summaryrefslogtreecommitdiff
path: root/tko/query_lib.py
blob: 35e72808c41ebd423f8f958a51da5083cbdcaa45 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
#!/usr/bin/python
"""
This library provides a bunch of miscellaneous parameter parsing,
sql generating and list cleanup library functions that are used
by both the reporting cli and web interface.
"""

import sys, os, re

tko = os.path.dirname(__file__)
sys.path.insert(0, tko)

import display, frontend, db

db = db.db()

def dprint(str):
    pass
    #print "! %s<br>" % str

def parse_scrub_and_gen_condition(condition, valid_field_dict):
    me = parse_scrub_and_gen_condition   # shorten the name
    compare_ops = {'=':'=', '<>':'<>', '==':'=', '!=':'<>', '>':'>',
                   '<':'<', '>=':'>=', '<=':'<=', '~':'LIKE', '#':'REGEXP'}

    # strip white space
    condition = condition.strip()

    # ()'s
    #match = re.match(r'^[(](.+)[)]$', condition)
    #if match:
    #       dprint("Matched () on %s" % condition)
    #       depth = 0
    #       for c in match.group(1):
    #               if c == '(':    depth += 1
    #               if c == ')':    depth -= 1
    #               if depth < 0:   break
    #       dprint("Depth is %d" % depth)
    #       if depth == 0:
    #               dprint("Match...stripping ()'s")
    #               return me(match.group(1), valid_field_dict)

    # OR
    match = re.match(r'^(.+)[|](.+)$', condition)
    if match:
        dprint("Matched | on %s" % condition)
        (a_sql, a_values) = me(match.group(1), valid_field_dict)
        (b_sql, b_values) = me(match.group(2), valid_field_dict)
        return (" (%s) OR (%s) " % (a_sql, b_sql),
                a_values + b_values)

    # AND
    match = re.match(r'^(.+)[&](.+)$', condition)
    if match:
        dprint("Matched & on %s" % condition)
        (a_sql, a_values) = me(match.group(1), valid_field_dict)
        (b_sql, b_values) = me(match.group(2), valid_field_dict)
        return (" (%s) AND (%s) " % (a_sql, b_sql),
                a_values + b_values)

    # NOT
    #match = re.match(r'^[!](.+)$', condition)
    #if match:
    #       dprint("Matched ! on %s" % condition)
    #       (sql, values) = me(match.group(1), valid_field_dict)
    #       return (" NOT (%s) " % (sql,), values)

    # '<field> <op> <value>' where value can be quoted
    # double quotes are escaped....i.e.  '''' is the same as "'"
    regex = r'^(%s)[ \t]*(%s)[ \t]*' + \
            r'(\'((\'\'|[^\'])*)\'|"((""|[^"])*)"|([^\'"].*))$'
    regex = regex % ('|'.join(valid_field_dict.keys()),
                     '|'.join(compare_ops.keys()))
    match = re.match(regex, condition)
    if match:
        field = valid_field_dict[match.group(1)]
        op = compare_ops[match.group(2)]
        if match.group(5):
            val = match.group(4).replace("''", "'")
        elif match.group(7):
            val = match.group(6).replace('""', '"')
        elif match.group(8):
            val = match.group(8)
        else:
            raise "Internal error"
        return ("%s %s %%s" % (field, op), [val])


    raise "Could not parse '%s' (%s)" % (condition, regex)