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"
38 #include "file-name-map.hh"
39 #include "international.hh"
44 using std::istringstream;
46 #define istringstream(x) std::istrstream (x, length ())
54 Source_file::load_stdin ()
58 while ((c = fgetc (stdin)) != EOF)
59 characters_.push_back ((char)c);
63 return contents of FILENAME. *Not 0-terminated!*
66 gulp_file (const string &filename, int desired_size)
68 /* "b" must ensure to open literally, avoiding text (CR/LF)
70 FILE *f = fopen (filename.c_str (), "rb");
73 warning (_f ("cannot open file: `%s'", filename.c_str ()));
79 fseek (f, 0, SEEK_END);
80 int real_size = ftell (f);
81 int read_count = real_size;
84 read_count = std::min (read_count, desired_size);
88 char *str = new char[read_count + 1];
91 int bytes_read = fread (str, sizeof (char), read_count, f);
92 if (bytes_read != read_count)
93 warning (_f ("expected to read %d characters, got %d", bytes_read,
96 int filesize = bytes_read;
99 cxx_arr.resize (filesize);
101 copy (str, str + filesize, cxx_arr.begin ());
116 Source_file::Source_file (const string &filename, const string &data)
122 characters_.resize (data.length ());
123 copy (data.begin (), data.end (), characters_.begin ());
125 characters_.push_back (0);
129 for (vsize i = 0; i < characters_.size (); i++)
130 if (characters_[i] == '\n')
131 newline_locations_.push_back (&characters_[0] + i);
134 Source_file::Source_file (const string &filename_string)
138 name_ = filename_string;
140 if (filename_string == "-")
144 characters_ = gulp_file (filename_string, -1);
147 characters_.push_back (0);
151 for (vsize i = 0; i < characters_.size (); i++)
152 if (characters_[i] == '\n')
153 newline_locations_.push_back (&characters_[0] + i);
157 Source_file::init_port ()
159 // This is somewhat icky: the string will in general be in utf8, but
160 // we do our own utf8 encoding and verification in the parser, so we
161 // use the no-conversion equivalent of latin1
162 SCM str = scm_from_latin1_string (c_str ());
163 str_port_ = scm_mkstrport (SCM_INUM0, str, SCM_OPN | SCM_RDNG, __FUNCTION__);
164 scm_set_port_filename_x (str_port_, ly_string2scm (name_));
168 Source_file::get_istream ()
172 if (length ()) // can-t this be done without such a hack?
173 istream_ = new istringstream (c_str ());
176 istream_ = new istringstream ("");
177 istream_->setstate (std::ios::eofbit);
178 // istream_->set (std::ios::eofbit);
185 Source_file::file_line_column_string (char const *context_str0) const
188 return " (" + _ ("position unknown") + ")";
191 int l, ch, col, offset;
192 get_counts (context_str0, &l, &ch, &col, &offset);
194 return name_string () + ":" + ::to_string (l)
195 + ":" + ::to_string (col + 1);
200 Source_file::quote_input (char const *pos_str0) const
202 if (!contains (pos_str0))
203 return " (" + _ ("position unknown") + ")";
205 int l, ch, col, offset;
206 get_counts (pos_str0, &l, &ch, &col, &offset);
207 string line = line_string (pos_str0);
208 string context = line.substr (0, offset)
210 + ::to_string (' ', col)
211 + line.substr (offset, line.length () - offset);
216 Source_file::name_string () const
218 return map_file_name (name_);
221 Source_file::~Source_file ()
227 Source_file::line_slice (char const *pos_str0) const
229 if (!contains (pos_str0))
232 char const *data_str0 = c_str ();
233 char const *eof_C_ = data_str0 + length ();
235 if (pos_str0 == eof_C_)
237 char const *begin_str0 = pos_str0;
238 while (begin_str0 > data_str0)
239 if (*--begin_str0 == '\n')
245 char const *end_str0 = pos_str0;
246 while (end_str0 < eof_C_)
247 if (*end_str0++ == '\n')
253 return Slice (begin_str0 - data_str0, end_str0 - data_str0);
257 Source_file::line_string (char const *pos_str0) const
259 if (!contains (pos_str0))
262 Slice line = line_slice (pos_str0);
263 char const *data_str0 = c_str ();
264 return string (data_str0 + line[LEFT], line.length ());
268 Source_file::get_counts (char const *pos_str0,
272 int *byte_offset) const
274 // Initialize arguments to defaults, needed if pos_str0 is not in source
280 if (!contains (pos_str0))
283 *line_number = get_line (pos_str0);
285 Slice line = line_slice (pos_str0);
286 char const *data = c_str ();
287 char const *line_start = (char const *)data + line[LEFT];
289 ssize left = (char const *) pos_str0 - line_start;
292 string line_begin (line_start, left);
293 char const *line_chars = line_begin.c_str ();
295 for (; left > 0; --left, ++line_chars)
297 // Skip UTF-8 continuation bytes. This is simplistic but
298 // robust, and we warn against non-UTF-8 input in the lexer
299 // already. In the case of non-UTF-8 or of this function being
300 // called in mid-character, the results are somewhat arbitrary,
301 // but there is no really sane definition anyway.
302 if ((*line_chars & 0xc0) == 0x80)
305 if (*line_chars == '\t')
306 (*column) = (*column / 8 + 1) * 8;
315 Source_file::contains (char const *pos_str0) const
317 return (pos_str0 && (pos_str0 >= c_str ()) && (pos_str0 <= c_str () + length ()));
321 Source_file::get_line (char const *pos_str0) const
323 if (!contains (pos_str0))
326 if (!newline_locations_.size ())
327 return 1 + line_offset_;
329 /* this will find the '\n' character at the end of our line */
330 vsize lo = lower_bound (newline_locations_,
332 std::less<char const *> ());
334 /* the return value will be indexed from 1 */
335 return lo + 1 + line_offset_;
339 Source_file::set_line (char const *pos_str0, int line)
343 int current_line = get_line (pos_str0);
344 line_offset_ += line - current_line;
346 assert (line == get_line (pos_str0));
353 Source_file::length () const
355 return characters_.size ();
359 Source_file::c_str () const
361 return &characters_[0];
365 Source_file::get_port () const
370 /****************************************************************/
373 const char Source_file::type_p_name_[] = "ly:source-file?";
376 Source_file::mark_smob () const
382 Source_file::print_smob (SCM port, scm_print_state *) const
384 scm_puts ("#<Source_file ", port);
385 scm_puts (name_.c_str (), port);
387 /* Do not print properties, that is too much hassle. */
388 scm_puts (" >", port);