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 may use most character encodings based on an 8-bit unit.
60 UTF-16 and UTF-32, based on wider units, appear to be unacceptable.
61 @code{rec_type} in the file header record is sufficient to distinguish
62 between ASCII and EBCDIC based encodings. The best way to determine
63 the specific encoding in use is to consult the character encoding
64 record (@pxref{Character Encoding Record}), if present, and failing
65 that the @code{character_code} in the machine integer info record
66 (@pxref{Machine Integer Info Record}). The same encoding should be
67 used for the dictionary and the data in the file, although it is
68 possible to artificially synthesize files that use different encodings
69 (@pxref{Character Encoding Record}).
71 System files are divided into records, each of which begins with a
72 4-byte record type, usually regarded as an @code{int32}.
74 The records must appear in the following order:
84 All pairs of value labels records and value label variables records,
88 Document record, if present.
91 Extension (type 7) records, in ascending numerical order of their
95 Dictionary termination record.
101 Each type of record is described separately below.
104 * File Header Record::
106 * Value Labels Records::
108 * Machine Integer Info Record::
109 * Machine Floating-Point Info Record::
110 * Multiple Response Sets Records::
111 * Extra Product Info Record::
112 * Variable Display Parameter Record::
113 * Long Variable Names Record::
114 * Very Long String Record::
115 * Character Encoding Record::
116 * Long String Value Labels Record::
117 * Long String Missing Values Record::
118 * Data File and Variable Attributes Records::
119 * Extended Number of Cases Record::
120 * Miscellaneous Informational Records::
121 * Dictionary Termination Record::
123 * Encrypted System Files::
126 @node File Header Record
127 @section File Header Record
129 The file header is always the first record in the file. It has the
136 int32 nominal_case_size;
141 char creation_date[9];
142 char creation_time[8];
148 @item char rec_type[4];
149 Record type code, either @samp{$FL2} for system files with
150 uncompressed data or data compressed with simple bytecode compression,
151 or @samp{$FL3} for system files with ZLIB compressed data.
153 This is truly a character field that uses the character encoding as
154 other strings. Thus, in a file with an ASCII-based character encoding
155 this field contains @code{24 46 4c 32} or @code{24 46 4c 33}, and in a
156 file with an EBCDIC-based encoding this field contains @code{5b c6 d3
157 f2}. (No EBCDIC-based ZLIB-compressed files have been observed.)
159 @item char prod_name[60];
160 Product identification string. This always begins with the characters
161 @samp{@@(#) SPSS DATA FILE}. PSPP uses the remaining characters to
162 give its version and the operating system name; for example, @samp{GNU
163 pspp 0.1.4 - sparc-sun-solaris2.5.2}. The string is truncated if it
164 would be longer than 60 characters; otherwise it is padded on the right
168 @item int32 layout_code;
169 Normally set to 2, although a few system files have been spotted in
170 the wild with a value of 3 here. PSPP use this value to determine the
171 file's integer endianness (@pxref{System File Format}).
173 @item int32 nominal_case_size;
174 Number of data elements per case. This is the number of variables,
175 except that long string variables add extra data elements (one for every
176 8 characters after the first 8). However, string variables do not
177 contribute to this value beyond the first 255 bytes. Further, system
178 files written by some systems set this value to -1. In general, it is
179 unsafe for systems reading system files to rely upon this value.
181 @item int32 compression;
182 Set to 0 if the data in the file is not compressed, 1 if the data is
183 compressed with simple bytecode compression, 2 if the data is ZLIB
184 compressed. This field has value 2 if and only if @code{rec_type} is
187 @item int32 weight_index;
188 If one of the variables in the data set is used as a weighting
189 variable, set to the dictionary index of that variable, plus 1
190 (@pxref{Dictionary Index}). Otherwise, set to 0.
193 Set to the number of cases in the file if it is known, or -1 otherwise.
195 In the general case it is not possible to determine the number of cases
196 that will be output to a system file at the time that the header is
197 written. The way that this is dealt with is by writing the entire
198 system file, including the header, then seeking back to the beginning of
199 the file and writing just the @code{ncases} field. For files in which
200 this is not valid, the seek operation fails. In this case,
201 @code{ncases} remains -1.
205 Compression bias, ordinarily set to 100. Only integers between
206 @code{1 - bias} and @code{251 - bias} can be compressed.
208 By assuming that its value is 100, PSPP uses @code{bias} to determine
209 the file's floating-point format and endianness (@pxref{System File
210 Format}). If the compression bias is not 100, PSPP cannot auto-detect
211 the floating-point format and assumes that it is IEEE 754 format with
212 the same endianness as the system file's integers, which is correct
213 for all known system files.
215 @item char creation_date[9];
216 Date of creation of the system file, in @samp{dd mmm yy}
217 format, with the month as standard English abbreviations, using an
218 initial capital letter and following with lowercase. If the date is not
219 available then this field is arbitrarily set to @samp{01 Jan 70}.
221 @item char creation_time[8];
222 Time of creation of the system file, in @samp{hh:mm:ss}
223 format and using 24-hour time. If the time is not available then this
224 field is arbitrarily set to @samp{00:00:00}.
226 @item char file_label[64];
227 File label declared by the user, if any (@pxref{FILE LABEL,,,pspp,
228 PSPP Users Guide}). Padded on the right with spaces.
230 A product that identifies itself as @code{VOXCO INTERVIEWER 4.3} uses
231 CR-only line ends in this field, rather than the more usual LF-only or
234 @item char padding[3];
235 Ignored padding bytes to make the structure a multiple of 32 bits in
236 length. Set to zeros.
