Mercurial > octave-nkf
annotate src/ls-mat-ascii.cc @ 11944:278afaecddd4 release-3-0-x
fix leaving stray '\r' in stream when reading from CRLF data file
* * *
fix CRLF issues with text-mode reading in windows when loading ascii data
| author | Benjamin Lindner <lindnerb@users.sourceforge.net> |
|---|---|
| date | Wed, 18 Mar 2009 15:23:14 +0100 |
| parents | a1dbe9d80eee |
| children | 21f8e9c91b77 |
| rev | line source |
|---|---|
| 4634 | 1 /* |
| 2 | |
| 7017 | 3 Copyright (C) 1996, 1997, 2003, 2004, 2005, 2006, 2007 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" |
| 54 #include "oct-obj.h" | |
| 55 #include "oct-map.h" | |
| 56 #include "ov-cell.h" | |
| 57 #include "pager.h" | |
| 58 #include "pt-exp.h" | |
| 59 #include "symtab.h" | |
| 60 #include "sysdep.h" | |
| 61 #include "unwind-prot.h" | |
| 62 #include "utils.h" | |
| 63 #include "variables.h" | |
| 64 #include "version.h" | |
| 65 #include "dMatrix.h" | |
| 66 | |
| 67 #include "ls-mat-ascii.h" | |
|
11944
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
68 #include "ls-ascii-helper.h" |
| 4634 | 69 |
| 70 static std::string | |
| 71 get_mat_data_input_line (std::istream& is) | |
| 72 { | |
| 73 std::string retval; | |
| 74 | |
| 75 bool have_data = false; | |
| 76 | |
| 77 do | |
| 78 { | |
| 79 retval = ""; | |
| 80 | |
| 81 char c; | |
| 82 while (is.get (c)) | |
| 83 { | |
| 84 if (c == '\n' || c == '\r') | |
|
11944
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
85 { |
|
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
86 // Let skip_until_newline handle CR/LF issues... |
|
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
87 skip_until_newline (is, false); |
|
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
88 break; |
|
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
89 } |
| 4634 | 90 |
| 91 if (c == '%' || c == '#') | |
| 92 { | |
| 93 // skip to end of line | |
|
11944
278afaecddd4
fix leaving stray '\r' in stream when reading from CRLF data file
Benjamin Lindner <lindnerb@users.sourceforge.net>
parents:
7017
diff
changeset
|
94 skip_until_newline (is, false); |
| 4634 | 95 |
| 96 break; | |
| 97 } | |
| 98 | |
| 99 if (! is.eof ()) | |
| 100 { | |
| 101 if (! have_data && c != ' ' && c != '\t') | |
| 102 have_data = true; | |
| 103 | |
| 104 retval += c; | |
| 105 } | |
| 106 } | |
| 107 } | |
| 108 while (! (have_data || is.eof ())); | |
| 109 | |
| 110 return retval; | |
| 111 } | |
| 112 | |
| 113 static void | |
| 5275 | 114 get_lines_and_columns (std::istream& is, const std::string& filename, octave_idx_type& nr, octave_idx_type& nc) |
| 4634 | 115 { |
| 116 std::streampos pos = is.tellg (); | |
| 117 | |
| 118 int file_line_number = 0; | |
| 119 | |
| 120 nr = 0; | |
| 121 nc = 0; | |
| 122 | |
| 123 while (is && ! error_state) | |
| 124 { | |
| 125 OCTAVE_QUIT; | |
| 126 | |
| 127 std::string buf = get_mat_data_input_line (is); | |
| 128 | |
| 129 file_line_number++; | |
| 130 | |
| 131 size_t beg = buf.find_first_not_of (", \t"); | |
| 132 | |
| 133 // If we see a CR as the last character in the buffer, we had a | |
| 134 // CRLF pair as the line separator. Any other CR in the text | |
