Mercurial > octave-nkf
annotate src/ls-mat-ascii.cc @ 8946:e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
| author | Benjamin Lindner <lindnerb@users.sourceforge.net> |
|---|---|
| date | Tue, 10 Mar 2009 01:01:50 -0400 |
| parents | eb63fbe60fab |
| children | 985792c9e0da |
| rev | line source |
|---|---|
| 4634 | 1 /* |
| 2 | |
| 8920 | 3 Copyright (C) 1996, 1997, 2003, 2004, 2005, 2006, 2007, 2008 John W. Eaton |
| 4634 | 4 |
| 5 This file is part of Octave. | |
| 6 | |
| 7 Octave is free software; you can redistribute it and/or modify it | |
| 8 under the terms of the GNU General Public License as published by the | |
| 7016 | 9 Free Software Foundation; either version 3 of the License, or (at your |
| 10 option) any later version. | |
| 4634 | 11 |
| 12 Octave is distributed in the hope that it will be useful, but WITHOUT | |
| 13 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
| 14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
| 15 for more details. | |
| 16 | |
| 17 You should have received a copy of the GNU General Public License | |
| 7016 | 18 along with Octave; see the file COPYING. If not, see |
| 19 <http://www.gnu.org/licenses/>. | |
| 4634 | 20 |
| 21 */ | |
| 22 | |
| 23 #ifdef HAVE_CONFIG_H | |
| 24 #include <config.h> | |
| 25 #endif | |
| 26 | |
| 27 #include <cfloat> | |
| 28 #include <cstring> | |
| 29 #include <cctype> | |
| 30 | |
| 31 #include <fstream> | |
| 32 #include <iomanip> | |
| 33 #include <iostream> | |
| 5765 | 34 #include <sstream> |
| 4634 | 35 #include <string> |
| 36 | |
| 37 #include "byte-swap.h" | |
| 38 #include "data-conv.h" | |
| 39 #include "file-ops.h" | |
| 40 #include "glob-match.h" | |
| 41 #include "lo-mappers.h" | |
| 42 #include "mach-info.h" | |
| 43 #include "oct-env.h" | |
| 44 #include "oct-time.h" | |
| 45 #include "quit.h" | |
| 46 #include "str-vec.h" | |
| 47 | |
| 48 #include "Cell.h" | |
| 49 #include "defun.h" | |
| 50 #include "error.h" | |
| 51 #include "gripes.h" | |
| 4867 | 52 #include "lex.h" |
| 4634 | 53 #include "load-save.h" |
|
8946
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
54 #include "ls-ascii-helper.h" |
|
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
55 #include "ls-mat-ascii.h" |
| 4634 | 56 #include "oct-obj.h" |
| 57 #include "oct-map.h" | |
| 58 #include "ov-cell.h" | |
| 59 #include "pager.h" | |
| 60 #include "pt-exp.h" | |
| 61 #include "sysdep.h" | |
| 62 #include "unwind-prot.h" | |
| 63 #include "utils.h" | |
| 64 #include "variables.h" | |
| 65 #include "version.h" | |
| 66 #include "dMatrix.h" | |
| 67 | |
| 68 static std::string | |
| 69 get_mat_data_input_line (std::istream& is) | |
| 70 { | |
| 71 std::string retval; | |
| 72 | |
| 73 bool have_data = false; | |
| 74 | |
| 75 do | |
| 76 { | |
| 77 retval = ""; | |
| 78 | |
| 79 char c; | |
| 80 while (is.get (c)) | |
| 81 { | |
| 82 if (c == '\n' || c == '\r') | |
|
8946
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
83 { |
|
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
84 skip_until_newline (is, false); |
|
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
85 break; |
|
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
86 } |
| 4634 | 87 |
| 88 if (c == '%' || c == '#') | |
| 89 { | |
|
8946
e7e928088e90
fix CRLF issues with text-mode reading in windows when loading ascii data
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
8920
diff
changeset
|
90 skip_until_newline (is, false); |
| 4634 | 91 break; |
| 92 } | |
| 93 | |
| 94 if (! is.eof ()) | |
| 95 { | |
| 96 if (! have_data && c != ' ' && c != '\t') | |
| 97 have_data = true; | |
| 98 | |
| 99 retval += c; | |
| 100 } | |
| 101 } | |
| 102 } | |
| 103 while (! (have_data || is.eof ())); | |
| 104 | |
| 105 return retval; | |
| 106 } | |
| 107 | |
| 108 static void | |
| 5275 | 109 get_lines_and_columns (std::istream& is, const std::string& filename, octave_idx_type& nr, octave_idx_type& nc) |
| 4634 | 110 { |
| 111 std::streampos pos = is.tellg (); | |
| 112 | |
| 113 int file_line_number = 0; | |
| 114 | |
| 115 nr = 0; | |
| 116 nc = 0; | |