239 @node Variable Record
240 @section Variable Record
242 There must be one variable record for each numeric variable and each
243 string variable with width 8 bytes or less. String variables wider
244 than 8 bytes have one variable record for each 8 bytes, rounding up.
245 The first variable record for a long string specifies the variable's
246 correct dictionary information. Subsequent variable records for a
247 long string are filled with dummy information: a type of -1, no
248 variable label or missing values, print and write formats that are
249 ignored, and an empty string as name. A few system files have been
250 encountered that include a variable label on dummy variable records,
251 so readers should take care to parse dummy variable records in the
252 same way as other variable records.
254 @anchor{Dictionary Index}
255 The @dfn{dictionary index} of a variable is its offset in the set of
256 variable records, including dummy variable records for long string
257 variables. The first variable record has a dictionary index of 0, the
258 second has a dictionary index of 1, and so on.
260 The system file format does not directly support string variables
261 wider than 255 bytes. Such very long string variables are represented
262 by a number of narrower string variables. @xref{Very Long String
263 Record}, for details.
265 A system file should contain at least one variable and thus at least
266 one variable record, but system files have been observed in the wild
267 without any variables (thus, no data either).
273 int32 n_missing_values;
278 /* @r{Present only if @code{has_var_label} is 1.} */
282 /* @r{Present only if @code{n_missing_values} is nonzero}. */
283 flt64 missing_values[];
287 @item int32 rec_type;
288 Record type code. Always set to 2.
291 Variable type code. Set to 0 for a numeric variable. For a short
292 string variable or the first part of a long string variable, this is set
293 to the width of the string. For the second and subsequent parts of a
294 long string variable, set to -1, and the remaining fields in the
295 structure are ignored.
297 @item int32 has_var_label;
298 If this variable has a variable label, set to 1; otherwise, set to 0.
300 @item int32 n_missing_values;
301 If the variable has no missing values, set to 0. If the variable has
302 one, two, or three discrete missing values, set to 1, 2, or 3,
303 respectively. If the variable has a range for missing variables, set to
304 -2; if the variable has a range for missing variables plus a single
305 discrete value, set to -3.
307 A long string variable always has the value 0 here. A separate record
308 indicates missing values for long string variables (@pxref{Long String
309 Missing Values Record}).
312 Print format for this variable. See below.
315 Write format for this variable. See below.
318 Variable name. The variable name must begin with a capital letter or
319 the at-sign (@samp{@@}). Subsequent characters may also be digits, octothorpes
320 (@samp{#}), dollar signs (@samp{$}), underscores (@samp{_}), or full
321 stops (@samp{.}). The variable name is padded on the right with spaces.
323 The @samp{name} fields should be unique within a system file. System
324 files written by SPSS that contain very long string variables with
325 similar names sometimes contain duplicate names that are later
326 eliminated by resolving the very long string names (@pxref{Very Long
327 String Record}). PSPP handles duplicates by assigning them new,
330 @item int32 label_len;
331 This field is present only if @code{has_var_label} is set to 1. It is
332 set to the length, in characters, of the variable label. The
333 documented maximum length varies from 120 to 255 based on SPSS
334 version, but some files have been seen with longer labels. PSPP
335 accepts labels of any length.
338 This field is present only if @code{has_var_label} is set to 1. It has
339 length @code{label_len}, rounded up to the nearest multiple of 32 bits.
340 The first @code{label_len} characters are the variable's variable label.
342 @item flt64 missing_values[];
343 This field is present only if @code{n_missing_values} is nonzero. It
344 has the same number of 8-byte elements as the absolute value of
345 @code{n_missing_values}. Each element is interpreted as a number for
346 numeric variables (with HIGHEST and LOWEST indicated as described in
347 the chapter introduction). For string variables of width less than 8
348 bytes, elements are right-padded with spaces; for string variables
349 wider than 8 bytes, only the first 8 bytes of each missing value are
350 specified, with the remainder implicitly all spaces.
352 For discrete missing values, each element represents one missing
353 value. When a range is present, the first element denotes the minimum
354 value in the range, and the second element denotes the maximum value
355 in the range. When a range plus a value are present, the third
356 element denotes the additional discrete missing value.
359 @anchor{System File Output Formats}
360 The @code{print} and @code{write} members of sysfile_variable are output
361 formats coded into @code{int32} types. The least-significant byte
362 of the @code{int32} represents the number of decimal places, and the
363 next two bytes in order of increasing significance represent field width
364 and format type, respectively. The most-significant byte is not
365 used and should be set to zero.
367 Format types are defined as follows:
370 @multitable {Value} {@code{DATETIME}}
456 A few system files have been observed in the wild with invalid
457 @code{write} fields, in particular with value 0. Readers should
458 probably treat invalid @code{print} or @code{write} fields as some
461 @node Value Labels Records
462 @section Value Labels Records
464 The value label records documented in this section are used for
465 numeric and short string variables only. Long string variables may
466 have value labels, but their value labels are recorded using a
467 different record type (@pxref{Long String Value Labels Record}).
469 The value label record has the following format:
475 /* @r{Repeated @code{label_cnt} times}. */
482 @item int32 rec_type;
483 Record type. Always set to 3.
485 @item int32 label_count;
486 Number of value labels present in this record.
489 The remaining fields are repeated @code{count} times. Each
490 repetition specifies one value label.
494 A numeric value or a short string value padded as necessary to 8 bytes
495 in length. Its type and width cannot be determined until the
496 following value label variables record (see below) is read.
498 @item char label_len;
499 The label's length, in bytes. The documented maximum length varies
500 from 60 to 120 based on SPSS version. PSPP supports value labels up
504 @code{label_len} bytes of the actual label, followed by up to 7 bytes
505 of padding to bring @code{label} and @code{label_len} together to a
506 multiple of 8 bytes in length.
