword, must include 'work' in the content
- uid:pj4 - the uid should include 'pj4'
- upekkhindriyaṁ -source:cst4 - match 'upekkhindriyaṁ' in the content, exclude all cst4 documents
- calmness +source:thanissaro - match 'calmness' in the content, only in documents by Bh. Thanissaro
- +"buddhas of the past" +source:bodhi - must include the phrase 'buddhas of the past', only in documents by Bh. Bodhi
Fulltext with fuzzy search (~ icon)
This option allows matching words which may differ from the query by N number of characters (i.e. the Levenshtein Distance).
Fuzzy search is not availble together with regex patterns.
Contains Match
If any part of the text contains the query exactly, it matches.
bodhi will match the ‘bodhi’ part in bodhisatta.
dukkha will match the ‘dukkha’ part in dukkhassa, but will not match dukkho.
dukkh will match dukkho, since it contains ‘dukkh’.
The query is executed as an SQL %query% expression.
RegEx Match
If part of the text satisfies the query as a regular expression pattern, it matches.
To improve the time of RegEx queries, select a Language and Source filter, so that not all database records are scanned for matches.
Limitations
Stemming and accent folding happens only in Fulltext Match.
Contains Match and RegEx Match queries are used in the exact form given.
This means they don't stem Pāli words and don't fold accents.
bodhiya will not match bodhi or bodhiyā.
Search As You Type is disabled for these modes.
Press Enter or click the search button to run the query.
AND Expressions
The search query is split on AND (must be uppercase). This combines each expression as a filter.
Examples:
dukkha AND sukha "Contains Match" means a text has to match:
%dukkha%first, and then also%sukha%
dukkha AND sukh\w+ "RegEx Match" means a text has to match:
dukkhafirst, and then alsosukh\w+