1 @node System File Format
2 @appendix System File Format
4 A system file encapsulates a set of cases and dictionary information
5 that describes how they may be interpreted. This chapter describes
6 the format of a system file.
8 System files use four data types: 8-bit characters, 32-bit integers,
10 and 64-bit floating points, called here @code{char}, @code{int32},
12 @code{flt64}, respectively. Data is not necessarily aligned on a word
13 or double-word boundary: the long variable name record (@pxref{Long
14 Variable Names Record}) and very long string records (@pxref{Very Long
15 String Record}) have arbitrary byte length and can therefore cause all
16 data coming after them in the file to be misaligned.
18 Integer data in system files may be big-endian or little-endian. A
19 reader may detect the endianness of a system file by examining
20 @code{layout_code} in the file header record
21 (@pxref{layout_code,,@code{layout_code}}).
23 Floating-point data in system files may nominally be in IEEE 754, IBM,
24 or VAX formats. A reader may detect the floating-point format in use
25 by examining @code{bias} in the file header record
26 (@pxref{bias,,@code{bias}}).
28 PSPP detects big-endian and little-endian integer formats in system
29 files and translates as necessary. PSPP also detects the
30 floating-point format in use, as well as the endianness of IEEE 754
31 floating-point numbers, and translates as needed. However, only IEEE
32 754 numbers with the same endianness as integer data in the same file
33 has actually been observed in system files, and it is likely that
34 other formats are obsolete or were never used.
36 System files use a few floating point values for special purposes:
40 The system-missing value is represented by the largest possible
41 negative number in the floating point format (@code{-DBL_MAX}).
44 HIGHEST is used as the high end of a missing value range with an
45 unbounded maximum. It is represented by the largest possible positive
46 number (@code{DBL_MAX}).
49 LOWEST is used as the low end of a missing value range with an
50 unbounded minimum. It was originally represented by the
51 second-largest negative number (in IEEE 754 format,
52 @code{0xffeffffffffffffe}). System files written by SPSS 21 and later
53 instead use the largest negative number (@code{-DBL_MAX}), the same
54 value as SYSMIS. This does not lead to ambiguity because LOWEST
55 appears in system files only in missing value ranges, which never
59 System files are divided into records, each of which begins with a
60 4-byte record type, usually regarded as an @code{int32}.
62 The records must appear in the following order:
72 All pairs of value labels records and value label variables records,
76 Document record, if present.
79 Extension (type 7) records, in ascending numerical order of their
83 Dictionary termination record.
89 Each type of record is described separately below.
92 * File Header Record::
94 * Value Labels Records::
96 * Machine Integer Info Record::
97 * Machine Floating-Point Info Record::
98 * Multiple Response Sets Records::
99 * Extra Product Info Record::
100 * Variable Display Parameter Record::
101 * Long Variable Names Record::
102 * Very Long String Record::
103 * Character Encoding Record::
104 * Long String Value Labels Record::
105 * Data File and Variable Attributes Records::
106 * Extended Number of Cases Record::
107 * Miscellaneous Informational Records::
108 * Dictionary Termination Record::
112 @node File Header Record
113 @section File Header Record
115 The file header is always the first record in the file. It has the
122 int32 nominal_case_size;
127 char creation_date[9];
128 char creation_time[8];
134 @item char rec_type[4];
135 Record type code, set to @samp{$FL2}, that is, either @code{24 46 4c
136 32} if the file uses an ASCII-based character encoding, or @code{5b c6
137 d3 f2} if the file uses an EBCDIC-based character encoding.
139 @item char prod_name[60];
140 Product identification string. This always begins with the characters
141 @samp{@@(#) SPSS DATA FILE}. PSPP uses the remaining characters to
142 give its version and the operating system name; for example, @samp{GNU
143 pspp 0.1.4 - sparc-sun-solaris2.5.2}. The string is truncated if it
144 would be longer than 60 characters; otherwise it is padded on the right
148 @item int32 layout_code;
149 Normally set to 2, although a few system files have been spotted in
150 the wild with a value of 3 here. PSPP use this value to determine the
151 file's integer endianness (@pxref{System File Format}).
153 @item int32 nominal_case_size;
154 Number of data elements per case. This is the number of variables,
155 except that long string variables add extra data elements (one for every
156 8 characters after the first 8). However, string variables do not
157 contribute to this value beyond the first 255 bytes. Further, system
158 files written by some systems set this value to -1. In general, it is
159 unsafe for systems reading system files to rely upon this value.
