/* PSPP - computes sample statistics.
Copyright (C) 1997-2004, 2006 Free Software Foundation, Inc.
- Written by Ben Pfaff <blp@gnu.org>.
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License as
02110-1301, USA. */
#include <config.h>
+
#include <language/data-io/data-reader.h>
+
#include <ctype.h>
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>
-#include <libpspp/alloc.h>
-#include <language/command.h>
-#include <libpspp/message.h>
-#include <language/data-io/file-handle.h>
+
#include <data/file-handle-def.h>
#include <data/file-name.h>
-#include <language/line-buffer.h>
+#include <data/procedure.h>
+#include <language/command.h>
+#include <language/data-io/file-handle.h>
#include <language/lexer/lexer.h>
+#include <language/prompt.h>
+#include <libpspp/alloc.h>
+#include <libpspp/assertion.h>
+#include <libpspp/message.h>
#include <libpspp/str.h>
-#include <procedure.h>
+
+#include "minmax.h"
+#include "size_max.h"
#include "gettext.h"
#define _(msgid) gettext (msgid)
struct dfm_reader
{
struct file_handle *fh; /* File handle. */
- struct file_locator where; /* Current location in data file. */
+ struct msg_locator where; /* Current location in data file. */
struct string line; /* Current line. */
struct string scratch; /* Extra line buffer. */
enum dfm_reader_flags flags; /* Zero or more of DFM_*. */
FILE *file; /* Associated file. */
size_t pos; /* Offset in line of current character. */
unsigned eof_cnt; /* # of attempts to advance past EOF. */
+ struct lexer *lexer; /* The lexer reading the file */
};
/* Closes reader R opened by dfm_open_reader(). */
return;
is_inline = r->fh == fh_inline_file ();
- file_name = is_inline ? NULL : xstrdup (fh_get_filename (r->fh));
+ file_name = is_inline ? NULL : xstrdup (fh_get_file_name (r->fh));
still_open = fh_close (r->fh, "data file", "rs");
if (still_open)
{
file between BEGIN FILE and END FILE. Returns a reader if
successful, or a null pointer otherwise. */
struct dfm_reader *
-dfm_open_reader (struct file_handle *fh)
+dfm_open_reader (struct file_handle *fh, struct lexer *lexer)
{
struct dfm_reader *r;
void **rp;
r = xmalloc (sizeof *r);
r->fh = fh;
- ds_init (&r->line, 64);
- ds_init (&r->scratch, 0);
+ r->lexer = lexer ;
+ ds_init_empty (&r->line);
+ ds_init_empty (&r->scratch);
r->flags = DFM_ADVANCE;
r->eof_cnt = 0;
if (fh != fh_inline_file ())
{
- r->where.filename = fh_get_filename (fh);
+ r->where.file_name = fh_get_file_name (fh);
r->where.line_number = 0;
- r->file = fn_open (fh_get_filename (fh), "rb");
+ r->file = fn_open (fh_get_file_name (fh), "rb");
if (r->file == NULL)
{
msg (ME, _("Could not open \"%s\" for reading as a data file: %s."),
- fh_get_filename (r->fh), strerror (errno));
+ fh_get_file_name (r->fh), strerror (errno));
fh_close (fh,"data file", "rs");
free (r);
return NULL;
{
r->flags |= DFM_SAW_BEGIN_DATA;
- while (token == '.')
- lex_get ();
- if (!lex_force_match_id ("BEGIN") || !lex_force_match_id ("DATA"))
+ while (lex_token (r->lexer) == '.')
