Mercurial > hg > octave-terminal
changeset 13310:583940a28bfd
handle "split" option for regexp
* regexp.cc (octregexp_list, octregexp, Fregexp):
Handle "split" option.
(octregexp): Consolidate multiple loops for gathering outputs.
(Fregexp): New tests.
author | John W. Eaton <jwe@octave.org> |
---|---|
date | Mon, 10 Oct 2011 15:01:56 -0400 |
parents | 1f075333d9e8 |
children | d590d9df5596 |
files | src/DLD-FUNCTIONS/regexp.cc |
diffstat | 1 files changed, 98 insertions(+), 65 deletions(-) [+] |
line wrap: on
line diff
--- a/src/DLD-FUNCTIONS/regexp.cc +++ b/src/DLD-FUNCTIONS/regexp.cc @@ -168,9 +168,10 @@ freespacing = true; nopts--; } - else if (str.find ("start", 0) && str.find ("end", 0) && - str.find ("tokenextents", 0) && str.find ("match", 0) && - str.find ("tokens", 0) && str.find ("names", 0)) + else if (str.find ("start", 0) && str.find ("end", 0) + && str.find ("tokenextents", 0) && str.find ("match", 0) + && str.find ("tokens", 0) && str.find ("names", 0) + && str.find ("split", 0)) error ("%s: unrecognized option", nm.c_str ()); } @@ -524,6 +525,8 @@ octave_idx_type i = 0; octave_scalar_map nmap; + retval.resize (7); + if (sz == 1) { for (int j = 0; j < named.length (); j++) @@ -547,78 +550,64 @@ retval(5) = nmap; } + std::string buffer = args(0).string_value (); + if (once) - retval(4) = sz ? lst.front ().t : Cell (); + { + retval(4) = sz ? lst.front ().t : Cell (); + retval(3) = sz ? lst.front ().m : std::string (); + retval(2) = sz ? lst.front ().te : Matrix (); + + if (sz) + { + double e = lst.front ().e; + double s = lst.front ().s; + + Cell sp (dim_vector (1, 2)); + sp(0) = buffer.substr (0, s-1); + sp(1) = buffer.substr (e); + + retval(6) = sp; + retval(1) = e; + retval(0) = s; + } + else + { + retval(6) = buffer; + retval(1) = Matrix (); + retval(0) = Matrix (); + } + } else { Cell t (dim_vector (1, sz)); - - i = 0; - for (const_iterator p = lst.begin (); p != lst.end (); p++) - t(i++) = p->t; - - retval(4) = t; - } - - if (once) - retval(3) = sz ? lst.front ().m : std::string (); - else - { Cell m (dim_vector (1, sz)); - - i = 0; - for (const_iterator p = lst.begin (); p != lst.end (); p++) - m(i++) = p->m; - - retval(3) = m; - } - - if (once) - retval(2) = sz ? lst.front ().te : Matrix (); - else - { Cell te (dim_vector (1, sz)); + NDArray e (dim_vector (1, sz)); + NDArray s (dim_vector (1, sz)); + Cell sp (dim_vector (1, sz+1)); + size_t sp_start = 0; i = 0; for (const_iterator p = lst.begin (); p != lst.end (); p++) - te(i++) = p->te; - - retval(2) = te; - } - - if (once) - { - if (sz) - retval(1) = lst.front ().e; - else - retval(1) = Matrix (); - } - else - { - NDArray e (dim_vector (1, sz)); + { + t(i) = p->t; + m(i) = p->m; + te(i) = p->te; + e(i) = p->e; + s(i) = p->s; + sp(i) = buffer.substr (sp_start, p->s-sp_start-1); + sp_start = p->e; + i++; + } - i = 0; - for (const_iterator p = lst.begin (); p != lst.end (); p++) - e(i++) = p->e; - - retval(1) = e; - } + sp(i) = buffer.substr (sp_start); - if (once) - { - if (sz) - retval(0) = lst.front ().s; - else - retval(0) = Matrix (); - } - else - { - NDArray s (dim_vector (1, sz)); - - i = 0; - for (const_iterator p = lst.begin (); p != lst.end (); p++) - s(i++) = p->s; - + retval(6) = sp; + retval(4) = t; + retval(3) = m; + retval(2) = te; + retval(1) = e; retval(0) = s; } @@ -661,6 +650,8 @@ k = 4; else if (str.find ("names", 0) == 0) k = 5; + else if (str.find ("split", 0) == 0) + k = 6; new_retval(n++) = retval(k); arg_used[k] = true; @@ -947,6 +938,8 @@ A structure containing the text of each matched named token, with the name\n\ being used as the fieldname. A named token is denoted by\n\ @code{(?<name>@dots{})}.\n\ +@item sp\n\ +A cell array of the text not returned by match.\n\ @end table\n\ \n\ Particular output arguments, or the order of the output arguments, can be\n\ @@ -961,6 +954,7 @@ @item @tab 'match' @tab @var{m} @tab\n\ @item @tab 'tokens' @tab @var{t} @tab\n\ @item @tab 'names' @tab @var{nm} @tab\n\ +@item @tab 'split' @tab @var{sp} @tab\n\ @end multitable\n\ \n\ Additional arguments are summarized below.\n\ @@ -1200,6 +1194,45 @@ %! assert(regexp("qit",'q(?=u*)','match'), {'q'}) %! assert(regexp('thingamabob','(?<=a)b'), 9) +## Tests for split option. +%!shared str +%! str = "foo bar foo"; +%!test +%! [a, b] = regexp (str, "f..", "match", "split"); +%! assert (a, {"foo", "foo"}); +%! assert (b, {"", " bar ", ""}); +%!test +%! [a, b] = regexp (str, "f..", "match", "split", "once"); +%! assert (a, "foo"); +%! assert (b, {"", " bar foo"}); +%!test +%! [a, b] = regexp (str, "fx.", "match", "split"); +%! assert (a, cell (1, 0)); +%! assert (b, {"foo bar foo"}); +%!test +%! [a, b] = regexp (str, "fx.", "match", "split", "once"); +%! assert (a, ""); +%! assert (b, "foo bar foo") + +%!shared str +%! str = "foo bar"; +%!test +%! [a, b] = regexp (str, "f..", "match", "split"); +%! assert (a, {"foo"}); +%! assert (b, {"", " bar"}); +%!test +%! [a, b] = regexp (str, "b..", "match", "split"); +%! assert (a, {"bar"}); +%! assert (b, {"foo ", ""}); +%!test +%! [a, b] = regexp (str, "x", "match", "split"); +%! assert (a, cell (1, 0)); +%! assert (b, {"foo bar"}); +%!test +%! [a, b] = regexp (str, "[o]+", "match", "split"); +%! assert (a, {"oo"}); +%! assert (b, {"f", " bar"}); + */ DEFUN_DLD (regexpi, args, nargout, @@ -1395,7 +1428,7 @@ const std::string opt = args(i).string_value (); if (opt != "tokenize" && opt != "start" && opt != "end" && opt != "tokenextents" && opt != "match" && opt != "tokens" - && opt != "names" && opt != "warnings") + && opt != "names" && opt != "split" && opt != "warnings") { regexpargs(len++) = args(i); }