From 597fead93976d0ac408b30f0f96dd8e50e514294 Mon Sep 17 00:00:00 2001
From: "Adam J. Gamble" <mail@adamgamble.com>
Date: Mon, 26 Nov 2012 23:32:01 +0000
Subject: [PATCH] Added some [belated] exception handling. * namedtuple mapping
 catch errors with headings, partial fix. * fix: pass instead of break to send
 options.headings for all ARGS, lazier.

---
 colly/commands/__init__.py |  5 +++--
 colly/commands/diff.py     |  5 +++++
 colly/core.py              | 25 +++++++++++++++++--------
 colly/exceptions.py        |  2 ++
 4 files changed, 27 insertions(+), 10 deletions(-)
 create mode 100644 colly/exceptions.py

diff --git a/colly/commands/__init__.py b/colly/commands/__init__.py
index a1431f2..82eeb07 100644
--- a/colly/commands/__init__.py
+++ b/colly/commands/__init__.py
@@ -5,6 +5,7 @@ import optparse
 
 from colly.optbase import parser
 from colly.core import Collate
+from colly.exceptions import CommandError
 
 __all__ = ['command_dict', 'Command', 'load_command',
            'load_all_commands', 'command_names']
@@ -101,12 +102,12 @@ class CsvCommand(Command):
                 try:
                     kwargs['headings'] = options.headings[n]
                 except IndexError:
-                    break
+                    pass
             elif options.index:
                  try:
                      kwargs['headings'] = [] # leave out index for now
                  except IndexError:
-                     break
+                     pass
             else:
                  pass
             
diff --git a/colly/commands/diff.py b/colly/commands/diff.py
index 8440292..ad87bd7 100644
--- a/colly/commands/diff.py
+++ b/colly/commands/diff.py
@@ -2,6 +2,7 @@ import sys
 import logging
 
 from colly.commands import CsvCommand
+from colly.exceptions import CommandError
 
 class DiffCommand(CsvCommand):
     name = 'diff'
@@ -48,6 +49,10 @@ class DiffCommand(CsvCommand):
             'all' : A.column.union(B.column)
         }[options.method]
 
+        if not ('pk' in A.headings or 'pk' in B.headings):
+           logging.warning("pk should be set, for csv file A and B")
+           raise CommandError("Please set 'pk' columns for both files")
+        
         for row in filtered:
             try:
                 print A.get_row(row)
diff --git a/colly/core.py b/colly/core.py
index 6424f49..24d7535 100644
--- a/colly/core.py
+++ b/colly/core.py
@@ -4,11 +4,14 @@ import logging
 import csv
 from collections import namedtuple
 
+from colly.exceptions import CsvImportError
+
 import simplejson as json
 
 class Collate(object):
     
     _rows = {}
+    headings = ""
     
     ''' Csv parsing and formatting.
     '''
@@ -24,18 +27,19 @@ class Collate(object):
                 f.seek(0) # reset iter to first row ^
             else:
                 headings = first_row
-                if 'pk' in options:
-                    try:
-                        pk = int(options['pk'])
-                        headings[pk] = 'pk' #: denote primary key
-                    except ValueError:
-                        logging.warning('Ignoring pk, should be a number denoting the primary key column')
+                logging.info('Interpretting headings as row #1')
+#                if 'pk' in options:
+#                    try:
+#                        pk = int(options['pk'])
+#                        headings[pk] = 'pk' #: denote primary key
+#                    except ValueError:
+#                        logging.warning('Ignoring pk, should be a number denoting the primary key column')
             
             cL = len(first_row) #: column length
             hL = len(headings)
             if hL != cL:
                 if cL < hL: 
-                    raise "Given headings exceeded those in CSV"
+                    raise CsvImportError("Given headings exceeded those in CSV")
                 else:
                     ''' Pad extra columns alphabeticaly
                         * fieldnames have to be unique, & not empty, or number.
@@ -43,14 +47,19 @@ class Collate(object):
                     pad = list(string.uppercase[hL:cL]) #:FIX/ limited to 26 column CSV
                     headings += pad 
             
+            self.headings = headings
             logging.info("CSV columns given headings: %s" % (headings))
             
             ''' OK, columns should be in order, ready to generate a map of the
                 @csv_file.
             '''
-            Head = namedtuple("row", ",".join(headings), verbose=False)
             index, pk = {}, 0 #: set empty vars
             
+            try:
+                Head = namedtuple("row", ",".join(headings), verbose=False)
+            except ValueError, err:
+                raise CsvImportError(err)
+            
             for row in map(Head._make, raw):
                 ''' Perhaps not immediately obvious, index contains the full 
                     dataset. The "pk" (primary key) should be unique, or if 
diff --git a/colly/exceptions.py b/colly/exceptions.py
new file mode 100644
index 0000000..269616e
--- /dev/null
+++ b/colly/exceptions.py
@@ -0,0 +1,2 @@
+class CsvImportError(Exception): pass
+class CommandError(Exception): pass
-- 
2.11.4.GIT