509 The value label record is always immediately followed by a value label
510 variables record with the following format:
519 @item int32 rec_type;
520 Record type. Always set to 4.
522 @item int32 var_count;
523 Number of variables that the associated value labels from the value
524 label record are to be applied.
527 A list of dictionary indexes of variables to which to apply the value
528 labels (@pxref{Dictionary Index}). There are @code{var_count}
531 String variables wider than 8 bytes may not be specified in this list.
534 @node Document Record
535 @section Document Record
537 The document record, if present, has the following format:
546 @item int32 rec_type;
547 Record type. Always set to 6.
550 Number of lines of documents present.
552 @item char lines[][80];
553 Document lines. The number of elements is defined by @code{n_lines}.
554 Lines shorter than 80 characters are padded on the right with spaces.
557 @node Machine Integer Info Record
558 @section Machine Integer Info Record
560 The integer info record, if present, has the following format:
572 int32 version_revision;
574 int32 floating_point_rep;
575 int32 compression_code;
577 int32 character_code;
581 @item int32 rec_type;
582 Record type. Always set to 7.
585 Record subtype. Always set to 3.
588 Size of each piece of data in the data part, in bytes. Always set to 4.
591 Number of pieces of data in the data part. Always set to 8.
593 @item int32 version_major;
594 PSPP major version number. In version @var{x}.@var{y}.@var{z}, this
597 @item int32 version_minor;
598 PSPP minor version number. In version @var{x}.@var{y}.@var{z}, this
601 @item int32 version_revision;
602 PSPP version revision number. In version @var{x}.@var{y}.@var{z},
605 @item int32 machine_code;
606 Machine code. PSPP always set this field to value to -1, but other
609 @item int32 floating_point_rep;
610 Floating point representation code. For IEEE 754 systems this is 1.
611 IBM 370 sets this to 2, and DEC VAX E to 3.
613 @item int32 compression_code;
614 Compression code. Always set to 1, regardless of whether or how the
617 @item int32 endianness;
618 Machine endianness. 1 indicates big-endian, 2 indicates little-endian.
620 @item int32 character_code;
621 @anchor{character-code} Character code. The following values have
622 been actually observed in system files:
632 The @code{windows-1250} code page for Central European and Eastern
636 The @code{windows-1252} code page for Western European languages.
645 The following additional values are known to be defined:
655 Other Windows code page numbers are known to be generally valid.
657 Old versions of SPSS for Unix and Windows always wrote value 2 in this
658 field, regardless of the encoding in use. Newer versions also write
659 the character encoding as a string (see @ref{Character Encoding
663 @node Machine Floating-Point Info Record
664 @section Machine Floating-Point Info Record
666 The floating-point info record, if present, has the following format:
682 @item int32 rec_type;
683 Record type. Always set to 7.
686 Record subtype. Always set to 4.
689 Size of each piece of data in the data part, in bytes. Always set to 8.
692 Number of pieces of data in the data part. Always set to 3.
695 The system missing value.
698 The value used for HIGHEST in missing values.
701 The value used for LOWEST in missing values.
704 @node Multiple Response Sets Records
705 @section Multiple Response Sets Records
707 The system file format has two different types of records that
708 represent multiple response sets (@pxref{MRSETS,,,pspp, PSPP Users
709 Guide}). The first type of record describes multiple response sets
710 that can be understood by SPSS before version 14. The second type of
711 record, with a closely related format, is used for multiple dichotomy
712 sets that use the CATEGORYLABELS=COUNTEDVALUES feature added in
722 /* @r{Exactly @code{count} bytes of data.} */
727 @item int32 rec_type;
728 Record type. Always set to 7.
731 Record subtype. Set to 7 for records that describe multiple response
732 sets understood by SPSS before version 14, or to 19 for records that
733 describe dichotomy sets that use the CATEGORYLABELS=COUNTEDVALUES
734 feature added in version 14.
737 The size of each element in the @code{mrsets} member. Always set to 1.
740 The total number of bytes in @code{mrsets}.
743 Zero or more line feeds (byte 0x0a), followed by a series of multiple
744 response sets, each of which consists of the following:
748 The set's name (an identifier that begins with @samp{$}), in mixed
749 upper and lower case.
752 An equals sign (@samp{=}).
755 @samp{C} for a multiple category set, @samp{D} for a multiple
756 dichotomy set with CATEGORYLABELS=VARLABELS, or @samp{E} for a
757 multiple dichotomy set with CATEGORYLABELS=COUNTEDVALUES.
760 For a multiple dichotomy set with CATEGORYLABELS=COUNTEDVALUES, a
761 space, followed by a number expressed as decimal digits, followed by a
762 space. If LABELSOURCE=VARLABEL was specified on MRSETS, then the
763 number is 11; otherwise it is 1.@footnote{This part of the format may
764 not be fully understood, because only a single example of each
765 possibility has been examined.}
768 For either kind of multiple dichotomy set, the counted value, as a
769 positive integer count specified as decimal digits, followed by a
770 space, followed by as many string bytes as specified in the count. If
771 the set contains numeric variables, the string consists of the counted
772 integer value expressed as decimal digits. If the set contains string
773 variables, the string contains the counted string value. Either way,
774 the string may be padded on the right with spaces (older versions of
775 SPSS seem to always pad to a width of 8 bytes; newer versions don't).
781 The multiple response set's label, using the same format as for the
782 counted value for multiple dichotomy sets. A string of length 0 means
783 that the set does not have a label. A string of length 0 is also
784 written if LABELSOURCE=VARLABEL was specified.
