view libinterp/corefcn/strfind.cc @ 21200:fcac5dbbf9ed

maint: Indent #ifdef blocks in libinterp. * builtins.h, Cell.cc, __contourc__.cc, __dispatch__.cc, __dsearchn__.cc, __ichol__.cc, __ilu__.cc, __lin_interpn__.cc, __pchip_deriv__.cc, __qp__.cc, balance.cc, besselj.cc, betainc.cc, bitfcns.cc, bsxfun.cc, c-file-ptr-stream.cc, c-file-ptr-stream.h, cellfun.cc, colloc.cc, comment-list.cc, conv2.cc, daspk.cc, dasrt.cc, dassl.cc, data.cc, debug.cc, defaults.cc, defaults.in.h, defun-dld.h, defun.cc, defun.h, det.cc, dirfns.cc, display.cc, dlmread.cc, dot.cc, dynamic-ld.cc, eig.cc, ellipj.cc, error.cc, errwarn.cc, event-queue.cc, fft.cc, fft2.cc, fftn.cc, file-io.cc, filter.cc, find.cc, gammainc.cc, gcd.cc, getgrent.cc, getpwent.cc, getrusage.cc, givens.cc, gl-render.cc, gl2ps-print.cc, graphics.cc, graphics.in.h, gripes.cc, hash.cc, help.cc, hess.cc, hex2num.cc, input.cc, inv.cc, jit-ir.cc, jit-typeinfo.cc, jit-util.cc, jit-util.h, kron.cc, load-path.cc, load-save.cc, lookup.cc, ls-ascii-helper.cc, ls-hdf5.cc, ls-mat-ascii.cc, ls-mat4.cc, ls-mat5.cc, ls-oct-binary.cc, ls-oct-text.cc, ls-oct-text.h, ls-utils.cc, ls-utils.h, lsode.cc, lu.cc, luinc.cc, mappers.cc, matrix_type.cc, max.cc, mex.h, mexproto.h, mgorth.cc, nproc.cc, oct-errno.in.cc, oct-fstrm.cc, oct-hdf5-types.cc, oct-hdf5.h, oct-hist.cc, oct-iostrm.cc, oct-lvalue.cc, oct-map.cc, oct-prcstrm.cc, oct-procbuf.cc, oct-stream.cc, oct-strstrm.cc, octave-link.cc, ordschur.cc, pager.cc, pinv.cc, pr-output.cc, procstream.cc, profiler.cc, psi.cc, pt-jit.cc, quad.cc, quadcc.cc, qz.cc, rand.cc, rcond.cc, regexp.cc, schur.cc, sighandlers.cc, sparse-xdiv.cc, sparse-xpow.cc, sparse.cc, spparms.cc, sqrtm.cc, str2double.cc, strfind.cc, strfns.cc, sub2ind.cc, svd.cc, sylvester.cc, symtab.cc, syscalls.cc, sysdep.cc, sysdep.h, time.cc, toplev.cc, tril.cc, tsearch.cc, txt-eng-ft.cc, txt-eng.cc, typecast.cc, urlwrite.cc, utils.cc, variables.cc, xdiv.cc, xnorm.cc, xpow.cc, zfstream.cc, __delaunayn__.cc, __eigs__.cc, __fltk_uigetfile__.cc, __glpk__.cc, __init_fltk__.cc, __init_gnuplot__.cc, __magick_read__.cc, __osmesa_print__.cc, __voronoi__.cc, amd.cc, audiodevinfo.cc, audioread.cc, ccolamd.cc, chol.cc, colamd.cc, convhulln.cc, dmperm.cc, fftw.cc, oct-qhull.h, qr.cc, symbfact.cc, symrcm.cc, oct-conf.in.cc, ov-base-diag.cc, ov-base-int.cc, ov-base-mat.cc, ov-base-scalar.cc, ov-base-sparse.cc, ov-base.cc, ov-bool-mat.cc, ov-bool-sparse.cc, ov-bool.cc, ov-builtin.cc, ov-cell.cc, ov-ch-mat.cc, ov-class.cc, ov-classdef.cc, ov-colon.cc, ov-complex.cc