2 This file is part of LilyPond, the GNU music typesetter.
4 Copyright (C) 1997--2014 Jan Nieuwenhuizen <janneke@gnu.org>
5 Han-Wen Nienhuys <hanwen@xs4all.nl>
7 LilyPond is free software: you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation, either version 3 of the License, or
10 (at your option) any later version.
12 LilyPond is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
17 You should have received a copy of the GNU General Public License
18 along with LilyPond. If not, see <http://www.gnu.org/licenses/>.
22 #define _GLIBCXX_HAVE_MBSTATE_T
24 #endif /* GCC_MAJOR < 4 */
26 #include "source-file.hh"
36 #define istringstream(x) istrstream (x, length ())
40 #include "file-name-map.hh"
41 #include "international.hh"
46 Source_file::load_stdin ()
50 while ((c = fgetc (stdin)) != EOF)
51 characters_.push_back ((char)c);
55 return contents of FILENAME. *Not 0-terminated!*
58 gulp_file (const string &filename, int desired_size)
60 /* "b" must ensure to open literally, avoiding text (CR/LF)
62 FILE *f = fopen (filename.c_str (), "rb");
65 warning (_f ("cannot open file: `%s'", filename.c_str ()));
71 fseek (f, 0, SEEK_END);
72 int real_size = ftell (f);
73 int read_count = real_size;
76 read_count = min (read_count, desired_size);
80 char *str = new char[read_count + 1];
83 int bytes_read = fread (str, sizeof (char), read_count, f);
84 if (bytes_read != read_count)
85 warning (_f ("expected to read %d characters, got %d", bytes_read,
88 int filesize = bytes_read;
91 cxx_arr.resize (filesize);
93 copy (str, str + filesize, cxx_arr.begin ());
109 Source_file::Source_file (const string &filename, const string &data)
115 characters_.resize (data.length ());
116 copy (data.begin (), data.end (), characters_.begin ());
118 characters_.push_back (0);
122 for (vsize i = 0; i < characters_.size (); i++)
123 if (characters_[i] == '\n')
124 newline_locations_.push_back (&characters_[0] + i);
127 Source_file::Source_file (const string &filename_string)
131 name_ = filename_string;
133 if (filename_string == "-")
137 characters_ = gulp_file (filename_string, -1);
140 characters_.push_back (0);
144 for (vsize i = 0; i < characters_.size (); i++)
145 if (characters_[i] == '\n')
146 newline_locations_.push_back (&characters_[0] + i);
150 Source_file::init_port ()
152 SCM str = scm_from_locale_string (c_str ());
153 str_port_ = scm_mkstrport (SCM_INUM0, str, SCM_OPN | SCM_RDNG, __FUNCTION__);
154 scm_set_port_filename_x (str_port_, ly_string2scm (name_));
158 Source_file::get_istream ()
162 if (length ()) // can-t this be done without such a hack?
163 istream_ = new istringstream (c_str ());
166 istream_ = new istringstream ("");
167 istream_->setstate (ios::eofbit);
168 // istream_->set (ios::eofbit);
175 Source_file::file_line_column_string (char const *context_str0) const
178 return " (" + _ ("position unknown") + ")";
181 int l, ch, col, offset;
182 get_counts (context_str0, &l, &ch, &col, &offset);
184 return name_string () + ":" + ::to_string (l)
185 + ":" + ::to_string (col + 1);
190 Source_file::quote_input (char const *pos_str0) const
192 if (!contains (pos_str0))
193 return " (" + _ ("position unknown") + ")";
195 int l, ch, col, offset;
196 get_counts (pos_str0, &l, &ch, &col, &offset);
197 string line = line_string (pos_str0);
198 string context = line.substr (0, offset)
200 + ::to_string (' ', col)
201 + line.substr (offset, line.length () - offset);
206 Source_file::name_string () const
208 return map_file_name (name_);
211 Source_file::~Source_file ()
217 Source_file::line_slice (char const *pos_str0) const
219 if (!contains (pos_str0))
222 char const *data_str0 = c_str ();
223 char const *eof_C_ = data_str0 + length ();
225 if (pos_str0 == eof_C_)
227 char const *begin_str0 = pos_str0;
228 while (begin_str0 > data_str0)
229 if (*--begin_str0 == '\n')
235 char const *end_str0 = pos_str0;
236 while (end_str0 < eof_C_)
237 if (*end_str0++ == '\n')
243 return Slice (begin_str0 - data_str0, end_str0 - data_str0);
247 Source_file::line_string (char const *pos_str0) const
249 if (!contains (pos_str0))
252 Slice line = line_slice (pos_str0);
253 char const *data_str0 = c_str ();
254 return string (data_str0 + line[LEFT], line.length ());
258 Source_file::get_counts (char const *pos_str0,
262 int *byte_offset) const
264 // Initialize arguments to defaults, needed if pos_str0 is not in source
270 if (!contains (pos_str0))
273 *line_number = get_line (pos_str0);
275 Slice line = line_slice (pos_str0);
276 char const *data = c_str ();
277 char const *line_start = (char const *)data + line[LEFT];
279 ssize left = (char const *) pos_str0 - line_start;
282 string line_begin (line_start, left);
283 char const *line_chars = line_begin.c_str ();
285 for (; left > 0; --left, ++line_chars)
287 // Skip UTF-8 continuation bytes. This is simplistic but
288 // robust, and we warn against non-UTF-8 input in the lexer
289 // already. In the case of non-UTF-8 or of this function being
290 // called in mid-character, the results are somewhat arbitrary,
291 // but there is no really sane definition anyway.
292 if ((*line_chars & 0xc0) == 0x80)
295 if (*line_chars == '\t')
296 (*column) = (*column / 8 + 1) * 8;
305 Source_file::contains (char const *pos_str0) const
307 return (pos_str0 && (pos_str0 >= c_str ()) && (pos_str0 <= c_str () + length ()));
311 Source_file::get_line (char const *pos_str0) const
313 if (!contains (pos_str0))
316 if (!newline_locations_.size ())
317 return 1 + line_offset_;
319 /* this will find the '\n' character at the end of our line */
320 vsize lo = lower_bound (newline_locations_,
322 less<char const *> ());
324 /* the return value will be indexed from 1 */
325 return lo + 1 + line_offset_;
329 Source_file::set_line (char const *pos_str0, int line)
333 int current_line = get_line (pos_str0);
334 line_offset_ += line - current_line;
336 assert (line == get_line (pos_str0));
343 Source_file::length () const
345 return characters_.size ();
349 Source_file::c_str () const
351 return &characters_[0];
355 Source_file::get_port () const
360 /****************************************************************/
362 #include "ly-smobs.icc"
364 IMPLEMENT_SMOBS (Source_file);
365 IMPLEMENT_DEFAULT_EQUAL_P (Source_file);
366 IMPLEMENT_TYPE_P (Source_file, "ly:source-file?");
369 Source_file::mark_smob (SCM smob)
371 Source_file *sc = (Source_file *) SCM_CELL_WORD_1 (smob);
373 return sc->str_port_;
377 Source_file::print_smob (SCM smob, SCM port, scm_print_state *)
379 Source_file *sc = (Source_file *) SCM_CELL_WORD_1 (smob);
381 scm_puts ("#<Source_file ", port);
382 scm_puts (sc->name_.c_str (), port);
384 /* Do not print properties, that is too much hassle. */
385 scm_puts (" >", port);