790 The short names of the variables in the set, converted to lowercase,
791 each separated from the previous by a single space.
793 Even though a multiple response set must have at least two variables,
794 some system files contain multiple response sets with no variables at
795 all. The source and meaning of these multiple response sets is
796 unknown. (Perhaps they arise from creating a multiple response set
797 then deleting all the variables that it contains?)
800 One line feed (byte 0x0a). Sometimes multiple, even hundreds, of line
805 Example: Given appropriate variable definitions, consider the
806 following MRSETS command:
809 MRSETS /MCGROUP NAME=$a LABEL='my mcgroup' VARIABLES=a b c
810 /MDGROUP NAME=$b VARIABLES=g e f d VALUE=55
811 /MDGROUP NAME=$c LABEL='mdgroup #2' VARIABLES=h i j VALUE='Yes'
812 /MDGROUP NAME=$d LABEL='third mdgroup' CATEGORYLABELS=COUNTEDVALUES
813 VARIABLES=k l m VALUE=34
814 /MDGROUP NAME=$e CATEGORYLABELS=COUNTEDVALUES LABELSOURCE=VARLABEL
815 VARIABLES=n o p VALUE='choice'.
818 The above would generate the following multiple response set record of
822 $a=C 10 my mcgroup a b c
824 $c=D3 Yes 10 mdgroup #2 h i j
827 It would also generate the following multiple response set record with
831 $d=E 1 2 34 13 third mdgroup k l m
832 $e=E 11 6 choice 0 n o p
835 @node Extra Product Info Record
836 @section Extra Product Info Record
838 This optional record appears to contain a text string that describes
839 the program that wrote the file and the source of the data. (This is
840 redundant with the file label and product info found in the file
850 /* @r{Exactly @code{count} bytes of data.} */
855 @item int32 rec_type;
856 Record type. Always set to 7.
859 Record subtype. Always set to 10.
862 The size of each element in the @code{info} member. Always set to 1.
865 The total number of bytes in @code{info}.
868 A text string. A product that identifies itself as @code{VOXCO
869 INTERVIEWER 4.3} uses CR-only line ends in this field, rather than the
870 more usual LF-only or CR LF line ends.
873 @node Variable Display Parameter Record
874 @section Variable Display Parameter Record
876 The variable display parameter record, if present, has the following
886 /* @r{Repeated @code{count} times}. */
888 int32 width; /* @r{Not always present.} */
893 @item int32 rec_type;
894 Record type. Always set to 7.
897 Record subtype. Always set to 11.
900 The size of @code{int32}. Always set to 4.
903 The number of sets of variable display parameters (ordinarily the
904 number of variables in the dictionary), times 2 or 3.
907 The remaining members are repeated @code{count} times, in the same
908 order as the variable records. No element corresponds to variable
909 records that continue long string variables. The meanings of these
910 members are as follows:
914 The measurement type of the variable:
924 SPSS sometimes writes a @code{measure} of 0. PSPP interprets this as
928 The width of the display column for the variable in characters.
930 This field is present if @var{count} is 3 times the number of
931 variables in the dictionary. It is omitted if @var{count} is 2 times
932 the number of variables.
934 @item int32 alignment;
935 The alignment of the variable for display purposes:
947 @node Long Variable Names Record
948 @section Long Variable Names Record
950 If present, the long variable names record has the following format:
959 /* @r{Exactly @code{count} bytes of data.} */
960 char var_name_pairs[];
964 @item int32 rec_type;
965 Record type. Always set to 7.
968 Record subtype. Always set to 13.
971 The size of each element in the @code{var_name_pairs} member. Always set to 1.
974 The total number of bytes in @code{var_name_pairs}.
976 @item char var_name_pairs[];
977 A list of @var{key}--@var{value} tuples, where @var{key} is the name
978 of a variable, and @var{value} is its long variable name.
979 The @var{key} field is at most 8 bytes long and must match the
980 name of a variable which appears in the variable record (@pxref{Variable
982 The @var{value} field is at most 64 bytes long.
983 The @var{key} and @var{value} fields are separated by a @samp{=} byte.
984 Each tuple is separated by a byte whose value is 09. There is no
985 trailing separator following the last tuple.
986 The total length is @code{count} bytes.
989 @node Very Long String Record
990 @section Very Long String Record
992 Old versions of SPSS limited string variables to a width of 255 bytes.
993 For backward compatibility with these older versions, the system file
994 format represents a string longer than 255 bytes, called a @dfn{very
995 long string}, as a collection of strings no longer than 255 bytes
996 each. The strings concatenated to make a very long string are called
997 its @dfn{segments}; for consistency, variables other than very long
998 strings are considered to have a single segment.
1000 A very long string with a width of @var{w} has @var{n} =
1001 (@var{w} + 251) / 252 segments, that is, one segment for every
1002 252 bytes of width, rounding up. It would be logical, then, for each
1003 of the segments except the last to have a width of 252 and the last
1004 segment to have the remainder, but this is not the case. In fact,
1005 each segment except the last has a width of 255 bytes. The last
1006 segment has width @var{w} - (@var{n} - 1) * 252; some versions
1007 of SPSS make it slightly wider, but not wide enough to make the last
1008 segment require another 8 bytes of data.
1010 Data is packed tightly into segments of a very long string, 255 bytes
1011 per segment. Because 255 bytes of segment data are allocated for
1012 every 252 bytes of the very long string's width (approximately), some
1013 unused space is left over at the end of the allocated segments. Data
1014 in unused space is ignored.
1016 Example: Consider a very long string of width 20,000. Such a very
1017 long string has 20,000 / 252 = 80 (rounding up) segments. The first
1018 79 segments have width 255; the last segment has width 20,000 - 79 *
1019 252 = 92 or slightly wider (up to 96 bytes, the next multiple of 8).