161 @item int32 compressed;
162 Set to 1 if the data in the file is compressed, 0 otherwise.
164 @item int32 weight_index;
165 If one of the variables in the data set is used as a weighting
166 variable, set to the dictionary index of that variable, plus 1
167 (@pxref{Dictionary Index}). Otherwise, set to 0.
170 Set to the number of cases in the file if it is known, or -1 otherwise.
172 In the general case it is not possible to determine the number of cases
173 that will be output to a system file at the time that the header is
174 written. The way that this is dealt with is by writing the entire
175 system file, including the header, then seeking back to the beginning of
176 the file and writing just the @code{ncases} field. For files in which
177 this is not valid, the seek operation fails. In this case,
178 @code{ncases} remains -1.
182 Compression bias, ordinarily set to 100. Only integers between
183 @code{1 - bias} and @code{251 - bias} can be compressed.
185 By assuming that its value is 100, PSPP uses @code{bias} to determine
186 the file's floating-point format and endianness (@pxref{System File
187 Format}). If the compression bias is not 100, PSPP cannot auto-detect
188 the floating-point format and assumes that it is IEEE 754 format with
189 the same endianness as the system file's integers, which is correct
190 for all known system files.
192 @item char creation_date[9];
193 Date of creation of the system file, in @samp{dd mmm yy}
194 format, with the month as standard English abbreviations, using an
195 initial capital letter and following with lowercase. If the date is not
196 available then this field is arbitrarily set to @samp{01 Jan 70}.
198 @item char creation_time[8];
199 Time of creation of the system file, in @samp{hh:mm:ss}
200 format and using 24-hour time. If the time is not available then this
201 field is arbitrarily set to @samp{00:00:00}.
203 @item char file_label[64];
204 File label declared by the user, if any (@pxref{FILE LABEL,,,pspp,
205 PSPP Users Guide}). Padded on the right with spaces.
207 A product that identifies itself as @code{VOXCO INTERVIEWER 4.3} uses
208 CR-only line ends in this field, rather than the more usual LF-only or
211 @item char padding[3];
212 Ignored padding bytes to make the structure a multiple of 32 bits in
213 length. Set to zeros.
216 @node Variable Record
217 @section Variable Record
219 There must be one variable record for each numeric variable and each
220 string variable with width 8 bytes or less. String variables wider
221 than 8 bytes have one variable record for each 8 bytes, rounding up.
222 The first variable record for a long string specifies the variable's
223 correct dictionary information. Subsequent variable records for a
224 long string are filled with dummy information: a type of -1, no
225 variable label or missing values, print and write formats that are
226 ignored, and an empty string as name. A few system files have been
227 encountered that include a variable label on dummy variable records,
228 so readers should take care to parse dummy variable records in the
229 same way as other variable records.
231 @anchor{Dictionary Index}
232 The @dfn{dictionary index} of a variable is its offset in the set of
233 variable records, including dummy variable records for long string
234 variables. The first variable record has a dictionary index of 0, the
235 second has a dictionary index of 1, and so on.
237 The system file format does not directly support string variables
238 wider than 255 bytes. Such very long string variables are represented
239 by a number of narrower string variables. @xref{Very Long String
240 Record}, for details.
246 int32 n_missing_values;
251 /* @r{Present only if @code{has_var_label} is 1.} */
255 /* @r{Present only if @code{n_missing_values} is nonzero}. */
256 flt64 missing_values[];
260 @item int32 rec_type;
261 Record type code. Always set to 2.
264 Variable type code. Set to 0 for a numeric variable. For a short
265 string variable or the first part of a long string variable, this is set
266 to the width of the string. For the second and subsequent parts of a
267 long string variable, set to -1, and the remaining fields in the
268 structure are ignored.
270 @item int32 has_var_label;
271 If this variable has a variable label, set to 1; otherwise, set to 0.
273 @item int32 n_missing_values;
274 If the variable has no missing values, set to 0. If the variable has
275 one, two, or three discrete missing values, set to 1, 2, or 3,
276 respectively. If the variable has a range for missing variables, set to
277 -2; if the variable has a range for missing variables plus a single
278 discrete value, set to -3.
