Revision as of 22:18, 20 July 2022 (view source) Hout (talk \| contribs) m (→‎{{header\|Haskell}}: Tidied) ← Older edit		Revision as of 01:24, 23 July 2022 (view source) Rdm (talk \| contribs) (J: simpler) Newer edit →
Line 819: =={{header\|J}}== <lang j>multisplit=: 4{{ NB. x:0 text, y: separators '~~sep~~ begin sep'=. \|: tbs=. y _,~/:~;(,.&.~~:(\|."1)@;@(~~>i.@#@[) ~~,.L:0"0~~y I.@E.L:0) x ~~end~~r=. ~~begin~~i.2 +0 NB. ~~sep~~partially {formed ~~#@>y~~result j=. k=. 0 NB. j indexes text, k indexes of (potential) instances of separators in text ~~last=. next=. 0~~ r=while.j<#x ~~2 0$0~~do. while. ~~next<#~~j>k{begin do. k=.k+1 end. NB. ignore overlapped separators 'b s'=. k{bs NB. unpack details about next separator ~~r=. r,.(last}.x{.~next{begin);next{t~~ if. _=b do. r,.(j}.x);'' return. end. NB. finish if no more separators ~~last=. next{end~~ txt=. (j + i. b-j){x NB. extract text between separators ~~next=. 1 i.~(begin>next{begin)*.begin>:last~~ j=. b+s{#@>y NB. advance to position after seprator r=.r,.txt;s,b NB. update partial result end. )}}</lang>▼ ~~r=. r,.'';~last}.x~~ ▲)</lang> Explanation: First find all potentially relevant separator instances, and sort them in increasing order, by starting location and separator index. <code>sep</code> is separator index, and <code>begin</code> is starting ~~location. <code>end</code> is ending~~ location. Then, loop through the possibilities, skipping over those separators which would overlap with previously used separators.

Multisplit: Difference between revisions

Multisplit (view source)

Revision as of 01:24, 23 July 2022