added length checking to field matching criteria for parser

pyaccuwage/parser.py

@@ -1,7 +1,13 @@
 #!/usr/bin/python
 # coding=UTF-8
+"""
+Parser utility to read data from Publication 1220 and
+convert it into python classes.
+
+"""
 import re
 
+
 class SimpleDefParser(object):
     def __init__(self):
         pass
@@ -35,6 +41,36 @@ class SimpleDefParser(object):
             yield item
         
 
+class LengthExpression(object):
+    import operator
+    REG = re.compile(r'([^\d]*)(\d+)')
+    OPS = {
+        '<': operator.lt,
+        '>': operator.gt,
+        '<=': operator.le,
+        '>=': operator.ge,
+        '=': operator.eq,
+    }
+
+    def __init__(self):
+        self.exp_cache = {}
+
+    def __call__(self, value, exps):
+        return len(exps) == sum(map(lambda x: self.check(value, x), exps))
+
+    def compile_exp(self, exp):
+        op, val = self.REG.match(exp).groups()
+        val = int(val)
+        return (self.OPS.get(op, None), val)
+
+    def check(self, value, exp):
+        if exp not in self.exp_cache:
+            (op, opval) = self.exp_cache[exp] = self.compile_exp(exp)
+        else:
+            (op, opval) = self.exp_cache[exp]
+        return op(value, opval)
+
+
 class BaseToken(object):
     regexp = re.compile('(.*)')
 
@@ -85,17 +121,41 @@ class PastedDefParser(object):
     ]
 
     FIELD_TYPES = [
-        (fields.BlankField, {'name': [
-                                re.compile(r'^blank$'),
-                                ]}),
-        (fields.MoneyField, {'desc': [
-                                re.compile(r'right\-justified'),
-                                re.compile(r'amount'),
-                                re.compile(r'zero\-filled'),
-                                ]}),
-        (fields.TextField,  {'desc': [
-                                re.compile(r'enter blanks')
-                                ]})
+        (fields.BlankField, {
+            'regexp': {
+                'name': [
+                    re.compile(r'^blank$'),
+                ],
+            },
+        }),
+    
+        (fields.MoneyField, {
+            'regexp': {
+                'desc': [
+                    re.compile(r'right\-justified'),
+                    re.compile(r'amount'),
+                    re.compile(r'zero\-filled'),
+                ],
+            },
+        }),
+    
+        (fields.TextField, {
+            'regexp': {
+                'desc': [
+                    re.compile(r'enter blanks'),
+                ],
+            },
+        }),
+
+        (fields.StateField, {
+            'regexp': {
+                'desc': [
+                    re.compile(r'state'),
+                    re.compile(r'postal'),
+                ],
+            },
+            'length': ['=2'],
+        })
     ]
 
     def load(self, infile):
@@ -204,13 +264,20 @@ class PastedDefParser(object):
 
 
     def _guess_field_types(self, entries):
+        lengthexp = LengthExpression()
+
         for entry in entries:
             matches = dict(map(lambda x:(x[0],0), self.FIELD_TYPES))
             
             for (classtype, criteria) in self.FIELD_TYPES:
-                for crit_key, crit_values in criteria.items():
-                    for crit_re in crit_values:
-                        matches[classtype] += 1 if crit_re.search(entry[crit_key]) else 0
+                if 'length' in criteria:
+                    if not lengthexp(int(entry['length']), criteria['length']):
+                        continue
+            
+                if 'regexp' in criteria:
+                    for crit_key, crit_values in criteria['regexp'].items():
+                        for crit_re in crit_values:
+                            matches[classtype] += 1 if crit_re.search(entry[crit_key]) else 0
                     
             
             matches = list(matches.items())
@@ -230,10 +297,10 @@ class PastedDefParser(object):
 
             # FIELD NAME
             if entry['name'] == 'blank':
-                add('blank%d' % blank_count)
+                add( (u'blank%d' % blank_count).ljust(40) )
                 blank_count += 1
             else:
-                add(entry['name'])
+                add(entry['name'].ljust(40))
 
             add(' = ')