| 135 // will not be considered as whitespace. | |
| 136 | |
| 137 if (beg != NPOS && buf[beg] == '\r' && beg == buf.length () - 1) | |
| 138 { | |
| 139 // We had a blank line ending with a CRLF. Handle it the | |
| 140 // same as an empty line. | |
| 141 beg = NPOS; | |
| 142 } | |
| 143 | |
| 5275 | 144 octave_idx_type tmp_nc = 0; |
| 4634 | 145 |
| 146 while (beg != NPOS) | |
| 147 { | |
| 148 tmp_nc++; | |
| 149 | |
| 150 size_t end = buf.find_first_of (", \t", beg); | |
| 151 | |
| 152 if (end != NPOS) | |
| 153 { | |
| 154 beg = buf.find_first_not_of (", \t", end); | |
| 155 | |
| 6856 | 156 if (beg == NPOS || (buf[beg] == '\r' && |
| 157 beg == buf.length () - 1)) | |
| 4634 | 158 { |
| 159 // We had a line with trailing spaces and | |
| 160 // ending with a CRLF, so this should look like EOL, | |
| 161 // not a new colum. | |
| 162 break; | |
| 163 } | |
| 164 } | |
| 165 else | |
| 166 break; | |
| 167 } | |
| 168 | |
| 169 if (tmp_nc > 0) | |
| 170 { | |
| 171 if (nc == 0) | |
| 172 { | |
| 173 nc = tmp_nc; | |
| 174 nr++; | |
| 175 } | |
| 176 else if (nc == tmp_nc) | |
| 177 nr++; | |
| 178 else | |
| 179 error ("load: %s: inconsistent number of columns near line %d", | |
| 180 filename.c_str (), file_line_number); | |
| 181 } | |
| 182 } | |
| 183 | |
| 184 if (nr == 0 || nc == 0) | |
| 185 error ("load: file `%s' seems to be empty!", filename.c_str ()); | |
| 186 | |
| 187 is.clear (); | |
| 4643 | 188 is.seekg (pos); |
| 4634 | 189 } |
| 190 | |
| 191 // Extract a matrix from a file of numbers only. | |
| 192 // | |
| 193 // Comments are not allowed. The file should only have numeric values. | |
| 194 // | |
| 195 // Reads the file twice. Once to find the number of rows and columns, | |
| 196 // and once to extract the matrix. | |
| 197 // | |
| 198 // FILENAME is used for error messages. | |
| 199 // | |
| 200 // This format provides no way to tag the data as global. | |
| 201 | |
| 202 std::string | |
| 203 read_mat_ascii_data (std::istream& is, const std::string& filename, | |
| 204 octave_value& tc) | |
| 205 { | |
| 206 std::string retval; | |
| 207 | |
| 208 std::string varname; | |
| 209 | |
| 210 size_t pos = filename.rfind ('/'); | |
| 211 | |
| 212 if (pos != NPOS) | |
| 213 varname = filename.substr (pos+1); | |
| 214 else | |
| 215 varname = filename; | |
| 216 | |
| 4867 | 217 pos = varname.rfind ('.'); |
| 4634 | 218 |
| 219 if (pos != NPOS) | |
| 220 varname = varname.substr (0, pos); | |
| 221 | |
| 222 size_t len = varname.length (); | |
| 223 for (size_t i = 0; i < len; i++) | |
| 224 { | |
| 225 char c = varname[i]; | |
| 226 if (! (isalnum (c) || c == '_')) | |
| 227 varname[i] = '_'; | |
| 228 } | |
| 229 | |
| 4867 | 230 if (is_keyword (varname) || ! isalpha (varname[0])) |
| 4634 | 231 varname.insert (0, "X"); |
| 232 | |
| 233 if (valid_identifier (varname)) | |
| 234 { | |
| 5275 | 235 octave_idx_type nr = 0; |
| 236 octave_idx_type nc = 0; | |
| 4634 | 237 |
| 238 int total_count = 0; | |
| 239 | |
| 240 get_lines_and_columns (is, filename, nr, nc); | |
| 241 | |
| 242 OCTAVE_QUIT; | |
| 243 | |
| 244 if (! error_state && nr > 0 && nc > 0) | |
| 245 { | |
| 246 Matrix tmp (nr, nc); | |