| 117 | |
| 118 while (is && ! error_state) | |
| 119 { | |
| 120 OCTAVE_QUIT; | |
| 121 | |
| 122 std::string buf = get_mat_data_input_line (is); | |
| 123 | |
| 124 file_line_number++; | |
| 125 | |
| 126 size_t beg = buf.find_first_not_of (", \t"); | |
| 127 | |
| 128 // If we see a CR as the last character in the buffer, we had a | |
| 129 // CRLF pair as the line separator. Any other CR in the text | |
| 130 // will not be considered as whitespace. | |
| 131 | |
| 8021 | 132 if (beg != std::string::npos && buf[beg] == '\r' && beg == buf.length () - 1) |
| 4634 | 133 { |
| 134 // We had a blank line ending with a CRLF. Handle it the | |
| 135 // same as an empty line. | |
| 8021 | 136 beg = std::string::npos; |
| 4634 | 137 } |
| 138 | |
| 5275 | 139 octave_idx_type tmp_nc = 0; |
| 4634 | 140 |
| 8021 | 141 while (beg != std::string::npos) |
| 4634 | 142 { |
| 143 tmp_nc++; | |
| 144 | |
| 145 size_t end = buf.find_first_of (", \t", beg); | |
| 146 | |
| 8021 | 147 if (end != std::string::npos) |
| 4634 | 148 { |
| 149 beg = buf.find_first_not_of (", \t", end); | |
| 150 | |
| 8021 | 151 if (beg == std::string::npos || (buf[beg] == '\r' && |
| 6856 | 152 beg == buf.length () - 1)) |
| 4634 | 153 { |
| 154 // We had a line with trailing spaces and | |
| 155 // ending with a CRLF, so this should look like EOL, | |
| 156 // not a new colum. | |
| 157 break; | |
| 158 } | |
| 159 } | |
| 160 else | |
| 161 break; | |
| 162 } | |
| 163 | |
| 164 if (tmp_nc > 0) | |
| 165 { | |
| 166 if (nc == 0) | |
| 167 { | |
| 168 nc = tmp_nc; | |
| 169 nr++; | |
| 170 } | |
| 171 else if (nc == tmp_nc) | |
| 172 nr++; | |
| 173 else | |
| 174 error ("load: %s: inconsistent number of columns near line %d", | |
| 175 filename.c_str (), file_line_number); | |
| 176 } | |
| 177 } | |
| 178 | |
| 179 if (nr == 0 || nc == 0) | |
| 180 error ("load: file `%s' seems to be empty!", filename.c_str ()); | |
| 181 | |
| 182 is.clear (); | |
| 4643 | 183 is.seekg (pos); |
| 4634 | 184 } |
| 185 | |
| 186 // Extract a matrix from a file of numbers only. | |
| 187 // | |
| 188 // Comments are not allowed. The file should only have numeric values. | |
| 189 // | |
| 190 // Reads the file twice. Once to find the number of rows and columns, | |
| 191 // and once to extract the matrix. | |
| 192 // | |
| 193 // FILENAME is used for error messages. | |
| 194 // | |
| 195 // This format provides no way to tag the data as global. | |
| 196 | |
| 197 std::string | |
| 198 read_mat_ascii_data (std::istream& is, const std::string& filename, | |
| 199 octave_value& tc) | |
| 200 { | |
| 201 std::string retval; | |
| 202 | |
| 203 std::string varname; | |
| 204 | |
| 205 size_t pos = filename.rfind ('/'); | |
| 206 | |
| 8021 | 207 if (pos != std::string::npos) |
| 4634 | 208 varname = filename.substr (pos+1); |
| 209 else | |
| 210 varname = filename; | |
| 211 | |
| 4867 | 212 pos = varname.rfind ('.'); |
| 4634 | 213 |
| 8021 | 214 if (pos != std::string::npos) |
| 4634 | 215 varname = varname.substr (0, pos); |
| 216 | |
| 217 size_t len = varname.length (); | |
| 218 for (size_t i = 0; i < len; i++) | |
| 219 { | |
| 220 char c = varname[i]; | |
| 221 if (! (isalnum (c) || c == '_')) | |
| 222 varname[i] = '_'; | |
| 223 } | |
| 224 | |
| 4867 | 225 if (is_keyword (varname) || ! isalpha (varname[0])) |
| 4634 | 226 varname.insert (0, "X"); |
| 227 | |
| 228 if (valid_identifier (varname)) | |
| 229 { | |
| 5275 | 230 octave_idx_type nr = 0; |
| 231 octave_idx_type nc = 0; | |
| 4634 | 232 |
| 233 int total_count = 0; | |
| 234 | |
| 235 get_lines_and_columns (is, filename, nr, nc); | |
| 236 | |
| 237 OCTAVE_QUIT; | |
| 238 | |
| 239 if (! error_state && nr > 0 && nc > 0) | |
| 240 { | |
| 241 Matrix tmp (nr, nc); | |
| 242 | |
| 243 if (nr < 1 || nc < 1) | |
| 244 is.clear (std::ios::badbit); | |
| 245 else | |
| 246 { | |
| 247 double d; | |
| 5275 | 248 for (octave_idx_type i = 0; i < nr; i++) |
| 4634 | 249 { |
| 250 std::string buf = get_mat_data_input_line (is); | |
| 251 | |
| 252 std::istringstream tmp_stream (buf); | |
| 253 | |