1020 The very long string's data is actually stored in the 19,890 bytes in
1021 the first 78 segments, plus the first 110 bytes of the 79th segment
1022 (19,890 + 110 = 20,000). The remaining 145 bytes of the 79th segment
1023 and all 92 bytes of the 80th segment are unused.
1025 The very long string record explains how to stitch together segments
1026 to obtain very long string data. For each of the very long string
1027 variables in the dictionary, it specifies the name of its first
1028 segment's variable and the very long string variable's actual width.
1029 The remaining segments immediately follow the named variable in the
1030 system file's dictionary.
1032 The very long string record, which is present only if the system file
1033 contains very long string variables, has the following format:
1042 /* @r{Exactly @code{count} bytes of data.} */
1043 char string_lengths[];
1047 @item int32 rec_type;
1048 Record type. Always set to 7.
1050 @item int32 subtype;
1051 Record subtype. Always set to 14.
1054 The size of each element in the @code{string_lengths} member. Always set to 1.
1057 The total number of bytes in @code{string_lengths}.
1059 @item char string_lengths[];
1060 A list of @var{key}--@var{value} tuples, where @var{key} is the name
1061 of a variable, and @var{value} is its length.
1062 The @var{key} field is at most 8 bytes long and must match the
1063 name of a variable which appears in the variable record (@pxref{Variable
1065 The @var{value} field is exactly 5 bytes long. It is a zero-padded,
1066 ASCII-encoded string that is the length of the variable.
1067 The @var{key} and @var{value} fields are separated by a @samp{=} byte.
1068 Tuples are delimited by a two-byte sequence @{00, 09@}.
1069 After the last tuple, there may be a single byte 00, or @{00, 09@}.
1070 The total length is @code{count} bytes.
1073 @node Character Encoding Record
1074 @section Character Encoding Record
1076 This record, if present, indicates the character encoding for string data,
1077 long variable names, variable labels, value labels and other strings in the
1087 /* @r{Exactly @code{count} bytes of data.} */
1092 @item int32 rec_type;
1093 Record type. Always set to 7.
1095 @item int32 subtype;
1096 Record subtype. Always set to 20.
1099 The size of each element in the @code{encoding} member. Always set to 1.
1102 The total number of bytes in @code{encoding}.
1104 @item char encoding[];
1105 The name of the character encoding. Normally this will be an official
1106 IANA character set name or alias.
1107 See @url{http://www.iana.org/assignments/character-sets}.
1108 Character set names are not case-sensitive, but SPSS appears to write
1109 them in all-uppercase.
1112 This record is not present in files generated by older software. See
1113 also the @code{character_code} field in the machine integer info
1114 record (@pxref{character-code}).
1116 When the character encoding record and the machine integer info record
1117 are both present, all system files observed in practice indicate the
1118 same character encoding, e.g.@: 1252 as @code{character_code} and
1119 @code{windows-1252} as @code{encoding}, 65001 and @code{UTF-8}, etc.
1121 If, for testing purposes, a file is crafted with different
1122 @code{character_code} and @code{encoding}, it seems that
1123 @code{character_code} controls the encoding for all strings in the
1124 system file before the dictionary termination record, including
1125 strings in data (e.g.@: string missing values), and @code{encoding}
1126 controls the encoding for strings following the dictionary termination
1129 @node Long String Value Labels Record
1130 @section Long String Value Labels Record
1132 This record, if present, specifies value labels for long string
1142 /* @r{Repeated up to exactly @code{count} bytes.} */
1147 long_string_label labels[];
1151 @item int32 rec_type;
1152 Record type. Always set to 7.
1154 @item int32 subtype;
1155 Record subtype. Always set to 21.
1161 The number of bytes following the header until the next header.
1163 @item int32 var_name_len;
1164 @itemx char var_name[];
1165 The number of bytes in the name of the variable that has long string
1166 value labels, plus the variable name itself, which consists of exactly
1167 @code{var_name_len} bytes. The variable name is not padded to any
1168 particular boundary, nor is it null-terminated.
1170 @item int32 var_width;
1171 The width of the variable, in bytes, which will be between 9 and
1174 @item int32 n_labels;
1175 @itemx long_string_label labels[];
1176 The long string labels themselves. The @code{labels} array contains
1177 exactly @code{n_labels} elements, each of which has the following
1188 @item int32 value_len;
1189 @itemx char value[];
1190 The string value being labeled. @code{value_len} is the number of
1191 bytes in @code{value}; it is equal to @code{var_width}. The
1192 @code{value} array is not padded or null-terminated.
1194 @item int32 label_len;
1195 @itemx char label[];
1196 The label for the string value. @code{label_len}, which must be
1197 between 0 and 120, is the number of bytes in @code{label}. The
1198 @code{label} array is not padded or null-terminated.
1202 @node Long String Missing Values Record
1203 @section Long String Missing Values Record
1205 This record, if present, specifies missing values for long string
1215 /* @r{Repeated up to exactly @code{count} bytes.} */
1218 char n_missing_values;
1219 long_string_missing_value values[];
1223 @item int32 rec_type;
1224 Record type. Always set to 7.
1226 @item int32 subtype;
1227 Record subtype. Always set to 22.
1233 The number of bytes following the header until the next header.
1235 @item int32 var_name_len;
1236 @itemx char var_name[];
1237 The number of bytes in the name of the long string variable that has
1238 missing values, plus the variable name itself, which consists of
1239 exactly @code{var_name_len} bytes. The variable name is not padded to
1240 any particular boundary, nor is it null-terminated.
