/* PSPP - a program for statistical analysis.
- Copyright (C) 2007, 2008, 2009 Free Software Foundation, Inc.
+ Copyright (C) 2007, 2008, 2009, 2010 Free Software Foundation, Inc.
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
size_t size, size_t count);
static void read_machine_float_info (struct sfm_reader *,
size_t size, size_t count);
+static void read_mrsets (struct sfm_reader *, size_t size, size_t count);
static void read_display_parameters (struct sfm_reader *,
size_t size, size_t count);
static void read_long_var_name_map (struct sfm_reader *r,
size_t size, size_t count);
static void read_variable_attributes (struct sfm_reader *r,
size_t size, size_t count);
+static void read_ncases64 (struct sfm_reader *, size_t size, size_t count);
static void read_character_encoding (struct sfm_reader *r,
size_t size, size_t count);
static void read_long_string_value_labels (struct sfm_reader *r,
size_t size, size_t count);
+static void read_unknown_extension (struct sfm_reader *,
+ size_t size, size_t count);
static void read_compressed_data (struct sfm_reader *);
static struct text_record *open_text_record (
char **key, char **value);
static char *text_tokenize (struct text_record *, int delimiter);
static bool text_match (struct text_record *text, int c);
+static const char *text_parse_counted_string (struct text_record *);
+static size_t text_pos (const struct text_record *);
static void usage (int exit_code);
static void sys_warn (struct sfm_reader *, const char *, ...)
static void read_bytes (struct sfm_reader *, void *, size_t);
static int read_int (struct sfm_reader *);
+static int64_t read_int64 (struct sfm_reader *);
static double read_float (struct sfm_reader *);
static void read_string (struct sfm_reader *, char *, size_t);
static void skip_bytes (struct sfm_reader *, size_t);
break;
case 7:
- /* Unknown purpose. */
- break;
+ case 19:
+ read_mrsets (r, size, count);
+ return;
case 11:
read_display_parameters (r, size, count);
return;
case 16:
- /* New in SPSS v14? Unknown purpose. */
- break;
+ read_ncases64 (r, size, count);
+ return;
case 17:
read_datafile_attributes (r, size, count);
default:
sys_warn (r, _("Unrecognized record type 7, subtype %d."), subtype);
- break;
+ read_unknown_extension (r, size, count);
+ return;
}
skip_bytes (r, bytes);
lowest, "LOWEST");
}
+/* Read record type 7, subtype 7. */
+static void
+read_mrsets (struct sfm_reader *r, size_t size, size_t count)
+{
+ struct text_record *text;
+
+ printf ("%08lx: multiple response sets\n", ftell (r->file));
+ text = open_text_record (r, size * count);
+ for (;;)
+ {
+ const char *name;
+ enum { MRSET_MC, MRSET_MD } type;
+ bool cat_label_from_counted_values = false;
+ bool label_from_var_label = false;
+ const char *counted;
+ const char *label;
+ const char *variables;
+
+ name = text_tokenize (text, '=');
+ if (name == NULL)
+ break;
+
+ if (text_match (text, 'C'))
+ {
+ type = MRSET_MC;
+ counted = NULL;
+ if (!text_match (text, ' '))
+ {
+ sys_warn (r, "missing space following 'C' at offset %zu "
+ "in mrsets record", text_pos (text));
+ break;
+ }
+ }
+ else if (text_match (text, 'D'))
+ {
+ type = MRSET_MD;
+ }
+ else if (text_match (text, 'E'))
+ {
+ char *number;
+
+ type = MRSET_MD;
+ cat_label_from_counted_values = true;
+
+ if (!text_match (text, ' '))
+ {
+ sys_warn (r, _("Missing space following 'E' at offset %zu "
+ "in MRSETS record"), text_pos (text));
+ break;
+ }
+
+ number = text_tokenize (text, ' ');
+ if (!strcmp (number, "11"))
+ label_from_var_label = true;
+ else if (strcmp (number, "1"))
+ sys_warn (r, _("Unexpected label source value \"%s\" "
+ "following 'E' at offset %zu in MRSETS record"),
+ number, text_pos (text));
+
+ }
+ else
+ {
+ sys_warn (r, "missing 'C', 'D', or 'E' at offset %zu "
+ "in mrsets record", text_pos (text));
+ break;
+ }
+
+ if (type == MRSET_MD)
+ {
+ counted = text_parse_counted_string (text);
+ if (counted == NULL)
+ break;
+ }
+
+ label = text_parse_counted_string (text);
+ if (label == NULL)
+ break;
+
+ variables = text_tokenize (text, '\n');
+ if (variables == NULL)
+ {
+ sys_warn (r, "missing variable names following label "
+ "at offset %zu in mrsets record", text_pos (text));
+ break;
+ }
+
+ printf ("\t\"%s\": multiple %s set",
+ name, type == MRSET_MC ? "category" : "dichotomy");
+ if (counted != NULL)
+ printf (", counted value \"%s\"", counted);
+ if (cat_label_from_counted_values)
+ printf (", category labels from counted values");
+ if (label[0] != '\0')
+ printf (", label \"%s\"", label);
+ if (label_from_var_label)
+ printf (", label from variable label");
+ printf(", variables \"%s\"\n", variables);
+ }
+ close_text_record (text);
+}
+
/* Read record type 7, subtype 11. */
static void