+ lex_get (r->lexer);
+ if (!lex_force_match_id (r->lexer, "BEGIN") || !lex_force_match_id (r->lexer, "DATA"))
return false;
- getl_set_prompt_style (GETL_PROMPT_DATA);
+ prompt_set_style (PROMPT_DATA);
}
- if (!getl_read_line (NULL))
+ if (!lex_get_line_raw (r->lexer, NULL))
{
msg (SE, _("Unexpected end-of-file while reading data in BEGIN "
"DATA. This probably indicates "
return false;
}
- if (ds_length (&getl_buf) >= 8
- && !strncasecmp (ds_c_str (&getl_buf), "end data", 8))
+ if (ds_length (lex_entire_line_ds (r->lexer) ) >= 8
+ && !strncasecmp (lex_entire_line (r->lexer), "end data", 8))
{
- lex_set_prog (ds_c_str (&getl_buf) + ds_length (&getl_buf));
+ lex_discard_line (r->lexer);
return false;
}
- ds_assign_string (&r->line, &getl_buf);
+ ds_assign_string (&r->line, lex_entire_line_ds (r->lexer) );
+
return true;
}
read_file_record (struct dfm_reader *r)
{
assert (r->fh != fh_inline_file ());
+ ds_clear (&r->line);
if (fh_get_mode (r->fh) == FH_MODE_TEXT)
{
- ds_clear (&r->line);
- if (!ds_gets (&r->line, r->file))
+ if (!ds_read_line (&r->line, r->file))
{
if (ferror (r->file))
msg (ME, _("Error reading file %s: %s."),
fh_get_name (r->fh), strerror (errno));
return false;
}
+ ds_chomp (&r->line, '\n');
}
else if (fh_get_mode (r->fh) == FH_MODE_BINARY)
{
size_t record_width = fh_get_record_width (r->fh);
- size_t amt;
-
- if (ds_length (&r->line) < record_width)
- ds_rpad (&r->line, record_width, 0);
-
- amt = fread (ds_c_str (&r->line), 1, record_width, r->file);
+ size_t amt = ds_read_stream (&r->line, 1, record_width, r->file);
if (record_width != amt)
{
if (ferror (r->file))
}
}
else
- abort ();
+ NOT_REACHED ();
r->where.line_number++;
{
r->flags &= ~DFM_ADVANCE;
- if (r->eof_cnt == 0 && read_record (r))
+ if (r->eof_cnt == 0 && read_record (r) )
{
r->pos = 0;
return 0;
/* Returns the current record in the file corresponding to
HANDLE. Aborts if reading from the file is necessary or at
- end of file, so call dfm_eof() first. Sets *LINE to the line,
- which is not null-terminated. The caller must not free or
- modify the returned string. */
-void
-dfm_get_record (struct dfm_reader *r, struct fixed_string *line)
+ end of file, so call dfm_eof() first. */
+struct substring
+dfm_get_record (struct dfm_reader *r)
{
assert ((r->flags & DFM_ADVANCE) == 0);
assert (r->eof_cnt == 0);
- assert (r->pos <= ds_length (&r->line));
- line->string = ds_data (&r->line) + r->pos;
- line->length = ds_length (&r->line) - r->pos;
+ return ds_substr (&r->line, r->pos, SIZE_MAX);
}
/* Expands tabs in the current line into the equivalent number of
void
dfm_expand_tabs (struct dfm_reader *r)
{
- struct string temp;
size_t ofs, new_pos, tab_width;
assert ((r->flags & DFM_ADVANCE) == 0);
assert (r->eof_cnt == 0);
- assert (r->pos <= ds_length (&r->line));
if (r->flags & DFM_TABS_EXPANDED)
return;
if (r->fh != fh_inline_file ()
&& (fh_get_mode (r->fh) == FH_MODE_BINARY
|| fh_get_tab_width (r->fh) == 0
- || memchr (ds_c_str (&r->line), '\t', ds_length (&r->line)) == NULL))
+ || ds_find_char (&r->line, '\t') == SIZE_MAX))
return;
/* Expand tabs from r->line into r->scratch, and figure out
new value for r->pos. */
tab_width = fh_get_tab_width (r->fh);
ds_clear (&r->scratch);
- new_pos = 0;
+ new_pos = SIZE_MAX;
for (ofs = 0; ofs < ds_length (&r->line); ofs++)
{
unsigned char c;