, ov-cs-list.cc, ov-cx-diag.cc, ov-cx-mat.cc, ov-cx-sparse.cc, ov-dld-fcn.cc, ov-fcn-handle.cc, ov-fcn-inline.cc, ov-fcn.cc, ov-float.cc, ov-flt-complex.cc, ov-flt-cx-diag.cc, ov-flt-cx-mat.cc, ov-flt-re-diag.cc, ov-flt-re-mat.cc, ov-int16.cc, ov-int32.cc, ov-int64.cc, ov-int8.cc, ov-java.cc, ov-lazy-idx.cc, ov-mex-fcn.cc, ov-null-mat.cc, ov-oncleanup.cc, ov-perm.cc, ov-range.cc, ov-re-diag.cc, ov-re-mat.cc, ov-re-sparse.cc, ov-scalar.cc, ov-str-mat.cc, ov-struct.cc, ov-typeinfo.cc, ov-uint16.cc, ov-uint32.cc, ov-uint64.cc, ov-uint8.cc, ov-usr-fcn.cc, ov.cc, ovl.cc, octave.cc, op-b-b.cc, op-b-bm.cc, op-b-sbm.cc, op-bm-b.cc, op-bm-bm.cc, op-bm-sbm.cc, op-cdm-cdm.cc, op-cell.cc, op-chm.cc, op-class.cc, op-cm-cm.cc, op-cm-cs.cc, op-cm-m.cc, op-cm-s.cc, op-cm-scm.cc, op-cm-sm.cc, op-cs-cm.cc, op-cs-cs.cc, op-cs-m.cc, op-cs-s.cc, op-cs-scm.cc, op-cs-sm.cc, op-dm-dm.cc, op-dm-scm.cc, op-dm-sm.cc, op-dm-template.cc, op-dms-template.cc, op-double-conv.cc, op-fcdm-fcdm.cc, op-fcdm-fdm.cc, op-fcm-fcm.cc, op-fcm-fcs.cc, op-fcm-fm.cc, op-fcm-fs.cc, op-fcn.cc, op-fcs-fcm.cc, op-fcs-fcs.cc, op-fcs-fm.cc, op-fcs-fs.cc, op-fdm-fdm.cc, op-float-conv.cc, op-fm-fcm.cc, op-fm-fcs.cc, op-fm-fm.cc, op-fm-fs.cc, op-fs-fcm.cc, op-fs-fcs.cc, op-fs-fm.cc, op-fs-fs.cc, op-i16-i16.cc, op-i32-i32.cc, op-i64-i64.cc, op-i8-i8.cc, op-int-concat.cc, op-int-conv.cc, op-m-cm.cc, op-m-cs.cc, op-m-m.cc, op-m-s.cc, op-m-scm.cc, op-m-sm.cc, op-pm-pm.cc, op-pm-scm.cc, op-pm-sm.cc, op-pm-template.cc, op-range.cc, op-s-cm.cc, op-s-cs.cc, op-s-m.cc, op-s-s.cc, op-s-scm.cc, op-s-sm.cc, op-sbm-b.cc, op-sbm-bm.cc, op-sbm-sbm.cc, op-scm-cm.cc, op-scm-cs.cc, op-scm-m.cc, op-scm-s.cc, op-scm-scm.cc, op-scm-sm.cc, op-sm-cm.cc, op-sm-cs.cc, op-sm-m.cc, op-sm-s.cc, op-sm-scm.cc, op-sm-sm.cc, op-str-m.cc, op-str-s.cc, op-str-str.cc, op-struct.cc, op-ui16-ui16.cc, op-ui32-ui32.cc, op-ui64-ui64.cc, op-ui8-ui8.cc, pt-arg-list.cc, pt-array-list.cc, pt-assign.cc, pt-binop.cc, pt-bp.cc, pt-cbinop.cc, pt-cell.cc, pt-check.cc, pt-classdef.cc, pt-cmd.cc, pt-colon.cc, pt-colon.h, pt-const.cc, pt-decl.cc, pt-eval.cc, pt-except.cc, pt-exp.cc, pt-fcn-handle.cc, pt-funcall.cc, pt-id.cc, pt-idx.cc, pt-jump.cc, pt-loop.cc, pt-mat.cc, pt-misc.cc, pt-pr-code.cc, pt-select.cc, pt-stmt.cc, pt-unop.cc, pt.cc, token.cc, Array-jit.cc, Array-os.cc, Array-sym.cc, Array-tc.cc, version.cc: Indent #ifdef blocks in libinterp.
author Rik <rik@octave.org>
date Fri, 05 Feb 2016 16:29:08 -0800
parents e39e05d90788
children 40de9f8f23a6
line wrap: on
line source