1242 @item char n_missing_values;
1243 The number of missing values, either 1, 2, or 3. (This is, unusually,
1244 a single byte instead of a 32-bit number.)
1246 @item long_string_missing_value values[];
1247 The missing values themselves. This array contains exactly
1248 @code{n_missing_values} elements, each of which has the following
1257 @item int32 value_len;
1258 The length of the missing value string, in bytes. This value should
1259 be 8, because long string variables are at least 8 bytes wide (by
1260 definition), only the first 8 bytes of a long string variable's
1261 missing values are allowed to be non-spaces, and any spaces within the
1262 first 8 bytes are included in the missing value here.
1265 The missing value string, exactly @code{value_len} bytes, without
1266 any padding or null terminator.
1270 @node Data File and Variable Attributes Records
1271 @section Data File and Variable Attributes Records
1273 The data file and variable attributes records represent custom
1274 attributes for the system file or for individual variables in the
1275 system file, as defined on the DATAFILE ATTRIBUTE (@pxref{DATAFILE
1276 ATTRIBUTE,,,pspp, PSPP Users Guide}) and VARIABLE ATTRIBUTE commands
1277 (@pxref{VARIABLE ATTRIBUTE,,,pspp, PSPP Users Guide}), respectively.
1286 /* @r{Exactly @code{count} bytes of data.} */
1291 @item int32 rec_type;
1292 Record type. Always set to 7.
1294 @item int32 subtype;
1295 Record subtype. Always set to 17 for a data file attribute record or
1296 to 18 for a variable attributes record.
1299 The size of each element in the @code{attributes} member. Always set to 1.
1302 The total number of bytes in @code{attributes}.
1304 @item char attributes[];
1305 The attributes, in a text-based format.
1307 In record type 17, this field contains a single attribute set. An
1308 attribute set is a sequence of one or more attributes concatenated
1309 together. Each attribute consists of a name, which has the same
1310 syntax as a variable name, followed by, inside parentheses, a sequence
1311 of one or more values. Each value consists of a string enclosed in
1312 single quotes (@code{'}) followed by a line feed (byte 0x0a). A value
1313 may contain single quote characters, which are not themselves escaped
1314 or quoted or required to be present in pairs. There is no apparent
1315 way to embed a line feed in a value. There is no distinction between
1316 an attribute with a single value and an attribute array with one
1319 In record type 18, this field contains a sequence of one or more
1320 variable attribute sets. If more than one variable attribute set is
1321 present, each one after the first is delimited from the previous by
1322 @code{/}. Each variable attribute set consists of a long
1324 followed by @code{:}, followed by an attribute set with the same
1325 syntax as on record type 17.
1327 The total length is @code{count} bytes.
1332 A system file produced with the following VARIABLE ATTRIBUTE commands
1336 VARIABLE ATTRIBUTE VARIABLES=dummy ATTRIBUTE=fred[1]('23') fred[2]('34').
1337 VARIABLE ATTRIBUTE VARIABLES=dummy ATTRIBUTE=bert('123').
1341 will contain a variable attribute record with the following contents:
1344 0000 07 00 00 00 12 00 00 00 01 00 00 00 22 00 00 00 |............"...|
1345 0010 64 75 6d 6d 79 3a 66 72 65 64 28 27 32 33 27 0a |dummy:fred('23'.|
1346 0020 27 33 34 27 0a 29 62 65 72 74 28 27 31 32 33 27 |'34'.)bert('123'|
1354 @node Variable Roles
1355 @subsection Variable Roles
1357 A variable's role is represented as an attribute named @code{$@@Role}.
1358 This attribute has a single element whose values and their meanings
1363 Input. This, the default, is the most common role.
1376 @node Extended Number of Cases Record
1377 @section Extended Number of Cases Record
1379 The file header record expresses the number of cases in the system
1380 file as an int32 (@pxref{File Header Record}). This record allows the
1381 number of cases in the system file to be expressed as a 64-bit number.
1393 @item int32 rec_type;
1394 Record type. Always set to 7.
1396 @item int32 subtype;
1397 Record subtype. Always set to 16.
1400 Size of each element. Always set to 8.
1403 Number of pieces of data in the data part. Alway set to 2.
1405 @item int64 unknown;
1406 Meaning unknown. Always set to 1.
1408 @item int64 ncases64;
1409 Number of cases in the file as a 64-bit integer. Presumably this
1410 could be -1 to indicate that the number of cases is unknown, for the
1411 same reason as @code{ncases} in the file header record, but this has
1412 not been observed in the wild.
1415 @node Miscellaneous Informational Records
1416 @section Miscellaneous Informational Records
1418 Some specific types of miscellaneous informational records are
1419 documented here, but others are known to exist. PSPP ignores unknown
1420 miscellaneous informational records when reading system files.
1429 /* @r{Exactly @code{size * count} bytes of data.} */
1434 @item int32 rec_type;
1435 Record type. Always set to 7.
1437 @item int32 subtype;
1438 Record subtype. May take any value. According to Aapi
1439 H@"am@"al@"ainen, value 5 indicates a set of grouped variables and 6
1440 indicates date info (probably related to USE). Subtype 24 appears to
1441 contain XML that describes how data in the file should be displayed
1445 Size of each piece of data in the data part. Should have the value 1,
1446 4, or 8, for @code{char}, @code{int32}, and @code{flt64} format data,
1450 Number of pieces of data in the data part.
1453 Arbitrary data. There must be @code{size} times @code{count} bytes of
1457 @node Dictionary Termination Record
1458 @section Dictionary Termination Record
1460 The dictionary termination record separates all other records from the
1469 @item int32 rec_type;
1470 Record type. Always set to 999.
1473 Ignored padding. Should be set to 0.
1477 @section Data Record
1479 The data record must follow all other records in the system file.
1480 Every system file must have a data record that specifies data for at
1481 least one case. The format of the data record varies depending on the
1482 value of @code{compression} in the file header record:
1485 @item 0: no compression
1486 Data is arranged as a series of 8-byte elements.
1487 Each element corresponds to
1488 the variable declared in the respective variable record (@pxref{Variable
1489 Record}). Numeric values are given in @code{flt64} format; string
1490 values are literal characters string, padded on the right when
1491 necessary to fill out 8-byte units.