read_display_parameters (struct sfm_reader *r, size_t size, size_t count)
}
}
+/* Read extended number of cases record. */
+static void
+read_ncases64 (struct sfm_reader *r, size_t size, size_t count)
+{
+ int64_t unknown, ncases64;
+
+ if (size != 8)
+ {
+ sys_warn (r, _("Bad size %zu for extended number of cases."), size);
+ skip_bytes (r, size * count);
+ return;
+ }
+ if (count != 2)
+ {
+ sys_warn (r, _("Bad count %zu for extended number of cases."), size);
+ skip_bytes (r, size * count);
+ return;
+ }
+ unknown = read_int64 (r);
+ ncases64 = read_int64 (r);
+ printf ("%08lx: extended number of cases: "
+ "unknown=%"PRId64", ncases64=%"PRId64"\n",
+ ftell (r->file), unknown, ncases64);
+}
+
static void
read_datafile_attributes (struct sfm_reader *r, size_t size, size_t count)
{
}
}
+static void
+hex_dump (size_t offset, const void *buffer_, size_t buffer_size)
+{
+ const uint8_t *buffer = buffer_;
+
+ while (buffer_size > 0)
+ {
+ size_t n = MIN (buffer_size, 16);
+ size_t i;
+
+ printf ("%04zx", offset);
+ for (i = 0; i < 16; i++)
+ {
+ if (i < n)
+ printf ("%c%02x", i == 8 ? '-' : ' ', buffer[i]);
+ else
+ printf (" ");
+ }
+
+ printf (" |");
+ for (i = 0; i < 16; i++)
+ {
+ unsigned char c = i < n ? buffer[i] : ' ';
+ putchar (isprint (c) ? c : '.');
+ }
+ printf ("|\n");
+
+ offset += n;
+ buffer += n;
+ buffer_size -= n;
+ }
+}
+
+/* Reads and prints any type 7 record that we don't understand. */
+static void
+read_unknown_extension (struct sfm_reader *r, size_t size, size_t count)
+{
+ unsigned char *buffer;
+ size_t i;
+
+ if (size == 0 || count > 65536 / size)
+ skip_bytes (r, size * count);
+ else if (size != 1)
+ {
+ buffer = xmalloc (size);
+ for (i = 0; i < count; i++)
+ {
+ read_bytes (r, buffer, size);
+ hex_dump (i * size, buffer, size);
+ }
+ free (buffer);
+ }
+ else
+ {
+ buffer = xmalloc (count);
+ read_bytes (r, buffer, count);
+ if (memchr (buffer, 0, count) == 0)
+ for (i = 0; i < count; i++)
+ {
+ unsigned char c = buffer[i];
+
+ if (c == '\\')
+ printf ("\\\\");
+ else if (c == '\n' || isprint (c))
+ putchar (c);
+ else
+ printf ("\\%02x", c);
+ }
+ else
+ hex_dump (0, buffer, count);
+ free (buffer);
+ }
+}
+
static void
read_variable_attributes (struct sfm_reader *r, size_t size, size_t count)
{
/* State. */
struct text_record
{
+ struct sfm_reader *reader; /* Reader. */
char *buffer; /* Record contents. */
size_t size; /* Size of buffer. */
size_t pos; /* Current position in buffer. */
struct text_record *text = xmalloc (sizeof *text);
char *buffer = xmalloc (size + 1);
read_bytes (r, buffer, size);
+ buffer[size] = '\0';
+ text->reader = r;
text->buffer = buffer;
text->size = size;
text->pos = 0;
return false;
}
+/* Reads a integer value expressed in decimal, then a space, then a string that
+ consists of exactly as many bytes as specified by the integer, then a space,
+ from TEXT. Returns the string, null-terminated, as a subset of TEXT's
+ buffer (so the caller should not free the string). */
+static const char *
+text_parse_counted_string (struct text_record *text)
+{
+ size_t start;
+ size_t n;
+ char *s;
+
+ start = text->pos;
+ n = 0;
+ while (isdigit ((unsigned char) text->buffer[text->pos]))
+ n = (n * 10) + (text->buffer[text->pos++] - '0');
+ if (start == text->pos)
+ {
+ sys_error (text->reader, "expecting digit at offset %zu in record",
+ text->pos);
+ return NULL;
+ }
+
+ if (!text_match (text, ' '))
+ {
+ sys_error (text->reader, "expecting space at offset %zu in record",
+ text->pos);
+ return NULL;
+ }
+
+ if (text->pos + n > text->size)
+ {
+ sys_error (text->reader, "%zu-byte string starting at offset %zu "
+ "exceeds record length %zu", n, text->pos, text->size);
+ return NULL;
+ }
+
+ s = &text->buffer[text->pos];
+ if (s[n] != ' ')
+ {
+ sys_error (text->reader, "expecting space at offset %zu following "
+ "%zu-byte string", text->pos + n, n);
+ return NULL;
+ }
+ s[n] = '\0';
+ text->pos += n + 1;
+ return s;
+}
+
/* Reads a variable=value pair from TEXT.
Looks up the variable in DICT and stores it into *VAR.
Stores a null-terminated value into *VALUE. */
text->pos++;
return true;
}
+
+/* Returns the current byte offset inside the TEXT's string. */
+static size_t
+text_pos (const struct text_record *text)
+{
+ return text->pos;
+}
\f
static void
usage (int exit_code)
return integer_get (r->integer_format, integer, sizeof integer);
}
+/* Reads a 64-bit signed integer from R and returns its value in
+ host format. */
+static int64_t
+read_int64 (struct sfm_reader *r)
+{
+ uint8_t integer[8];
+ read_bytes (r, integer, sizeof integer);
+ return integer_get (r->integer_format, integer, sizeof integer);
+}
+
/* Reads a 64-bit floating-point number from R and returns its
value in host format. */
static double
projects / pspp-builds.git / blobdiff