Add --date-format and --datetime-format options to set strptime forma…

…t strings.

CHANGELOG.rst

@@ -6,6 +6,8 @@ Improvements:
 * Add a :code:`--version` flag.
 * Add a :code:`--skip-lines` option to skip initial lines (e.g. comments, copyright notices, empty rows).
 * Add a :code:`--locale` option to set the locale of any formatted numbers.
+* Add a :code:`--date-format` option to set a strptime date format string.
+* Add a :code:`--datetime-format` option to set a strptime datetime format string.
 * :code:`-I` is the short option for :code:`--no-inference`.
 * :doc:`/scripts/csvjoin` supports :code:`--snifflimit` and :code:`--no-inference`.
 * :doc:`/scripts/csvstat` adds a :code:`--freq-count` option to set the maximum number of frequent values to display.

csvkit/cli.py

@@ -166,6 +166,12 @@ def _init_common_parser(self):
         if 'S' not in self.override_flags:
             self.argparser.add_argument('-S', '--skipinitialspace', dest='skipinitialspace', action='store_true',
                                         help='Ignore whitespace immediately following the delimiter.')
+        if 'date-format' not in self.override_flags:
+            self.argparser.add_argument('--date-format', dest='date_format',
+                                        help='Specify a strptime date format string like "%%m/%%d/%%Y".')
+        if 'datetime-format' not in self.override_flags:
+            self.argparser.add_argument('--datetime-format', dest='datetime_format',
+                                        help='Specify a strptime datetime format string like "%%m/%%d/%%Y %%I:%%M %%p".')
         if 'H' not in self.override_flags:
             self.argparser.add_argument('-H', '--no-header-row', dest='no_header_row', action='store_true',
                                         help='Specify that the input CSV file has no header row. Will create default headers (a,b,c,...).')
@@ -288,8 +294,8 @@ def get_column_types(self):
                 agate.Boolean(),
                 agate.Number(locale=self.args.locale),
                 agate.TimeDelta(),
-                agate.Date(),
-                agate.DateTime(),
+                agate.Date(date_format=self.args.date_format),
+                agate.DateTime(datetime_format=self.args.datetime_format),
                 text_type
             ])
 

csvkit/utilities/csvclean.py

@@ -11,7 +11,7 @@
 
 class CSVClean(CSVKitUtility):
     description = 'Fix common errors in a CSV file.'
-    override_flags = ['H', 'L']
+    override_flags = ['H', 'L', 'date-format', 'datetime-format']
 
     def add_arguments(self):
         self.argparser.add_argument('-n', '--dry-run', dest='dryrun', action='store_true',

csvkit/utilities/csvcut.py

@@ -16,7 +16,7 @@
 
 class CSVCut(CSVKitUtility):
     description = 'Filter and truncate CSV files. Like the Unix "cut" command, but for tabular data.'
-    override_flags = ['L']
+    override_flags = ['L', 'date-format', 'datetime-format']
 
     def add_arguments(self):
         self.argparser.add_argument('-n', '--names', dest='names_only', action='store_true',

csvkit/utilities/csvformat.py

@@ -7,7 +7,7 @@
 
 class CSVFormat(CSVKitUtility):
     description = 'Convert a CSV file to a custom output format.'
-    override_flags = ['l', 'zero', 'H', 'L']
+    override_flags = ['l', 'zero', 'H', 'L', 'date-format', 'datetime-format']
 
     def add_arguments(self):
         self.argparser.add_argument('-D', '--out-delimiter', dest='out_delimiter',

csvkit/utilities/csvgrep.py

@@ -11,7 +11,7 @@
 
 class CSVGrep(CSVKitUtility):
     description = 'Search CSV files. Like the Unix "grep" command, but for tabular data.'
-    override_flags = ['L']
+    override_flags = ['L', 'date-format', 'datetime-format']
 
     def add_arguments(self):
         self.argparser.add_argument('-n', '--names', dest='names_only', action='store_true',

csvkit/utilities/csvjson.py

@@ -37,7 +37,7 @@ def add_arguments(self):
         self.argparser.add_argument('-y', '--snifflimit', dest='sniff_limit', type=int,
                                     help='Limit CSV dialect sniffing to the specified number of bytes. Specify "0" to disable sniffing entirely.')
         self.argparser.add_argument('-I', '--no-inference', dest='no_inference', action='store_true',
-                                    help='Disable type inference when parsing CSV input.')
+                                    help='Disable type inference (and --locale, --date-format, --datetime-format) when parsing CSV input.')
 