/*

Copyright (C) 2009-2015 Jaroslav Hajek
Copyright (C) 2009-2010 VZLU Prague

This file is part of Octave.

Octave is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 3 of the License, or (at your
option) any later version.

Octave is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
for more details.

You should have received a copy of the GNU General Public License
along with Octave; see the file COPYING.  If not, see
<http://www.gnu.org/licenses/>.

*/

#ifdef HAVE_CONFIG_H
#  include <config.h>
#endif

#include <algorithm>
#include <deque>
#include <limits>
#include <string>

#include "oct-locbuf.h"

#include "Cell.h"
#include "ov.h"
#include "defun.h"
#include "unwind-prot.h"
#include "errwarn.h"
#include "utils.h"

// This allows safe indexing with char.
// In C++, char may be (and often is) signed!
#define ORD(ch) static_cast<unsigned char>(ch)
#define TABSIZE (std::numeric_limits<unsigned char>::max () + 1)

// This is the quick search algorithm, as described at
// http://www-igm.univ-mlv.fr/~lecroq/string/node19.html
static void
qs_preprocess (const Array<char>& needle,
               octave_idx_type *table)
{
  const char *x = needle.data ();
  octave_idx_type m = needle.numel ();

  for (octave_idx_type i = 0; i < TABSIZE; i++)
    table[i] = m + 1;
  for (octave_idx_type i = 0; i < m; i++)
    table[ORD(x[i])] = m - i;
}


static Array<octave_idx_type>
qs_search (const Array<char>& needle,
           const Array<char>& haystack,
           const octave_idx_type *table,
           bool overlaps = true)
{
  const char *x = needle.data ();
  octave_idx_type m = needle.numel ();
  const char *y = haystack.data ();
  octave_idx_type n = haystack.numel ();

  // We'll use deque because it typically has the most favorable properties for
  // the operation we need.
  std::deque<octave_idx_type> accum;
  if (m == 1)
    {
      // Looking for a single character.
      for (octave_idx_type i = 0; i < n; i++)
        {
          if (y[i] == x[0])
            accum.push_back (i);
        }
    }
  else if (m == 2)
    {
      // Two characters.
      if (overlaps)
        {
          for (octave_idx_type i = 0; i < n-1; i++)
            {
              if (y[i] == x[0] && y[i+1] == x[1])
                accum.push_back (i);
            }
        }
      else
        {
          for (octave_idx_type i = 0; i < n-1; i++)
            {
              if (y[i] == x[0] && y[i+1] == x[1])
                accum.push_back (i++);
            }
        }
    }
  else if (n >= m)
    {
      // General case.
      octave_idx_type j = 0;

      if (overlaps)
        {
          while (j < n - m)
            {
              if (std::equal (x, x + m, y + j))
                accum.push_back (j);
              j += table[ORD(y[j + m])];
            }
        }
      else
        {
          while (j < n - m)
            {
              if (std::equal (x, x + m, y + j))
                {
                  accum.push_back (j);
                  j += m;
                }
              else
                j += table[ORD(y[j + m])];
            }
        }

      if (j == n - m && std::equal (x, x + m, y + j))
        accum.push_back (j);
    }

  octave_idx_type nmatch = accum.size ();
  octave_idx_type one = 1;
  Array<octave_idx_type> result (dim_vector (std::min (one, nmatch), nmatch));
  octave_idx_type k = 0;
  for (std::deque<octave_idx_type>::const_iterator iter = accum.begin ();
       iter != accum.end (); iter++)
    {
      result.xelem (k++) = *iter;
    }

  return result;
}

DEFUN (strfind, args, ,
       "-*- texinfo -*-\n\
@deftypefn  {} {@var{idx} =} strfind (@var{str}, @var{pattern})\n\
@deftypefnx {} {@var{idx} =} strfind (@var{cellstr}, @var{pattern})\n\
@deftypefnx {} {@var{idx} =} strfind (@dots{}, \"overlaps\", @var{val})\n\
Search for @var{pattern} in the string @var{str} and return the starting\n\
index of every such occurrence in the vector @var{idx}.\n\
\n\
If there is no such occurrence, or if @var{pattern} is longer than\n\
@var{str}, or if @var{pattern} itself is empty, then @var{idx} is the empty\n\
array @code{[]}.\n\
\n\
The optional argument @qcode{\"overlaps\"} determines whether the pattern\n\
can match at every position in @var{str} (true), or only for unique\n\
occurrences of the complete pattern (false).  The default is true.\n\
\n\
If a cell array of strings @var{cellstr} is specified then @var{idx} is a\n\
cell array of vectors, as specified above.\n\
\n\
Examples:\n\
\n\
@example\n\
@group\n\
strfind (\"abababa\", \"aba\")\n\
     @result{} [1, 3, 5]\n\
\n\
strfind (\"abababa\", \"aba\", \"overlaps\", false)\n\
     @result{} [1, 5]\n\
\n\
strfind (@{\"abababa\", \"bebebe\", \"ab\"@}, \"aba\")\n\
     @result{}\n\
        @{\n\
          [1,1] =\n\
\n\
             1   3   5\n\
\n\
          [1,2] = [](1x0)\n\
          [1,3] = [](1x0)\n\
        @}\n\
@end group\n\
@end example\n\
@seealso{findstr, strmatch, regexp, regexpi, find}\n\
@end deftypefn")
{
  int nargin = args.length ();

