While the Wordlist is useful for tracking occurences of individual words, the Search offers extended ways of searching through the manuscripts.
All searches are performed page-by-page, and results are returned as page links.
The search operates as follows:
- wildcards: use ‘?’ to match any single character, and ‘*’ to match any number of characters – e.g. gwnaeth?nt
- phrases: put a phrase in double quotes to search for phrase as a whole – e.g. "sef a oruc"
- fuzzy search: to do a fuzzy search, use the tilde at the end of a single-word term – e.g. tywyssawc~
- proximity search: to do a proximity search, use the tilde at the end of the phrase – e.g. to find gwynt and heul within 50 words of each other, use "gwynt heul"~50
- required / prohibited: the ‘+’ symbol makes a term required so that only pages containing that page are returned; the ‘-’ symbol makes a term prohibited, so that pages with that term are excluded from the results – e.g. eneit +corff
- boolean operators: operators ‘and’, ‘or’, ‘not’ can be used to join terms – e.g. uffern or nef will return pages with either uffern or nef mentioned; uffern not nef will return pages including uffern but not nef
- grouping: for even greater control, tems can be grouped in parentheses – e.g. (uffern or nef) and eneit
Note that as searching is done page-by-page, phrases which span pages will not be found; similarly for proximity searches.
Fuzzy searches can return overly ‘broad’ results; to narrow the ‘fuzzyness’, add a similarity factor after the tilde – e.g. tywyssawc~0.8 (the default similarity factor is 0.5).
- C & K are considered equivalent, and are represented as C.
- the digraphs Ch, Dd, Ff, Ng, Ll, Ph, Rh, and Th are treated as character pairs (unlike the Wordlist).
- the ‘middle welsh v’ character ỽ and ‘middle welsh ll’ character ỻ appear as w and ll respectively;
- the character represented by ð in Peniarth 20 appears as dd.