     def main(self):
         # We need to do this dance here, because we aren't writing through agate.

csvkit/utilities/csvstack.py

@@ -9,7 +9,7 @@
 
 class CSVStack(CSVKitUtility):
     description = 'Stack up the rows from multiple CSV files, optionally adding a grouping value.'
-    override_flags = ['f', 'K', 'L']
+    override_flags = ['f', 'K', 'L', 'date-format', 'datetime-format']
 
     def add_arguments(self):
         self.argparser.add_argument(metavar="FILE", nargs='+', dest='input_paths', default=['-'],

csvkit/utilities/csvstat.py

@@ -63,7 +63,7 @@
 
 class CSVStat(CSVKitUtility):
     description = 'Print descriptive statistics for each column in a CSV file.'
-    override_flags = ['l', 'L']
+    override_flags = ['l', 'L', 'date-format', 'datetime-format']
 
     def add_arguments(self):
         self.argparser.add_argument('--csv', dest='csv_output', action='store_true',

csvkit/utilities/in2csv.py

@@ -35,7 +35,7 @@ def add_arguments(self):
         self.argparser.add_argument('-y', '--snifflimit', dest='sniff_limit', type=int,
                                     help='Limit CSV dialect sniffing to the specified number of bytes. Specify "0" to disable sniffing entirely.')
         self.argparser.add_argument('-I', '--no-inference', dest='no_inference', action='store_true',
-                                    help='Disable type inference when parsing CSV input.')
+                                    help='Disable type inference (and --locale, --date-format, --datetime-format) when parsing CSV input.')
 
     def main(self):
         # Determine the file type.

csvkit/utilities/sql2csv.py

@@ -10,7 +10,7 @@
 
 class SQL2CSV(CSVKitUtility):
     description = 'Execute an SQL query on a database and output the result to a CSV file.'
-    override_flags = 'f,b,d,e,H,K,L,p,q,S,t,u,z,zero'.split(',')
+    override_flags = 'f,b,d,e,H,K,L,p,q,S,t,u,z,date-format,datetime-format,zero'.split(',')
 
     def add_arguments(self):
         self.argparser.add_argument('--db', dest='connection_string', default='sqlite://',

docs/common_arguments.rst

@@ -29,6 +29,11 @@ All tools which accept CSV as input share a set of common command-line arguments
                         Specify the locale (en_US) of any formatted numbers.
   -S, --skipinitialspace
                         Ignore whitespace immediately following the delimiter.
+  --date-format DATE_FORMAT
+                        Specify a strptime date format string like "%m/%d/%Y".
+  --datetime-format DATETIME_FORMAT
+                        Specify a strptime datetime format string like
+                        "%m/%d/%Y %I:%M %p".
   -H, --no-header-row   Specify that the input CSV file has no header row.
                         Will create default headers (A,B,C,...).
   -L SKIP_LINES, --skip-lines SKIP_LINES

tests/test_utilities/test_in2csv.py

@@ -37,6 +37,9 @@ def test_version(self):
     def test_locale(self):
         self.assertConverted('csv', 'examples/test_locale.csv', 'examples/test_locale_converted.csv', ['--locale', 'de_DE'])
 
+    def test_date_format(self):
+        self.assertConverted('csv', 'examples/test_date_format.csv', 'examples/test_date_format_converted.csv', ['--date-format', '%d/%m/%Y'])
+
     def test_convert_csv(self):
         self.assertConverted('csv', 'examples/testfixed_converted.csv', 'examples/testfixed_converted.csv')