if (ofs == r->pos)
new_pos = ds_length (&r->scratch);
- c = ds_c_str (&r->line)[ofs];
+ c = ds_data (&r->line)[ofs];
if (c != '\t')
- ds_putc (&r->scratch, c);
+ ds_put_char (&r->scratch, c);
else
{
do
- ds_putc (&r->scratch, ' ');
+ ds_put_char (&r->scratch, ' ');
while (ds_length (&r->scratch) % tab_width != 0);
}
}
+ if (new_pos == SIZE_MAX)
+ {
+ /* Maintain the same relationship between position and line
+ length that we had before. DATA LIST uses a
+ beyond-the-end position to deal with an empty field at
+ the end of the line. */
+ assert (r->pos >= ds_length (&r->line));
+ new_pos = (r->pos - ds_length (&r->line)) + ds_length (&r->scratch);
+ }
/* Swap r->line and r->scratch and set new r->pos. */
- temp = r->line;
- r->line = r->scratch;
- r->scratch = temp;
+ ds_swap (&r->line, &r->scratch);
r->pos = new_pos;
}
-/* Causes dfm_get_record() to read in the next record the next time it
- is executed on file HANDLE. */
+/* Causes dfm_get_record() or dfm_get_whole_record() to read in
+ the next record the next time it is executed on file
+ HANDLE. */
void
dfm_forward_record (struct dfm_reader *r)
{
dfm_reread_record (struct dfm_reader *r, size_t column)
{
r->flags &= ~DFM_ADVANCE;
- if (column < 1)
- r->pos = 0;
- else if (column > ds_length (&r->line))
- r->pos = ds_length (&r->line);
- else
- r->pos = column - 1;
+ r->pos = MAX (column, 1) - 1;
}
/* Sets the current line to begin COLUMNS characters following
is set. Unless dfm_reread_record() or dfm_forward_columns()
have been called, this is 1. */
size_t
-dfm_column_start (struct dfm_reader *r)
+dfm_column_start (const struct dfm_reader *r)
{
return r->pos + 1;
}
-/* Pushes the filename and line number on the fn/ln stack. */
+/* Returns the number of columns we are currently beyond the end
+ of the line. At or before end-of-line, this is 0; one column
+ after end-of-line, this is 1; and so on. */
+size_t
+dfm_columns_past_end (const struct dfm_reader *r)
+{
+ return r->pos < ds_length (&r->line) ? 0 : ds_length (&r->line) - r->pos;
+}
+
+/* Returns the 1-based column within the current line that P
+ designates. */
+size_t
+dfm_get_column (const struct dfm_reader *r, const char *p)
+{
+ return ds_pointer_to_position (&r->line, p) + 1;
+}
+
+/* Pushes the file name and line number on the fn/ln stack. */
void
dfm_push (struct dfm_reader *r)
{
if (r->fh != fh_inline_file ())
- err_push_file_locator (&r->where);
+ msg_push_msg_locator (&r->where);
}
-/* Pops the filename and line number from the fn/ln stack. */
+/* Pops the file name and line number from the fn/ln stack. */
void
dfm_pop (struct dfm_reader *r)
{
if (r->fh != fh_inline_file ())
- err_pop_file_locator (&r->where);
+ msg_pop_msg_locator (&r->where);
}
\f
/* BEGIN DATA...END DATA procedure. */
/* Perform BEGIN DATA...END DATA as a procedure in itself. */
int
-cmd_begin_data (void)
+cmd_begin_data (struct lexer *lexer, struct dataset *ds)
{
struct dfm_reader *r;
bool ok;
}
/* Open inline file. */
- r = dfm_open_reader (fh_inline_file ());
+ r = dfm_open_reader (fh_inline_file (), lexer);
r->flags |= DFM_SAW_BEGIN_DATA;
/* Input procedure reads from inline file. */
- getl_set_prompt_style (GETL_PROMPT_DATA);
- ok = procedure (NULL, NULL);
+ prompt_set_style (PROMPT_DATA);
+ ok = procedure (ds, NULL, NULL);
dfm_close_reader (r);
projects / pspp-builds.git / blobdiff