281 Print format for this variable. See below.
284 Write format for this variable. See below.
287 Variable name. The variable name must begin with a capital letter or
288 the at-sign (@samp{@@}). Subsequent characters may also be digits, octothorpes
289 (@samp{#}), dollar signs (@samp{$}), underscores (@samp{_}), or full
290 stops (@samp{.}). The variable name is padded on the right with spaces.
292 @item int32 label_len;
293 This field is present only if @code{has_var_label} is set to 1. It is
294 set to the length, in characters, of the variable label. The
295 documented maximum length varies from 120 to 255 based on SPSS
296 version, but some files have been seen with longer labels. PSPP
297 accepts longer labels and truncates them to 255 bytes on input.
300 This field is present only if @code{has_var_label} is set to 1. It has
301 length @code{label_len}, rounded up to the nearest multiple of 32 bits.
302 The first @code{label_len} characters are the variable's variable label.
304 @item flt64 missing_values[];
305 This field is present only if @code{n_missing_values} is nonzero. It
306 has the same number of 8-byte elements as the absolute value of
307 @code{n_missing_values}. Each element is interpreted as a number for
308 numeric variables (with HIGHEST and LOWEST indicated as described in
309 the chapter introduction). For string variables of width less than 8
310 bytes, elements are right-padded with spaces; for string variables
311 wider than 8 bytes, only the first 8 bytes of each missing value are
312 specified, with the remainder implicitly all spaces.
314 For discrete missing values, each element represents one missing
315 value. When a range is present, the first element denotes the minimum
316 value in the range, and the second element denotes the maximum value
317 in the range. When a range plus a value are present, the third
318 element denotes the additional discrete missing value.
321 The @code{print} and @code{write} members of sysfile_variable are output
322 formats coded into @code{int32} types. The least-significant byte
323 of the @code{int32} represents the number of decimal places, and the
324 next two bytes in order of increasing significance represent field width
325 and format type, respectively. The most-significant byte is not
326 used and should be set to zero.
328 Format types are defined as follows:
331 @multitable {Value} {@code{DATETIME}}
417 A few system files have been observed in the wild with invalid
418 @code{write} fields, in particular with value 0. Readers should
419 probably treat invalid @code{print} or @code{write} fields as some
422 @node Value Labels Records
423 @section Value Labels Records
425 The value label records documented in this section are used for
426 numeric and short string variables only. Long string variables may
427 have value labels, but their value labels are recorded using a
428 different record type (@pxref{Long String Value Labels Record}).
430 The value label record has the following format:
436 /* @r{Repeated @code{label_cnt} times}. */
443 @item int32 rec_type;
444 Record type. Always set to 3.
446 @item int32 label_count;
447 Number of value labels present in this record.
450 The remaining fields are repeated @code{count} times. Each
451 repetition specifies one value label.
455 A numeric value or a short string value padded as necessary to 8 bytes
456 in length. Its type and width cannot be determined until the
457 following value label variables record (see below) is read.
459 @item char label_len;
460 The label's length, in bytes. The documented maximum length varies
461 from 60 to 120 based on SPSS version. PSPP supports value labels up
465 @code{label_len} bytes of the actual label, followed by up to 7 bytes
466 of padding to bring @code{label} and @code{label_len} together to a
467 multiple of 8 bytes in length.
470 The value label record is always immediately followed by a value label
471 variables record with the following format:
480 @item int32 rec_type;
481 Record type. Always set to 4.
483 @item int32 var_count;
484 Number of variables that the associated value labels from the value
485 label record are to be applied.
488 A list of dictionary indexes of variables to which to apply the value
489 labels (@pxref{Dictionary Index}). There are @code{var_count}
492 String variables wider than 8 bytes may not be specified in this list.
495 @node Document Record
496 @section Document Record
498 The document record, if present, has the following format:
507 @item int32 rec_type;
508 Record type. Always set to 6.
511 Number of lines of documents present.
513 @item char lines[][80];
514 Document lines. The number of elements is defined by @code{n_lines}.
515 Lines shorter than 80 characters are padded on the right with spaces.
518 @node Machine Integer Info Record
519 @section Machine Integer Info Record
521 The integer info record, if present, has the following format:
533 int32 version_revision;
535 int32 floating_point_rep;
536 int32 compression_code;
538 int32 character_code;
542 @item int32 rec_type;
