dnl PSPP - a program for statistical analysis.
dnl Copyright (C) 2017 Free Software Foundation, Inc.
dnl
dnl This program is free software: you can redistribute it and/or modify
dnl it under the terms of the GNU General Public License as published by
dnl the Free Software Foundation, either version 3 of the License, or
dnl (at your option) any later version.
dnl
dnl This program is distributed in the hope that it will be useful,
dnl but WITHOUT ANY WARRANTY; without even the implied warranty of
dnl MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
dnl GNU General Public License for more details.
dnl
dnl You should have received a copy of the GNU General Public License
dnl along with this program.  If not, see <http://www.gnu.org/licenses/>.
dnl
AT_BANNER([format guesser])

AT_SETUP([format guesser])
AT_DATA([format-guesser.txt], [dnl
# No data.
=> F8.2
"" => F8.2
"." => F8.2

# Numeric formats.
"1.2" => F3.1
"$1.2" => DOLLAR4.1
"1.2%" => PCT4.1
"$1.2%" => A5
"1e5" => E3.0
"1e+5" => E4.0
"1+5" => E3.0
"1-5" => E3.0
"1.2e5" => E5.1
"1.3e+5" => E6.1
"1.4+5" => E5.1
"1e" => A2
"1e+" => A3
"1+" => A2
"1-" => A2
"1.5-5" => E5.1
"1,123" => COMMA5.0         # Is , is grouping or decimal? Assume grouping.
"1.123" => F5.3		    # Ditto.
"1,12" => F4.2		    # Not a group of 3, so last delim must be decimal.
"1.12" => F4.2		    # Ditto.
"1,1234" => F6.4	    # Not a group of 3, so last delim must be decimal.
"1.1234" => F6.4	    # Ditto.
"$1.234" => DOLLAR6.3       # Dollar sign means decimal has to be '.'.
"$1,234" => DOLLAR6.0       # Ditto.
"1.234%" => PCT6.3          # Percent sign means decimal has to be '.'.
"1,234%" => PCT6.0          # Ditto.
"1,123.456" => COMMA9.3	    # Both '.' and ',', so last delim must be decimal.
"1.123,456" => DOT9.3	    # Ditto.
"1,123,456.45" => COMMA12.2 # Ditto.
"1.123.456,45" => DOT12.2   # Ditto.
"1,123,456" => COMMA9.0	    # Ditto.
"1.123.456" => DOT9.0	    # Ditto.

# Date and time formats.
"01-OCT-1978" => DATE11
"01-x-1978" => EDATE9		# Roman numeral for month.
"01-13-99" => ADATE8
"1-13-99" => ADATE7 (ADATE8)
"13-01-99" => EDATE8
"13-1-99" => EDATE7 (EDATE8)
"32-1-1" => SDATE6 (SDATE8)
"1q01" => QYR4
"1Q01" => QYR4
"1 q 01" => QYR6
"1 Q 01" => QYR6
"1q2001" => QYR6
"1Q2001" => QYR6
"1 q 2001" => QYR8
"1 Q 2001" => QYR8
"oct 05" => MOYR6
"oct 2005" => MOYR8
"1-1-01 1:2" => A10 		# Minute needs at least two digits.
"1-1-01 1:02" => DATETIME11.0 (DATETIME17.0)
"1-1-01 1:02:3" => A13		# Second needs at least two digits.
"1-1-01 1:02:03" => DATETIME20.0
"1-1-01 1:02:03.1" => DATETIME20.1 (DATETIME22.1)
"1-1-01 +1:02:03.1" => DATETIME20.1 (DATETIME22.1)
"1-1-01 -1:02:03.1" => DATETIME20.1 (DATETIME22.1)
# XXX YMDHMS formats
# XXX MTIME formats
"1:30" => TIME4.0 (TIME5.0)
"1:30:05" => TIME8.0
"-1:30" => TIME5.0
"+1:30" => TIME5.0
"-1:30:15" => TIME8.0
"+1:30:15" => TIME8.0
"-1:30:15.5" => TIME10.1
"+1:30:15.75" => TIME11.2
"1 1:30" => DTIME6.0 (DTIME8.0)
"+1 1:30" => DTIME7.0 (DTIME8.0)
"-1 1:30" => DTIME7.0 (DTIME8.0)
"-1-13-99" => A8
"+1-13-99" => A8
"1+13+99" => A7
"1:00:01.03" => TIME10.2 (TIME11.2)
"12 1:00:01.3" => DTIME12.1 (DTIME13.1)
"jan" => MONTH3
"Feb" => MONTH3
"MAR" => MONTH3
"i" => A1			# Not detected as MONTH format.
"v" => A1
"ix" => A2
"x" => A1
"january" => MONTH7
"janaury" => MONTH7		# Only first three letters are significant.
"february" => MONTH8
"febraury" => MONTH8
"march" => MONTH5
"marhc" => MONTH5
"april" => MONTH5
"may" => MONTH3
"june" => MONTH4
"july" => MONTH4
"august" => MONTH6
"september" => MONTH9
"october" => MONTH7
"november" => MONTH8
"decmeber" => MONTH8
"december" => MONTH8
"monady" => WKDAY6
"tuseday" => WKDAY7
"wedensday" => WKDAY9
"thurdsay" => WKDAY8
"fridya" => WKDAY6
"saturady" => WKDAY8
"sudnay" => WKDAY6

# Ambiguous; bias in favor of more sensible DD/MM/YY format:
"1/1/1978" => EDATE8
"01/01/01" => EDATE8

# Several ambiguous dates can be clarified by one unambiguous example:
"1/1/1978" "1/2/1978" "1/3/1978" "1/13/1978" => ADATE9 # MM/DD/YY
"01/01/01" "02/01/01" "03/01/01" "13/01/01" => EDATE8  # DD/MM/YY
"01/01/01" "02/01/01" "03/01/01" "2013/01/01" => SDATE10 # YY/MM/DD
])
AT_CHECK([[(echo "SET DECIMAL=DOT."
           sed -e 's/#.*//
s/[ 	]*$//
/^$/d
s,^\(.*\)=> \(.*\)$,DEBUG FORMAT GUESSER \1.,') \
           < format-guesser.txt > format-guesser.sps]])
AT_CHECK([pspp --testing-mode -O format=csv format-guesser.sps],
  [0], [], [stderr])
AT_CHECK([[sed -e 's/#.*//
s/[ 	]*$//
/^$/d' < format-guesser.txt > expout]])
AT_CHECK([cat stderr], [0], [expout])
AT_CLEANUP