1493 @item 1: bytecode compression
1495 of the data record is divided into a series of 1-byte command
1496 codes. These codes have meanings as described below:
1500 Ignored. If the program writing the system file accumulates compressed
1501 data in blocks of fixed length, 0 bytes can be used to pad out extra
1502 bytes remaining at the end of a fixed-size block.
1506 value @var{code} - @var{bias}, where
1507 @var{code} is the value of the compression code and @var{bias} is the
1508 variable @code{bias} from the file header. For example,
1509 code 105 with bias 100.0 (the normal value) indicates a numeric variable
1511 One file has been seen written by SPSS 14 that contained such a code
1512 in a @emph{string} field with the value 0 (after the bias is
1513 subtracted) as a way of encoding null bytes.
1516 End of file. This code may or may not appear at the end of the data
1517 stream. PSPP always outputs this code but its use is not required.
1520 A numeric or string value that is not
1521 compressible. The value is stored in the 8 bytes following the
1522 current block of command bytes. If this value appears twice in a block
1523 of command bytes, then it indicates the second group of 8 bytes following the
1524 command bytes, and so on.
1527 An 8-byte string value that is all spaces.
1530 The system-missing value.
1533 The end of the 8-byte group of bytecodes is followed by any 8-byte
1534 blocks of non-compressible values indicated by code 253. After that
1535 follows another 8-byte group of bytecodes, then those bytecodes'
1536 non-compressible values. The pattern repeats to the end of the file
1537 or a code with value 252.
1539 @item 2: ZLIB compression
1540 The data record consists of the following, in order:
1544 ZLIB data header, 24 bytes long.
1547 One or more variable-length blocks of ZLIB compressed data.
1550 ZLIB data trailer, with a 24-byte fixed header plus an additional 24
1551 bytes for each preceding ZLIB compressed data block.
1554 The ZLIB data header has the following format:
1563 @item int64 zheader_ofs;
1564 The offset, in bytes, of the beginning of this structure within the
1567 @item int64 ztrailer_ofs;
1568 The offset, in bytes, of the first byte of the ZLIB data trailer.
1570 @item int64 ztrailer_len;
1571 The number of bytes in the ZLIB data trailer. This and the previous
1572 field sum to the size of the system file in bytes.
1575 The data header is followed by @code{(ztrailer_ofs - 24) / 24} ZLIB
1576 compressed data blocks. Each ZLIB compressed data block begins with a
1577 ZLIB header as specified in RFC@tie{}1950, e.g.@: hex bytes @code{78
1578 01} (the only header yet observed in practice). Each block
1579 decompresses to a fixed number of bytes (in practice only
1580 @code{0x3ff000}-byte blocks have been observed), except that the last
1581 block of data may be shorter. The last ZLIB compressed data block
1582 gends just before offset @code{ztrailer_ofs}.
1584 The result of ZLIB decompression is bytecode compressed data as
1585 described above for compression format 1.
1587 The ZLIB data trailer begins with the following 24-byte fixed header:
1597 @item int64 int_bias;
1598 The compression bias as a negative integer, e.g.@: if @code{bias} in
1599 the file header record is 100.0, then @code{int_bias} is @minus{}100
1600 (this is the only value yet observed in practice).
1603 Always observed to be zero.
1605 @item int32 block_size;
1606 The number of bytes in each ZLIB compressed data block, except
1607 possibly the last, following decompression. Only @code{0x3ff000} has
1608 been observed so far.
1610 @item int32 n_blocks;
1611 The number of ZLIB compressed data blocks, always exactly
1612 @code{(ztrailer_ofs - 24) / 24}.
1615 The fixed header is followed by @code{n_blocks} 24-byte ZLIB data
1616 block descriptors, each of which describes the compressed data block
1617 corresponding to its offset. Each block descriptor has the following
1621 int64 uncompressed_ofs;
1622 int64 compressed_ofs;
1623 int32 uncompressed_size;
1624 int32 compressed_size;
1628 @item int64 uncompressed_ofs;
1629 The offset, in bytes, that this block of data would have in a similar
1630 system file that uses compression format 1. This is
1631 @code{zheader_ofs} in the first block descriptor, and in each
1632 succeeding block descriptor it is the sum of the previous desciptor's
1633 @code{uncompressed_ofs} and @code{uncompressed_size}.
1635 @item int64 compressed_ofs;
1636 The offset, in bytes, of the actual beginning of this compressed data
1637 block. This is @code{zheader_ofs + 24} in the first block descriptor,
1638 and in each succeeding block descriptor it is the sum of the previous
1639 descriptor's @code{compressed_ofs} and @code{compressed_size}. The
1640 final block descriptor's @code{compressed_ofs} and
1641 @code{compressed_size} sum to @code{ztrailer_ofs}.