543 Record type. Always set to 7.
546 Record subtype. Always set to 3.
549 Size of each piece of data in the data part, in bytes. Always set to 4.
552 Number of pieces of data in the data part. Always set to 8.
554 @item int32 version_major;
555 PSPP major version number. In version @var{x}.@var{y}.@var{z}, this
558 @item int32 version_minor;
559 PSPP minor version number. In version @var{x}.@var{y}.@var{z}, this
562 @item int32 version_revision;
563 PSPP version revision number. In version @var{x}.@var{y}.@var{z},
566 @item int32 machine_code;
567 Machine code. PSPP always set this field to value to -1, but other
570 @item int32 floating_point_rep;
571 Floating point representation code. For IEEE 754 systems this is 1.
572 IBM 370 sets this to 2, and DEC VAX E to 3.
574 @item int32 compression_code;
575 Compression code. Always set to 1.
577 @item int32 endianness;
578 Machine endianness. 1 indicates big-endian, 2 indicates little-endian.
580 @item int32 character_code;
581 @anchor{character-code} Character code. The following values have
582 been actually observed in system files:
592 The @code{windows-1250} code page for Central European and Eastern
596 The @code{windows-1252} code page for Western European languages.
605 The following additional values are known to be defined:
615 Other Windows code page numbers are known to be generally valid.
617 Old versions of SPSS for Unix and Windows always wrote value 2 in this
618 field, regardless of the encoding in use. Newer versions also write
619 the character encoding as a string (see @ref{Character Encoding
623 @node Machine Floating-Point Info Record
624 @section Machine Floating-Point Info Record
626 The floating-point info record, if present, has the following format:
642 @item int32 rec_type;
643 Record type. Always set to 7.
646 Record subtype. Always set to 4.
649 Size of each piece of data in the data part, in bytes. Always set to 8.
652 Number of pieces of data in the data part. Always set to 3.
655 The system missing value.
658 The value used for HIGHEST in missing values.
661 The value used for LOWEST in missing values.
664 @node Multiple Response Sets Records
665 @section Multiple Response Sets Records
667 The system file format has two different types of records that
668 represent multiple response sets (@pxref{MRSETS,,,pspp, PSPP Users
669 Guide}). The first type of record describes multiple response sets
670 that can be understood by SPSS before version 14. The second type of
671 record, with a closely related format, is used for multiple dichotomy
672 sets that use the CATEGORYLABELS=COUNTEDVALUES feature added in
682 /* @r{Exactly @code{count} bytes of data.} */
687 @item int32 rec_type;
688 Record type. Always set to 7.
691 Record subtype. Set to 7 for records that describe multiple response
692 sets understood by SPSS before version 14, or to 19 for records that
693 describe dichotomy sets that use the CATEGORYLABELS=COUNTEDVALUES
694 feature added in version 14.
697 The size of each element in the @code{mrsets} member. Always set to 1.
700 The total number of bytes in @code{mrsets}.
703 A series of multiple response sets, each of which consists of the
708 The set's name (an identifier that begins with @samp{$}), in mixed
709 upper and lower case.
712 An equals sign (@samp{=}).
715 @samp{C} for a multiple category set, @samp{D} for a multiple
716 dichotomy set with CATEGORYLABELS=VARLABELS, or @samp{E} for a
717 multiple dichotomy set with CATEGORYLABELS=COUNTEDVALUES.
720 For a multiple dichotomy set with CATEGORYLABELS=COUNTEDVALUES, a
721 space, followed by a number expressed as decimal digits, followed by a
722 space. If LABELSOURCE=VARLABEL was specified on MRSETS, then the
723 number is 11; otherwise it is 1.@footnote{This part of the format may
724 not be fully understood, because only a single example of each
725 possibility has been examined.}
728 For either kind of multiple dichotomy set, the counted value, as a
729 positive integer count specified as decimal digits, followed by a
730 space, followed by as many string bytes as specified in the count. If
731 the set contains numeric variables, the string consists of the counted
732 integer value expressed as decimal digits. If the set contains string
733 variables, the string contains the counted string value. Either way,
734 the string may be padded on the right with spaces (older versions of
735 SPSS seem to always pad to a width of 8 bytes; newer versions don't).
741 The multiple response set's label, using the same format as for the
742 counted value for multiple dichotomy sets. A string of length 0 means
743 that the set does not have a label. A string of length 0 is also
744 written if LABELSOURCE=VARLABEL was specified.