| 247 | |
| 248 if (nr < 1 || nc < 1) | |
| 249 is.clear (std::ios::badbit); | |
| 250 else | |
| 251 { | |
| 252 double d; | |
| 5275 | 253 for (octave_idx_type i = 0; i < nr; i++) |
| 4634 | 254 { |
| 255 std::string buf = get_mat_data_input_line (is); | |
| 256 | |
| 257 std::istringstream tmp_stream (buf); | |
| 258 | |
| 5275 | 259 for (octave_idx_type j = 0; j < nc; j++) |
| 4634 | 260 { |
| 261 OCTAVE_QUIT; | |
| 262 | |
| 263 d = octave_read_double (tmp_stream); | |
| 264 | |
| 265 if (tmp_stream || tmp_stream.eof ()) | |
| 266 { | |
| 267 tmp.elem (i, j) = d; | |
| 268 total_count++; | |
| 269 | |
| 270 // Skip whitespace and commas. | |
| 271 char c; | |
| 272 while (1) | |
| 273 { | |
| 274 tmp_stream >> c; | |
| 275 | |
| 276 if (! tmp_stream) | |
| 277 break; | |
| 278 | |
| 279 if (! (c == ' ' || c == '\t' || c == ',')) | |
| 280 { | |
| 281 tmp_stream.putback (c); | |
| 282 break; | |
| 283 } | |
| 284 } | |
| 285 | |
| 286 if (tmp_stream.eof ()) | |
| 287 break; | |
| 288 } | |
| 289 else | |
| 290 { | |
| 291 error ("load: failed to read matrix from file `%s'", | |
| 292 filename.c_str ()); | |
| 293 | |
| 294 return retval; | |
| 295 } | |
| 296 | |
| 297 } | |
| 298 } | |
| 299 } | |
| 300 | |
| 301 if (is || is.eof ()) | |
| 302 { | |
| 5775 | 303 // FIXME -- not sure this is best, but it works. |
| 4634 | 304 |
| 305 if (is.eof ()) | |
| 306 is.clear (); | |
| 307 | |
| 5275 | 308 octave_idx_type expected = nr * nc; |
| 4634 | 309 |
| 310 if (expected == total_count) | |
| 311 { | |
| 312 tc = tmp; | |
| 313 retval = varname; | |
| 314 } | |
| 315 else | |
| 316 error ("load: expected %d elements, found %d", | |
| 317 expected, total_count); | |
| 318 } | |
| 319 else | |
| 320 error ("load: failed to read matrix from file `%s'", | |
| 321 filename.c_str ()); | |
| 322 } | |
| 323 else | |
| 324 error ("load: unable to extract matrix size from file `%s'", | |
| 325 filename.c_str ()); | |
| 326 } | |
| 327 else | |
| 328 error ("load: unable to convert filename `%s' to valid identifier", | |
| 329 filename.c_str ()); | |
| 330 | |
| 331 return retval; | |
| 332 } | |
| 333 | |
| 5938 | 334 bool |
| 335 save_mat_ascii_data (std::ostream& os, const octave_value& val, | |
| 336 int precision) | |
| 337 { | |
| 338 bool success = true; | |
| 339 | |
| 340 if (val.is_complex_type ()) | |
| 341 warning ("save: omitting imaginary part for ASCII file"); | |
| 342 | |
| 343 Matrix m = val.matrix_value (true); | |
| 344 | |
| 345 if (error_state) | |
| 346 { | |
| 347 success = false; | |
| 348 | |
| 349 error_state = 0; | |
| 350 } | |
| 351 else | |
| 5951 | 352 { |
| 353 long old_precision = os.precision (); | |
| 354 | |
| 355 os.precision (precision); | |
| 5938 | 356 |
| 5951 | 357 std::ios::fmtflags oflags |
| 358 = os.flags (static_cast<std::ios::fmtflags> (std::ios::scientific)); | |
| 359 | |
| 360 os << m; | |
| 361 | |
| 362 os.flags (oflags); | |
| 363 | |
| 364 os.precision (old_precision); | |
| 365 } | |
| 5938 | 366 |
| 367 return (os && success); | |
| 368 } | |
| 369 | |
| 4634 | 370 /* |
| 371 ;;; Local Variables: *** | |
| 372 ;;; mode: C++ *** | |
| 373 ;;; End: *** | |
| 374 */ | |
| 375 |