1 AT_BANNER([format guesser])
3 AT_SETUP([format guesser])
4 AT_DATA([format-guesser.txt], [dnl
27 "1,123" => COMMA5.0 # Is , is grouping or decimal? Assume grouping.
28 "1.123" => F5.3 # Ditto.
29 "1,12" => F4.2 # Not a group of 3, so last delim must be decimal.
30 "1.12" => F4.2 # Ditto.
31 "1,1234" => F6.4 # Not a group of 3, so last delim must be decimal.
32 "1.1234" => F6.4 # Ditto.
33 "$1.234" => DOLLAR6.3 # Dollar sign means decimal has to be '.'.
34 "$1,234" => DOLLAR6.0 # Ditto.
35 "1.234%" => PCT6.3 # Percent sign means decimal has to be '.'.
36 "1,234%" => PCT6.0 # Ditto.
37 "1,123.456" => COMMA9.3 # Both '.' and ',', so last delim must be decimal.
38 "1.123,456" => DOT9.3 # Ditto.
39 "1,123,456.45" => COMMA12.2 # Ditto.
40 "1.123.456,45" => DOT12.2 # Ditto.
41 "1,123,456" => COMMA9.0 # Ditto.
42 "1.123.456" => DOT9.0 # Ditto.
44 # Date and time formats.
45 "01-OCT-1978" => DATE11
46 "01-x-1978" => EDATE9 # Roman numeral for month.
48 "1-13-99" => ADATE7 (ADATE8)
50 "13-1-99" => EDATE7 (EDATE8)
51 "32-1-1" => SDATE6 (SDATE8)
62 "1-1-01 1:2" => A10 # Minute needs at least two digits.
63 "1-1-01 1:02" => DATETIME11.0 (DATETIME17.0)
64 "1-1-01 1:02:3" => A13 # Second needs at least two digits.
65 "1-1-01 1:02:03" => DATETIME20.0
66 "1-1-01 1:02:03.1" => DATETIME20.1 (DATETIME22.1)
67 "1-1-01 +1:02:03.1" => DATETIME20.1 (DATETIME22.1)
68 "1-1-01 -1:02:03.1" => DATETIME20.1 (DATETIME22.1)
69 "1:30" => TIME4.0 (TIME5.0)
75 "-1:30:15.5" => TIME10.1
76 "+1:30:15.75" => TIME11.2
77 "1 1:30" => DTIME6.0 (DTIME8.0)
78 "+1 1:30" => DTIME7.0 (DTIME8.0)
79 "-1 1:30" => DTIME7.0 (DTIME8.0)
83 "1:00:01.03" => TIME10.2 (TIME11.2)
84 "12 1:00:01.3" => DTIME12.1 (DTIME13.1)
88 "i" => A1 # Not detected as MONTH format.
93 "janaury" => MONTH7 # Only first three letters are significant.
103 "september" => MONTH9
110 "wedensday" => WKDAY9
116 # Ambiguous; bias in favor of more sensible DD/MM/YY format:
120 # Several ambiguous dates can be clarified by one unambiguous example:
121 "1/1/1978" "1/2/1978" "1/3/1978" "1/13/1978" => ADATE9 # MM/DD/YY
122 "01/01/01" "02/01/01" "03/01/01" "13/01/01" => EDATE8 # DD/MM/YY
123 "01/01/01" "02/01/01" "03/01/01" "2013/01/01" => SDATE10 # YY/MM/DD
125 AT_CHECK([[(echo "SET DECIMAL=DOT."
129 s,^\(.*\)=> \(.*\)$,DEBUG FORMAT GUESSER \1.,') \
130 < format-guesser.txt > format-guesser.sps]])
131 AT_CHECK([pspp --testing-mode -O format=csv format-guesser.sps],
133 AT_CHECK([[sed -e 's/#.*//
135 /^$/d' < format-guesser.txt > expout]])
136 AT_CHECK([cat stderr], [0], [expout])