750 The short names of the variables in the set, converted to lowercase,
751 each separated from the previous by a single space.
754 A line feed (byte 0x0a).
758 Example: Given appropriate variable definitions, consider the
759 following MRSETS command:
762 MRSETS /MCGROUP NAME=$a LABEL='my mcgroup' VARIABLES=a b c
763 /MDGROUP NAME=$b VARIABLES=g e f d VALUE=55
764 /MDGROUP NAME=$c LABEL='mdgroup #2' VARIABLES=h i j VALUE='Yes'
765 /MDGROUP NAME=$d LABEL='third mdgroup' CATEGORYLABELS=COUNTEDVALUES
766 VARIABLES=k l m VALUE=34
767 /MDGROUP NAME=$e CATEGORYLABELS=COUNTEDVALUES LABELSOURCE=VARLABEL
768 VARIABLES=n o p VALUE='choice'.
771 The above would generate the following multiple response set record of
775 $a=C 10 my mcgroup a b c
777 $c=D3 Yes 10 mdgroup #2 h i j
780 It would also generate the following multiple response set record with
784 $d=E 1 2 34 13 third mdgroup k l m
785 $e=E 11 6 choice 0 n o p
788 @node Extra Product Info Record
789 @section Extra Product Info Record
791 This optional record appears to contain a text string that describes
792 the program that wrote the file and the source of the data. (This is
793 redundant with the file label and product info found in the file
803 /* @r{Exactly @code{count} bytes of data.} */
808 @item int32 rec_type;
809 Record type. Always set to 7.
812 Record subtype. Always set to 10.
815 The size of each element in the @code{info} member. Always set to 1.
818 The total number of bytes in @code{info}.
821 A text string. A product that identifies itself as @code{VOXCO
822 INTERVIEWER 4.3} uses CR-only line ends in this field, rather than the
823 more usual LF-only or CR LF line ends.
826 @node Variable Display Parameter Record
827 @section Variable Display Parameter Record
829 The variable display parameter record, if present, has the following
839 /* @r{Repeated @code{count} times}. */
841 int32 width; /* @r{Not always present.} */
846 @item int32 rec_type;
847 Record type. Always set to 7.
850 Record subtype. Always set to 11.
853 The size of @code{int32}. Always set to 4.
856 The number of sets of variable display parameters (ordinarily the
857 number of variables in the dictionary), times 2 or 3.
860 The remaining members are repeated @code{count} times, in the same
861 order as the variable records. No element corresponds to variable
862 records that continue long string variables. The meanings of these
863 members are as follows:
867 The measurement type of the variable:
877 SPSS sometimes writes a @code{measure} of 0. PSPP interprets this as
881 The width of the display column for the variable in characters.
883 This field is present if @var{count} is 3 times the number of
884 variables in the dictionary. It is omitted if @var{count} is 2 times
885 the number of variables.
887 @item int32 alignment;
888 The alignment of the variable for display purposes:
900 @node Long Variable Names Record
901 @section Long Variable Names Record
903 If present, the long variable names record has the following format:
912 /* @r{Exactly @code{count} bytes of data.} */
913 char var_name_pairs[];
917 @item int32 rec_type;
918 Record type. Always set to 7.
921 Record subtype. Always set to 13.
924 The size of each element in the @code{var_name_pairs} member. Always set to 1.
927 The total number of bytes in @code{var_name_pairs}.
929 @item char var_name_pairs[];
930 A list of @var{key}--@var{value} tuples, where @var{key} is the name
931 of a variable, and @var{value} is its long variable name.
932 The @var{key} field is at most 8 bytes long and must match the
933 name of a variable which appears in the variable record (@pxref{Variable
935 The @var{value} field is at most 64 bytes long.
936 The @var{key} and @var{value} fields are separated by a @samp{=} byte.
937 Each tuple is separated by a byte whose value is 09. There is no
938 trailing separator following the last tuple.
939 The total length is @code{count} bytes.
942 @node Very Long String Record
943 @section Very Long String Record
945 Old versions of SPSS limited string variables to a width of 255 bytes.
946 For backward compatibility with these older versions, the system file
947 format represents a string longer than 255 bytes, called a @dfn{very
948 long string}, as a collection of strings no longer than 255 bytes
949 each. The strings concatenated to make a very long string are called
950 its @dfn{segments}; for consistency, variables other than very long
951 strings are considered to have a single segment.