| 5275 | 254 for (octave_idx_type j = 0; j < nc; j++) |
| 4634 | 255 { |
| 256 OCTAVE_QUIT; | |
| 257 | |
| 258 d = octave_read_double (tmp_stream); | |
| 259 | |
| 260 if (tmp_stream || tmp_stream.eof ()) | |
| 261 { | |
| 262 tmp.elem (i, j) = d; | |
| 263 total_count++; | |
| 264 | |
| 265 // Skip whitespace and commas. | |
| 266 char c; | |
| 267 while (1) | |
| 268 { | |
| 269 tmp_stream >> c; | |
| 270 | |
| 271 if (! tmp_stream) | |
| 272 break; | |
| 273 | |
| 274 if (! (c == ' ' || c == '\t' || c == ',')) | |
| 275 { | |
| 276 tmp_stream.putback (c); | |
| 277 break; | |
| 278 } | |
| 279 } | |
| 280 | |
| 281 if (tmp_stream.eof ()) | |
| 282 break; | |
| 283 } | |
| 284 else | |
| 285 { | |
| 286 error ("load: failed to read matrix from file `%s'", | |
| 287 filename.c_str ()); | |
| 288 | |
| 289 return retval; | |
| 290 } | |
| 291 | |
| 292 } | |
| 293 } | |
| 294 } | |
| 295 | |
| 296 if (is || is.eof ()) | |
| 297 { | |
| 5775 | 298 // FIXME -- not sure this is best, but it works. |
| 4634 | 299 |
| 300 if (is.eof ()) | |
| 301 is.clear (); | |
| 302 | |
| 5275 | 303 octave_idx_type expected = nr * nc; |
| 4634 | 304 |
| 305 if (expected == total_count) | |
| 306 { | |
| 307 tc = tmp; | |
| 308 retval = varname; | |
| 309 } | |
| 310 else | |
| 311 error ("load: expected %d elements, found %d", | |
| 312 expected, total_count); | |
| 313 } | |
| 314 else | |
| 315 error ("load: failed to read matrix from file `%s'", | |
| 316 filename.c_str ()); | |
| 317 } | |
| 318 else | |
| 319 error ("load: unable to extract matrix size from file `%s'", | |
| 320 filename.c_str ()); | |
| 321 } | |
| 322 else | |
| 323 error ("load: unable to convert filename `%s' to valid identifier", | |
| 324 filename.c_str ()); | |
| 325 | |
| 326 return retval; | |
| 327 } | |
| 328 | |
| 5938 | 329 bool |
| 330 save_mat_ascii_data (std::ostream& os, const octave_value& val, | |
|
8425
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
331 int precision, bool tabs) |
| 5938 | 332 { |
| 333 bool success = true; | |
| 334 | |
| 335 if (val.is_complex_type ()) | |
| 336 warning ("save: omitting imaginary part for ASCII file"); | |
| 337 | |
| 338 Matrix m = val.matrix_value (true); | |
| 339 | |
| 340 if (error_state) | |
| 341 { | |
| 342 success = false; | |
| 343 | |
| 344 error_state = 0; | |
| 345 } | |
| 346 else | |
| 5951 | 347 { |
| 348 long old_precision = os.precision (); | |
| 349 | |
| 350 os.precision (precision); | |
| 5938 | 351 |
| 5951 | 352 std::ios::fmtflags oflags |
| 353 = os.flags (static_cast<std::ios::fmtflags> (std::ios::scientific)); | |
| 354 | |
|
8425
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
355 if (tabs) |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
356 { |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
357 for (octave_idx_type i = 0; i < m.rows (); i++) |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
358 { |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
359 for (octave_idx_type j = 0; j < m.cols (); j++) |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
360 { |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
361 // Omit leading tabs. |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
362 if (j != 0) os << '\t'; |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
363 octave_write_double (os, m (i, j)); |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
364 } |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
365 os << "\n"; |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
366 } |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
367 } |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
368 else |
|
2e777f5135a3
support -tabs option for save -ascii
Jaroslav Hajek <highegg@gmail.com>
parents:
8021
diff
changeset
|
369 os << m; |
| 5951 | 370 |
| 371 os.flags (oflags); | |
| 372 | |
| 373 os.precision (old_precision); | |
| 374 } | |
| 5938 | 375 |
| 376 return (os && success); | |
| 377 } | |
| 378 | |
| 4634 | 379 /* |
| 380 ;;; Local Variables: *** | |
| 381 ;;; mode: C++ *** | |
| 382 ;;; End: *** | |
| 383 */ | |
| 384 |