1643 @item int32 uncompressed_size;
1644 The number of bytes in this data block, after decompression. This is
1645 @code{block_size} in every data block except the last, which may be
1648 @item int32 compressed_size;
1649 The number of bytes in this data block, as stored compressed in this
1654 @setfilename ignored
1656 @node Encrypted System Files
1657 @section Encrypted System Files
1659 SPSS 21 and later support an encrypted system file format.
1662 The SPSS encrypted file format is poorly designed. It is much cheaper
1663 and faster to decrypt a file encrypted this way than if a well
1664 designed alternative were used. If you must use this format, use a
1665 10-byte randomly generated password.
1668 @subheading Encrypted File Format
1670 Encrypted system files begin with the following 36-byte fixed header:
1673 0000 1c 00 00 00 00 00 00 00 45 4e 43 52 59 50 54 45 |........ENCRYPTE|
1674 0010 44 53 41 56 15 00 00 00 00 00 00 00 00 00 00 00 |DSAV............|
1675 0020 00 00 00 00 |....|
1678 Following the fixed header is a complete system file in the usual
1679 format, except that each 16-byte block is encrypted with AES-256 in
1680 ECB mode. The AES-256 key is derived from a password in the following
1685 Start from the literal password typed by the user. Truncate it to at
1686 most 10 bytes, then append (between 1 and 22) null bytes until there
1687 are exactly 32 bytes. Call this @var{password}.
1690 Let @var{constant} be the following 73-byte constant:
1693 0000 00 00 00 01 35 27 13 cc 53 a7 78 89 87 53 22 11
1694 0010 d6 5b 31 58 dc fe 2e 7e 94 da 2f 00 cc 15 71 80
1695 0020 0a 6c 63 53 00 38 c3 38 ac 22 f3 63 62 0e ce 85
1696 0030 3f b8 07 4c 4e 2b 77 c7 21 f5 1a 80 1d 67 fb e1
1697 0040 e1 83 07 d8 0d 00 00 01 00
1701 Compute CMAC-AES-256(@var{password}, @var{constant}). Call the
1702 16-byte result @var{cmac}.
1705 The 32-byte AES-256 key is @var{cmac} || @var{cmac}, that is,
1706 @var{cmac} repeated twice.
1709 @subsubheading Example
1711 Consider the password @samp{pspp}. @var{password} is:
1714 0000 70 73 70 70 00 00 00 00 00 00 00 00 00 00 00 00 |pspp............|
1715 0010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
1722 0000 3e da 09 8e 66 04 d4 fd f9 63 0c 2c a8 6f b0 45
1729 0000 3e da 09 8e 66 04 d4 fd f9 63 0c 2c a8 6f b0 45
1730 0010 3e da 09 8e 66 04 d4 fd f9 63 0c 2c a8 6f b0 45
1733 @subheading Password Encoding
1735 SPSS also supports what it calls ``encrypted passwords.'' These are
1736 not encrypted. They are encoded with a simple, fixed scheme. An
1737 encoded password is always a multiple of 2 characters long, and never
1738 longer than 20 characters. The characters in an encoded password are
1739 always in the graphic ASCII range 33 through 126. Each successive
1740 pair of characters in the password encodes a single byte in the
1743 Use the following algorithm to decode a pair of characters:
1747 Let @var{a} be the ASCII code of the first character, and @var{b} be
1748 the ASCII code of the second character.
1751 Let @var{ah} be the most significant 4 bits of @var{a}. Find the line
1752 in the table below that has @var{ah} on the left side. The right side
1753 of the line is a set of possible values for the most significant 4
1754 bits of the decoded byte.
1757 @t{2 } @result{} @t{2367}
1758 @t{3 } @result{} @t{0145}
1759 @t{47} @result{} @t{89cd}
1760 @t{56} @result{} @t{abef}
1764 Let @var{bh} be the most significant 4 bits of @var{b}. Find the line
1765 in the second table below that has @var{bh} on the left side. The
1766 right side of the line is a set of possible values for the most
1767 significant 4 bits of the decoded byte. Together with the results of
1768 the previous step, only a single possibility is left.
1771 @t{2 } @result{} @t{139b}
1772 @t{3 } @result{} @t{028a}
1773 @t{47} @result{} @t{46ce}
1774 @t{56} @result{} @t{57df}
1778 Let @var{al} be the least significant 4 bits of @var{a}. Find the
1779 line in the table below that has @var{al} on the left side. The right
1780 side of the line is a set of possible values for the least significant
1781 4 bits of the decoded byte.
1784 @t{03cf} @result{} @t{0145}
1785 @t{12de} @result{} @t{2367}
1786 @t{478b} @result{} @t{89cd}
1787 @t{569a} @result{} @t{abef}
1791 Let @var{bl} be the least significant 4 bits of @var{b}. Find the
1792 line in the table below that has @var{bl} on the left side. The right
1793 side of the line is a set of possible values for the least significant
1794 4 bits of the decoded byte. Together with the results of the previous
1795 step, only a single possibility is left.
1798 @t{03cf} @result{} @t{028a}
1799 @t{12de} @result{} @t{139b}
1800 @t{478b} @result{} @t{46ce}
1801 @t{569a} @result{} @t{57df}
1805 @subsubheading Example
1807 Consider the encoded character pair @samp{-|}. @var{a} is
1808 0x2d and @var{b} is 0x7c, so @var{ah} is 2, @var{bh} is 7, @var{al} is
1809 0xd, and @var{bl} is 0xc. @var{ah} means that the most significant
1810 four bits of the decoded character is 2, 3, 6, or 7, and @var{bh}
1811 means that they are 4, 6, 0xc, or 0xe. The single possibility in
1812 common is 6, so the most significant four bits are 6. Similarly,
1813 @var{al} means that the least significant four bits are 2, 3, 6, or 7,
1814 and @var{bl} means they are 0, 2, 8, or 0xa, so the least significant
1815 four bits are 2. The decoded character is therefore 0x62, the letter