953 A very long string with a width of @var{w} has @var{n} =
954 (@var{w} + 251) / 252 segments, that is, one segment for every
955 252 bytes of width, rounding up. It would be logical, then, for each
956 of the segments except the last to have a width of 252 and the last
957 segment to have the remainder, but this is not the case. In fact,
958 each segment except the last has a width of 255 bytes. The last
959 segment has width @var{w} - (@var{n} - 1) * 252; some versions
960 of SPSS make it slightly wider, but not wide enough to make the last
961 segment require another 8 bytes of data.
963 Data is packed tightly into segments of a very long string, 255 bytes
964 per segment. Because 255 bytes of segment data are allocated for
965 every 252 bytes of the very long string's width (approximately), some
966 unused space is left over at the end of the allocated segments. Data
967 in unused space is ignored.
969 Example: Consider a very long string of width 20,000. Such a very
970 long string has 20,000 / 252 = 80 (rounding up) segments. The first
971 79 segments have width 255; the last segment has width 20,000 - 79 *
972 252 = 92 or slightly wider (up to 96 bytes, the next multiple of 8).
973 The very long string's data is actually stored in the 19,890 bytes in
974 the first 78 segments, plus the first 110 bytes of the 79th segment
975 (19,890 + 110 = 20,000). The remaining 145 bytes of the 79th segment
976 and all 92 bytes of the 80th segment are unused.
978 The very long string record explains how to stitch together segments
979 to obtain very long string data. For each of the very long string
980 variables in the dictionary, it specifies the name of its first
981 segment's variable and the very long string variable's actual width.
982 The remaining segments immediately follow the named variable in the
983 system file's dictionary.
985 The very long string record, which is present only if the system file
986 contains very long string variables, has the following format:
995 /* @r{Exactly @code{count} bytes of data.} */
996 char string_lengths[];
1000 @item int32 rec_type;
1001 Record type. Always set to 7.
1003 @item int32 subtype;
1004 Record subtype. Always set to 14.
1007 The size of each element in the @code{string_lengths} member. Always set to 1.
1010 The total number of bytes in @code{string_lengths}.
1012 @item char string_lengths[];
1013 A list of @var{key}--@var{value} tuples, where @var{key} is the name
1014 of a variable, and @var{value} is its length.
1015 The @var{key} field is at most 8 bytes long and must match the
1016 name of a variable which appears in the variable record (@pxref{Variable
1018 The @var{value} field is exactly 5 bytes long. It is a zero-padded,
1019 ASCII-encoded string that is the length of the variable.
1020 The @var{key} and @var{value} fields are separated by a @samp{=} byte.
1021 Tuples are delimited by a two-byte sequence @{00, 09@}.
1022 After the last tuple, there may be a single byte 00, or @{00, 09@}.
1023 The total length is @code{count} bytes.
1026 @node Character Encoding Record
1027 @section Character Encoding Record
1029 This record, if present, indicates the character encoding for string data,
1030 long variable names, variable labels, value labels and other strings in the
1040 /* @r{Exactly @code{count} bytes of data.} */
1045 @item int32 rec_type;
1046 Record type. Always set to 7.
1048 @item int32 subtype;
1049 Record subtype. Always set to 20.
1052 The size of each element in the @code{encoding} member. Always set to 1.
1055 The total number of bytes in @code{encoding}.
1057 @item char encoding[];
1058 The name of the character encoding. Normally this will be an official
1059 IANA character set name or alias.
1060 See @url{http://www.iana.org/assignments/character-sets}.
1061 Character set names are not case-sensitive, but SPSS appears to write
1062 them in all-uppercase.
1065 This record is not present in files generated by older software. See
1066 also the @code{character_code} field in the machine integer info
1067 record (@pxref{character-code}).
1069 When the character encoding record and the machine integer info record
1070 are both present, all system files observed in practice indicate the
1071 same character encoding, e.g.@: 1252 as @code{character_code} and
1072 @code{windows-1252} as @code{encoding}, 65001 and @code{UTF-8}, etc.
