X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=lily%2Fsource-file.cc;h=14fdf2beb3b6ef4c10ff835c60deeca2a5811157;hb=90e4d7057f3857da049dfda3d130017d4719bd6b;hp=041c046d2bfc2fa4761df56005f2e76692f962f7;hpb=a066a93ee74edebb9d238a1bac93c3bc7e8e6e4a;p=lilypond.git

diff --git a/lily/source-file.cc b/lily/source-file.cc
index 041c046d2b..14fdf2beb3 100644
--- a/lily/source-file.cc
+++ b/lily/source-file.cc
@@ -1,7 +1,7 @@
 /*
   This file is part of LilyPond, the GNU music typesetter.
 
-  Copyright (C) 1997--2011 Jan Nieuwenhuizen <janneke@gnu.org>
+  Copyright (C) 1997--2015 Jan Nieuwenhuizen <janneke@gnu.org>
   Han-Wen Nienhuys <hanwen@xs4all.nl>
 
   LilyPond is free software: you can redistribute it and/or modify
@@ -55,7 +55,7 @@ Source_file::load_stdin ()
   return contents of FILENAME. *Not 0-terminated!*
  */
 vector<char>
-gulp_file (string filename, int desired_size)
+gulp_file (const string &filename, int desired_size)
 {
   /* "b" must ensure to open literally, avoiding text (CR/LF)
      conversions.  */
@@ -102,11 +102,10 @@ Source_file::init ()
   istream_ = 0;
   line_offset_ = 0;
   str_port_ = SCM_EOL;
-  self_scm_ = SCM_EOL;
   smobify_self ();
 }
 
-Source_file::Source_file (string filename, string data)
+Source_file::Source_file (const string &filename, const string &data)
 {
   init ();
 
@@ -124,7 +123,7 @@ Source_file::Source_file (string filename, string data)
       newline_locations_.push_back (&characters_[0] + i);
 }
 
-Source_file::Source_file (string filename_string)
+Source_file::Source_file (const string &filename_string)
 {
   init ();
 
@@ -149,7 +148,10 @@ Source_file::Source_file (string filename_string)
 void
 Source_file::init_port ()
 {
-  SCM str = scm_from_locale_string (c_str ());
+  // This is somewhat icky: the string will in general be in utf8, but
+  // we do our own utf8 encoding and verification in the parser, so we
+  // use the no-conversion equivalent of latin1
+  SCM str = scm_from_latin1_string (c_str ());
   str_port_ = scm_mkstrport (SCM_INUM0, str, SCM_OPN | SCM_RDNG, __FUNCTION__);
   scm_set_port_filename_x (str_port_, ly_string2scm (name_));
 }
@@ -181,8 +183,8 @@ Source_file::file_line_column_string (char const *context_str0) const
       int l, ch, col, offset;
       get_counts (context_str0, &l, &ch, &col, &offset);
 
-      return name_string () + ":" + to_string (l)
-             + ":" + to_string (col);
+      return name_string () + ":" + ::to_string (l)
+             + ":" + ::to_string (col + 1);
     }
 }
 
@@ -196,8 +198,8 @@ Source_file::quote_input (char const *pos_str0) const
   get_counts (pos_str0, &l, &ch, &col, &offset);
   string line = line_string (pos_str0);
   string context = line.substr (0, offset)
-                   + to_string ('\n')
-                   + to_string (' ', col)
+                   + ::to_string ('\n')
+                   + ::to_string (' ', col)
                    + line.substr (offset, line.length () - offset);
   return context;
 }
@@ -277,30 +279,27 @@ Source_file::get_counts (char const *pos_str0,
   char const *line_start = (char const *)data + line[LEFT];
 
   ssize left = (char const *) pos_str0 - line_start;
+  *byte_offset = left;
+
   string line_begin (line_start, left);
   char const *line_chars = line_begin.c_str ();
 
-  while (left > 0)
+  for (; left > 0; --left, ++line_chars)
     {
-      size_t thislen = utf8_char_len (*line_chars);
-
-      if (thislen == 1 && line_chars[0] == '\t')
+      // Skip UTF-8 continuation bytes.  This is simplistic but
+      // robust, and we warn against non-UTF-8 input in the lexer
+      // already.  In the case of non-UTF-8 or of this function being
+      // called in mid-character, the results are somewhat arbitrary,
+      // but there is no really sane definition anyway.
+      if ((*line_chars & 0xc0) == 0x80)
+        continue;
+
+      if (*line_chars == '\t')
         (*column) = (*column / 8 + 1) * 8;
       else
         (*column)++;
 
       (*line_char)++;
-
-      /*
-        To have decent output in UTF-8 aware terminals,
-        we must keep track of the number of bytes from
-        the left edge of the terminal.
-      */
-      *byte_offset += thislen;
-
-      /* Advance past this character. */
-      line_chars += thislen;
-      left -= thislen;
     }
 }
 
@@ -317,7 +316,7 @@ Source_file::get_line (char const *pos_str0) const
     return 0;
 
   if (!newline_locations_.size ())
-    return 1;
+    return 1 + line_offset_;
 
   /* this will find the '\n' character at the end of our line */
   vsize lo = lower_bound (newline_locations_,
@@ -331,10 +330,15 @@ Source_file::get_line (char const *pos_str0) const
 void
 Source_file::set_line (char const *pos_str0, int line)
 {
-  int current_line = get_line (pos_str0);
-  line_offset_ += line - current_line;
+  if (pos_str0)
+    {
+      int current_line = get_line (pos_str0);
+      line_offset_ += line - current_line;
 
-  assert (line == get_line (pos_str0));
+      assert (line == get_line (pos_str0));
+    }
+  else
+    line_offset_ = line;
 }
 
 int
@@ -357,30 +361,22 @@ Source_file::get_port () const
 
 /****************************************************************/
 
-#include "ly-smobs.icc"
 
-IMPLEMENT_SMOBS (Source_file);
-IMPLEMENT_DEFAULT_EQUAL_P (Source_file);
-IMPLEMENT_TYPE_P (Source_file, "ly:source-file?");
+const char * const Source_file::type_p_name_ = "ly:source-file?";
 
 SCM
-Source_file::mark_smob (SCM smob)
+Source_file::mark_smob () const
 {
-  Source_file *sc = (Source_file *) SCM_CELL_WORD_1 (smob);
-
-  return sc->str_port_;
+  return str_port_;
 }
 
 int
-Source_file::print_smob (SCM smob, SCM port, scm_print_state *)
+Source_file::print_smob (SCM port, scm_print_state *) const
 {
-  Source_file *sc = (Source_file *) SCM_CELL_WORD_1 (smob);
-
   scm_puts ("#<Source_file ", port);
-  scm_puts (sc->name_.c_str (), port);
+  scm_puts (name_.c_str (), port);
 
   /* Do not print properties, that is too much hassle.  */
   scm_puts (" >", port);
   return 1;
 }
-