  if (nargin != 4 && nargin != 2)
    print_usage ();

  bool overlaps = true;
  if (nargin == 4)
    {
      if (! args(2).is_string () || ! args(3).is_scalar_type ())
        error ("strfind: invalid optional arguments");

      std::string opt = args(2).string_value ();

      if (opt == "overlaps")
        overlaps = args(3).bool_value ();
      else
        error ("strfind: unknown option: %s", opt.c_str ());
    }

  octave_value retval;

  octave_value argstr = args(0);
  octave_value argpat = args(1);

  if (argpat.is_string ())
    {
      Array<char> needle = argpat.char_array_value ();
      OCTAVE_LOCAL_BUFFER (octave_idx_type, table, TABSIZE);
      qs_preprocess (needle, table);

      if (argstr.is_string ())
        if (argpat.is_empty ())
          // Return a null matrix for null pattern for MW compatibility
          retval = Matrix ();
        else
          retval = octave_value (qs_search (needle,
                                            argstr.char_array_value (),
                                            table, overlaps),
                                 true, true);
      else if (argstr.is_cell ())
        {
          const Cell argsc = argstr.cell_value ();
          Cell retc (argsc.dims ());
          octave_idx_type ns = argsc.numel ();

          for (octave_idx_type i = 0; i < ns; i++)
            {
              octave_value argse = argsc(i);
              if (! argse.is_string ())
                error ("strfind: each element of CELLSTR must be a string");

              if (argpat.is_empty ())
                retc(i) = Matrix ();
              else
                retc(i) = octave_value (qs_search (needle,
                                                   argse.char_array_value (),
                                                   table, overlaps),
                                        true, true);
            }

          retval = retc;
        }
      else
        error ("strfind: first argument must be a string or cell array of strings");
    }
  else if (argpat.is_cell ())
    retval = do_simple_cellfun (Fstrfind, "strfind", args);
  else
    error ("strfind: PATTERN must be a string or cell array of strings");

  return retval;
}

/*
%!assert (strfind ("abababa", "aba"), [1, 3, 5])
%!assert (strfind ("abababa", "aba", "overlaps", false), [1, 5])
%!assert (strfind ({"abababa", "bla", "bla"}, "a"), {[1, 3, 5, 7], 3, 3})
%!assert (strfind ("Linux _is_ user-friendly. It just isn't ignorant-friendly or idiot-friendly.", "friendly"), [17, 50, 68])
%!assert (strfind ("abc", ""), [])
%!assert (strfind ("abc", {"", "b", ""}), {[], 2, []})
%!assert (strfind ({"abc", "def"}, ""), {[], []})

%!error strfind ()
%!error strfind ("foo", "bar", 1)
%!error <unknown option: foobar> strfind ("foo", 100, "foobar", 1)
%!error <each element of CELLSTR must be a string> strfind ({"A", 1}, "foo")
%!error <first argument must be a string> strfind (100, "foo")
%!error <PATTERN must be a string> strfind ("foo", 100)
*/

static Array<char>
qs_replace (const Array<char>& str, const Array<char>& pat,
            const Array<char>& rep,
            const octave_idx_type *table,
            bool overlaps = true)
{
  Array<char> ret = str;

  octave_idx_type siz = str.numel ();
  octave_idx_type psiz = pat.numel ();
  octave_idx_type rsiz = rep.numel ();

  if (psiz != 0)
    {
      // Look up matches, without overlaps.
      const Array<octave_idx_type> idx = qs_search (pat, str, table, overlaps);
      octave_idx_type nidx = idx.numel ();

      if (nidx)
        {
          // Compute result size.
          octave_idx_type retsiz;
          if (overlaps)
            {
              retsiz = 0;
              // OMG. Is this the "right answer" MW always looks for, or
              // someone was just lazy?
              octave_idx_type k = 0;
              for (octave_idx_type i = 0; i < nidx; i++)
                {
                  octave_idx_type j = idx(i);
                  if (j >= k)
                    retsiz += j - k;
                  retsiz += rsiz;
                  k = j + psiz;
                }

