2 This file is part of LilyPond, the GNU music typesetter.
4 Copyright (C) 1997--2015 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"
45 ADD_SMOB_INIT (Source_file);
48 Source_file::load_stdin ()
52 while ((c = fgetc (stdin)) != EOF)
53 characters_.push_back ((char)c);
57 return contents of FILENAME. *Not 0-terminated!*
60 gulp_file (const string &filename, int desired_size)
62 /* "b" must ensure to open literally, avoiding text (CR/LF)
64 FILE *f = fopen (filename.c_str (), "rb");
67 warning (_f ("cannot open file: `%s'", filename.c_str ()));
73 fseek (f, 0, SEEK_END);
74 int real_size = ftell (f);
75 int read_count = real_size;
78 read_count = min (read_count, desired_size);
82 char *str = new char[read_count + 1];
85 int bytes_read = fread (str, sizeof (char), read_count, f);
86 if (bytes_read != read_count)
87 warning (_f ("expected to read %d characters, got %d", bytes_read,
90 int filesize = bytes_read;
93 cxx_arr.resize (filesize);
95 copy (str, str + filesize, cxx_arr.begin ());
110 Source_file::Source_file (const string &filename, const string &data)
116 characters_.resize (data.length ());
117 copy (data.begin (), data.end (), characters_.begin ());
119 characters_.push_back (0);
123 for (vsize i = 0; i < characters_.size (); i++)
124 if (characters_[i] == '\n')
125 newline_locations_.push_back (&characters_[0] + i);
128 Source_file::Source_file (const string &filename_string)
132 name_ = filename_string;
134 if (filename_string == "-")
138 characters_ = gulp_file (filename_string, -1);
141 characters_.push_back (0);
145 for (vsize i = 0; i < characters_.size (); i++)
146 if (characters_[i] == '\n')
147 newline_locations_.push_back (&characters_[0] + i);
151 Source_file::init_port ()
153 // This is somewhat icky: the string will in general be in utf8, but
154 // we do our own utf8 encoding and verification in the parser, so we
155 // use the no-conversion equivalent of latin1
156 SCM str = scm_from_latin1_string (c_str ());
157 str_port_ = scm_mkstrport (SCM_INUM0, str, SCM_OPN | SCM_RDNG, __FUNCTION__);
158 scm_set_port_filename_x (str_port_, ly_string2scm (name_));
162 Source_file::get_istream ()
166 if (length ()) // can-t this be done without such a hack?
167 istream_ = new istringstream (c_str ());
170 istream_ = new istringstream ("");
171 istream_->setstate (ios::eofbit);
172 // istream_->set (ios::eofbit);
179 Source_file::file_line_column_string (char const *context_str0) const
182 return " (" + _ ("position unknown") + ")";
185 int l, ch, col, offset;
186 get_counts (context_str0, &l, &ch, &col, &offset);
188 return name_string () + ":" + ::to_string (l)
189 + ":" + ::to_string (col + 1);
194 Source_file::quote_input (char const *pos_str0) const
196 if (!contains (pos_str0))
197 return " (" + _ ("position unknown") + ")";
199 int l, ch, col, offset;
200 get_counts (pos_str0, &l, &ch, &col, &offset);
201 string line = line_string (pos_str0);
202 string context = line.substr (0, offset)
204 + ::to_string (' ', col)
205 + line.substr (offset, line.length () - offset);
210 Source_file::name_string () const
212 return map_file_name (name_);
215 Source_file::~Source_file ()
221 Source_file::line_slice (char const *pos_str0) const
223 if (!contains (pos_str0))
226 char const *data_str0 = c_str ();
227 char const *eof_C_ = data_str0 + length ();
229 if (pos_str0 == eof_C_)
231 char const *begin_str0 = pos_str0;
232 while (begin_str0 > data_str0)
233 if (*--begin_str0 == '\n')
239 char const *end_str0 = pos_str0;
240 while (end_str0 < eof_C_)
241 if (*end_str0++ == '\n')
247 return Slice (begin_str0 - data_str0, end_str0 - data_str0);
251 Source_file::line_string (char const *pos_str0) const
253 if (!contains (pos_str0))
256 Slice line = line_slice (pos_str0);
257 char const *data_str0 = c_str ();
258 return string (data_str0 + line[LEFT], line.length ());
262 Source_file::get_counts (char const *pos_str0,
266 int *byte_offset) const
268 // Initialize arguments to defaults, needed if pos_str0 is not in source
274 if (!contains (pos_str0))
277 *line_number = get_line (pos_str0);
279 Slice line = line_slice (pos_str0);
280 char const *data = c_str ();
281 char const *line_start = (char const *)data + line[LEFT];
283 ssize left = (char const *) pos_str0 - line_start;
286 string line_begin (line_start, left);
287 char const *line_chars = line_begin.c_str ();
289 for (; left > 0; --left, ++line_chars)
291 // Skip UTF-8 continuation bytes. This is simplistic but
292 // robust, and we warn against non-UTF-8 input in the lexer
293 // already. In the case of non-UTF-8 or of this function being
294 // called in mid-character, the results are somewhat arbitrary,
295 // but there is no really sane definition anyway.
296 if ((*line_chars & 0xc0) == 0x80)
299 if (*line_chars == '\t')
300 (*column) = (*column / 8 + 1) * 8;
309 Source_file::contains (char const *pos_str0) const
311 return (pos_str0 && (pos_str0 >= c_str ()) && (pos_str0 <= c_str () + length ()));
315 Source_file::get_line (char const *pos_str0) const
317 if (!contains (pos_str0))
320 if (!newline_locations_.size ())
321 return 1 + line_offset_;
323 /* this will find the '\n' character at the end of our line */
324 vsize lo = lower_bound (newline_locations_,
326 less<char const *> ());
328 /* the return value will be indexed from 1 */
329 return lo + 1 + line_offset_;
333 Source_file::set_line (char const *pos_str0, int line)
337 int current_line = get_line (pos_str0);
338 line_offset_ += line - current_line;
340 assert (line == get_line (pos_str0));
347 Source_file::length () const
349 return characters_.size ();
353 Source_file::c_str () const
355 return &characters_[0];
359 Source_file::get_port () const
364 /****************************************************************/
367 const char Source_file::type_p_name_[] = "ly:source-file?";
370 Source_file::mark_smob ()
376 Source_file::print_smob (SCM port, scm_print_state *)
378 scm_puts ("#<Source_file ", port);
379 scm_puts (name_.c_str (), port);
381 /* Do not print properties, that is too much hassle. */
382 scm_puts (" >", port);