1074 If, for testing purposes, a file is crafted with different
1075 @code{character_code} and @code{encoding}, it seems that
1076 @code{character_code} controls the encoding for all strings in the
1077 system file before the dictionary termination record, including
1078 strings in data (e.g.@: string missing values), and @code{encoding}
1079 controls the encoding for strings following the dictionary termination
1082 @node Long String Value Labels Record
1083 @section Long String Value Labels Record
1085 This record, if present, specifies value labels for long string
1095 /* @r{Repeated up to exactly @code{count} bytes.} */
1100 long_string_label labels[];
1104 @item int32 rec_type;
1105 Record type. Always set to 7.
1107 @item int32 subtype;
1108 Record subtype. Always set to 21.
1114 The number of bytes following the header until the next header.
1116 @item int32 var_name_len;
1117 @itemx char var_name[];
1118 The number of bytes in the name of the variable that has long string
1119 value labels, plus the variable name itself, which consists of exactly
1120 @code{var_name_len} bytes. The variable name is not padded to any
1121 particular boundary, nor is it null-terminated.
1123 @item int32 var_width;
1124 The width of the variable, in bytes, which will be between 9 and
1127 @item int32 n_labels;
1128 @itemx long_string_label labels[];
1129 The long string labels themselves. The @code{labels} array contains
1130 exactly @code{n_labels} elements, each of which has the following
1141 @item int32 value_len;
1142 @itemx char value[];
1143 The string value being labeled. @code{value_len} is the number of
1144 bytes in @code{value}; it is equal to @code{var_width}. The
1145 @code{value} array is not padded or null-terminated.
1147 @item int32 label_len;
1148 @itemx char label[];
1149 The label for the string value. @code{label_len}, which must be
1150 between 0 and 120, is the number of bytes in @code{label}. The
1151 @code{label} array is not padded or null-terminated.
1155 @node Data File and Variable Attributes Records
1156 @section Data File and Variable Attributes Records
1158 The data file and variable attributes records represent custom
1159 attributes for the system file or for individual variables in the
1160 system file, as defined on the DATAFILE ATTRIBUTE (@pxref{DATAFILE
1161 ATTRIBUTE,,,pspp, PSPP Users Guide}) and VARIABLE ATTRIBUTE commands
1162 (@pxref{VARIABLE ATTRIBUTE,,,pspp, PSPP Users Guide}), respectively.
1171 /* @r{Exactly @code{count} bytes of data.} */
1176 @item int32 rec_type;
1177 Record type. Always set to 7.
1179 @item int32 subtype;
1180 Record subtype. Always set to 17 for a data file attribute record or
1181 to 18 for a variable attributes record.
1184 The size of each element in the @code{attributes} member. Always set to 1.
1187 The total number of bytes in @code{attributes}.
1189 @item char attributes[];
1190 The attributes, in a text-based format.
1192 In record type 17, this field contains a single attribute set. An
1193 attribute set is a sequence of one or more attributes concatenated
1194 together. Each attribute consists of a name, which has the same
1195 syntax as a variable name, followed by, inside parentheses, a sequence
1196 of one or more values. Each value consists of a string enclosed in
1197 single quotes (@code{'}) followed by a line feed (byte 0x0a). A value
1198 may contain single quote characters, which are not themselves escaped
1199 or quoted or required to be present in pairs. There is no apparent
1200 way to embed a line feed in a value. There is no distinction between
1201 an attribute with a single value and an attribute array with one
1204 In record type 18, this field contains a sequence of one or more
1205 variable attribute sets. If more than one variable attribute set is
1206 present, each one after the first is delimited from the previous by
1207 @code{/}. Each variable attribute set consists of a long
1209 followed by @code{:}, followed by an attribute set with the same
1210 syntax as on record type 17.
1212 The total length is @code{count} bytes.
1217 A system file produced with the following VARIABLE ATTRIBUTE commands
1221 VARIABLE ATTRIBUTE VARIABLES=dummy ATTRIBUTE=fred[1]('23') fred[2]('34').
1222 VARIABLE ATTRIBUTE VARIABLES=dummy ATTRIBUTE=bert('123').
1226 will contain a variable attribute record with the following contents:
1229 00000000 07 00 00 00 12 00 00 00 01 00 00 00 22 00 00 00 |............"...|
1230 00000010 64 75 6d 6d 79 3a 66 72 65 64 28 27 32 33 27 0a |dummy:fred('23'.|
1231 00000020 27 33 34 27 0a 29 62 65 72 74 28 27 31 32 33 27 |'34'.)bert('123'|
1232 00000030 0a 29 |.) |
1239 @node Variable Roles
1240 @subsection Variable Roles
1242 A variable's role is represented as an attribute named @code{$@@Role}.
1243 This attribute has a single element whose values and their meanings
1248 Input. This, the default, is the most common role.
1261 @node Extended Number of Cases Record
1262 @section Extended Number of Cases Record
1264 The file header record expresses the number of cases in the system
1265 file as an int32 (@pxref{File Header Record}). This record allows the
1266 number of cases in the system file to be expressed as a 64-bit number.