              retsiz += siz - k;
            }
          else
            retsiz = siz + nidx * (rsiz - psiz);

          if (retsiz == 0)
            ret.clear (dim_vector (0, 0));
          else
            {
              ret.clear (dim_vector (1, retsiz));
              const char *src = str.data ();
              const char *reps = rep.data ();
              char *dest = ret.fortran_vec ();

              octave_idx_type k = 0;
              for (octave_idx_type i = 0; i < nidx; i++)
                {
                  octave_idx_type j = idx(i);
                  if (j >= k)
                    dest = std::copy (src + k, src + j, dest);
                  dest = std::copy (reps, reps + rsiz, dest);
                  k = j + psiz;
                }

              std::copy (src + k, src + siz, dest);
            }
        }
    }

  return ret;
}

DEFUN (strrep, args, ,
       "-*- texinfo -*-\n\
@deftypefn  {} {@var{newstr} =} strrep (@var{str}, @var{ptn}, @var{rep})\n\
@deftypefnx {} {@var{newstr} =} strrep (@var{cellstr}, @var{ptn}, @var{rep})\n\
@deftypefnx {} {@var{newstr} =} strrep (@dots{}, \"overlaps\", @var{val})\n\
Replace all occurrences of the pattern @var{ptn} in the string @var{str}\n\
with the string @var{rep} and return the result.\n\
\n\
The optional argument @qcode{\"overlaps\"} determines whether the pattern\n\
can match at every position in @var{str} (true), or only for unique\n\
occurrences of the complete pattern (false).  The default is true.\n\
\n\
@var{s} may also be a cell array of strings, in which case the replacement is\n\
done for each element and a cell array is returned.\n\
\n\
Example:\n\
\n\
@example\n\
@group\n\
strrep (\"This is a test string\", \"is\", \"&%$\")\n\
    @result{}  \"Th&%$ &%$ a test string\"\n\
@end group\n\
@end example\n\
\n\
@seealso{regexprep, strfind, findstr}\n\
@end deftypefn")
{
  int nargin = args.length ();

  if (nargin != 3 && nargin != 5)
    print_usage ();

  bool overlaps = true;

  if (nargin == 5)
    {
      if (! args(3).is_string () || ! args(4).is_scalar_type ())
        error ("strrep: invalid optional arguments");

      std::string opt = args(3).string_value ();
      if (opt == "overlaps")
        overlaps = args(4).bool_value ();
      else
        error ("strrep: unknown option: %s", opt.c_str ());
    }

  octave_value retval;

  octave_value argstr = args(0);
  octave_value argpat = args(1);
  octave_value argrep = args(2);

  if (argpat.is_string () && argrep.is_string ())
    {
      const Array<char> pat = argpat.char_array_value ();
      const Array<char> rep = argrep.char_array_value ();

      OCTAVE_LOCAL_BUFFER (octave_idx_type, table, TABSIZE);
      qs_preprocess (pat, table);

      if (argstr.is_string ())
        retval = qs_replace (argstr.char_array_value (), pat, rep,
                             table, overlaps);
      else if (argstr.is_cell ())
        {
          const Cell argsc = argstr.cell_value ();
          Cell retc (argsc.dims ());
          octave_idx_type ns = argsc.numel ();

          for (octave_idx_type i = 0; i < ns; i++)
            {
              octave_value argse = argsc(i);
              if (argse.is_string ())
                retc(i) = qs_replace (argse.char_array_value (), pat, rep,
                                      table, overlaps);
              else
                error ("strrep: each element of S must be a string");
            }

          retval = retc;
        }
      else
        error ("strrep: S must be a string or cell array of strings");
    }
  else if (argpat.is_cell () || argrep.is_cell ())
    retval = do_simple_cellfun (Fstrrep, "strrep", args);
  else
    error ("strrep: PTN and REP arguments must be strings or cell arrays of strings");

  return retval;
}

/*
%!assert (strrep ("This is a test string", "is", "&%$"),
%!                "Th&%$ &%$ a test string")
%!assert (strrep ("abababc", "abab", "xyz"), "xyzxyzc")
%!assert (strrep ("abababc", "abab", "xyz", "overlaps", false), "xyzabc")

%!assert (size (strrep ("a", "a", "")), [0 0])

%!error strrep ()
%!error strrep ("foo", "bar", 3, 4)
*/