1278 @item int32 rec_type;
1279 Record type. Always set to 7.
1281 @item int32 subtype;
1282 Record subtype. Always set to 16.
1285 Size of each element. Always set to 8.
1288 Number of pieces of data in the data part. Alway set to 2.
1290 @item int64 unknown;
1291 Meaning unknown. Always set to 1.
1293 @item int64 ncases64;
1294 Number of cases in the file as a 64-bit integer. Presumably this
1295 could be -1 to indicate that the number of cases is unknown, for the
1296 same reason as @code{ncases} in the file header record, but this has
1297 not been observed in the wild.
1300 @node Miscellaneous Informational Records
1301 @section Miscellaneous Informational Records
1303 Some specific types of miscellaneous informational records are
1304 documented here, but others are known to exist. PSPP ignores unknown
1305 miscellaneous informational records when reading system files.
1314 /* @r{Exactly @code{size * count} bytes of data.} */
1319 @item int32 rec_type;
1320 Record type. Always set to 7.
1322 @item int32 subtype;
1323 Record subtype. May take any value. According to Aapi
1324 H@"am@"al@"ainen, value 5 indicates a set of grouped variables and 6
1325 indicates date info (probably related to USE). Subtype 24 appears to
1326 contain XML that describes how data in the file should be displayed
1330 Size of each piece of data in the data part. Should have the value 1,
1331 4, or 8, for @code{char}, @code{int32}, and @code{flt64} format data,
1335 Number of pieces of data in the data part.
1338 Arbitrary data. There must be @code{size} times @code{count} bytes of
1342 @node Dictionary Termination Record
1343 @section Dictionary Termination Record
1345 The dictionary termination record separates all other records from the
1354 @item int32 rec_type;
1355 Record type. Always set to 999.
1358 Ignored padding. Should be set to 0.
1362 @section Data Record
1364 Data records must follow all other records in the system file. There must
1365 be at least one data record in every system file.
1367 The format of data records varies depending on whether the data is
1368 compressed. Regardless, the data is arranged in a series of 8-byte
1371 When data is not compressed,
1372 each element corresponds to
1373 the variable declared in the respective variable record (@pxref{Variable
1374 Record}). Numeric values are given in @code{flt64} format; string
1375 values are literal characters string, padded on the right when
1376 necessary to fill out 8-byte units.
1378 Compressed data is arranged in the following manner: the first 8 bytes
1379 in the data section is divided into a series of 1-byte command
1380 codes. These codes have meanings as described below:
1384 Ignored. If the program writing the system file accumulates compressed
1385 data in blocks of fixed length, 0 bytes can be used to pad out extra
1386 bytes remaining at the end of a fixed-size block.
1390 value @var{code} - @var{bias}, where
1391 @var{code} is the value of the compression code and @var{bias} is the
1392 variable @code{bias} from the file header. For example,
1393 code 105 with bias 100.0 (the normal value) indicates a numeric variable
1395 One file has been seen written by SPSS 14 that contained such a code
1396 in a @emph{string} field with the value 0 (after the bias is
1397 subtracted) as a way of encoding null bytes.
1400 End of file. This code may or may not appear at the end of the data
1401 stream. PSPP always outputs this code but its use is not required.
1404 A numeric or string value that is not
1405 compressible. The value is stored in the 8 bytes following the
1406 current block of command bytes. If this value appears twice in a block
1407 of command bytes, then it indicates the second group of 8 bytes following the
1408 command bytes, and so on.
1411 An 8-byte string value that is all spaces.
1414 The system-missing value.
1417 When the end of the an 8-byte group of command bytes is reached, any
1418 blocks of non-compressible values indicated by code 253 are skipped,
1419 and the next element of command bytes is read and interpreted, until
1420 the end of the file or a code with value 252 is reached.
1421 @setfilename ignored