1 | #!/usr/bin/env python |
---|
2 | # This tool takes a tab-delimited text file as input and creates filters on columns based on certain properties. |
---|
3 | # The tool will skip over invalid lines within the file, informing the user about the number of lines skipped. |
---|
4 | |
---|
5 | from __future__ import division |
---|
6 | import sys, re, os.path |
---|
7 | from galaxy import eggs |
---|
8 | |
---|
9 | # Older py compatibility |
---|
10 | try: |
---|
11 | set() |
---|
12 | except: |
---|
13 | from sets import Set as set |
---|
14 | |
---|
15 | assert sys.version_info[:2] >= ( 2, 4 ) |
---|
16 | |
---|
17 | def get_operands( filter_condition ): |
---|
18 | # Note that the order of all_operators is important |
---|
19 | items_to_strip = ['+', '-', '**', '*', '//', '/', '%', '<<', '>>', '&', '|', '^', '~', '<=', '<', '>=', '>', '==', '!=', '<>', ' and ', ' or ', ' not ', ' is ', ' is not ', ' in ', ' not in '] |
---|
20 | for item in items_to_strip: |
---|
21 | if filter_condition.find( item ) >= 0: |
---|
22 | filter_condition = filter_condition.replace( item, ' ' ) |
---|
23 | operands = set( filter_condition.split( ' ' ) ) |
---|
24 | return operands |
---|
25 | |
---|
26 | def stop_err( msg ): |
---|
27 | sys.stderr.write( msg ) |
---|
28 | sys.exit() |
---|
29 | |
---|
30 | in_fname = sys.argv[1] |
---|
31 | out_fname = sys.argv[2] |
---|
32 | cond_text = sys.argv[3] |
---|
33 | try: |
---|
34 | in_columns = int( sys.argv[4] ) |
---|
35 | assert sys.argv[5] #check to see that the column types varaible isn't null |
---|
36 | in_column_types = sys.argv[5].split( ',' ) |
---|
37 | except: |
---|
38 | stop_err( "Data does not appear to be tabular. This tool can only be used with tab-delimited data." ) |
---|
39 | |
---|
40 | # Unescape if input has been escaped |
---|
41 | mapped_str = { |
---|
42 | '__lt__': '<', |
---|
43 | '__le__': '<=', |
---|
44 | '__eq__': '==', |
---|
45 | '__ne__': '!=', |
---|
46 | '__gt__': '>', |
---|
47 | '__ge__': '>=', |
---|
48 | '__sq__': '\'', |
---|
49 | '__dq__': '"', |
---|
50 | } |
---|
51 | for key, value in mapped_str.items(): |
---|
52 | cond_text = cond_text.replace( key, value ) |
---|
53 | |
---|
54 | # Attempt to determine if the condition includes executable stuff and, if so, exit |
---|
55 | secured = dir() |
---|
56 | operands = get_operands(cond_text) |
---|
57 | for operand in operands: |
---|
58 | try: |
---|
59 | check = int( operand ) |
---|
60 | except: |
---|
61 | if operand in secured: |
---|
62 | stop_err( "Illegal value '%s' in condition '%s'" % ( operand, cond_text ) ) |
---|
63 | |
---|
64 | # Prepare the column variable names and wrappers for column data types |
---|
65 | cols, type_casts = [], [] |
---|
66 | for col in range( 1, in_columns + 1 ): |
---|
67 | col_name = "c%d" % col |
---|
68 | cols.append( col_name ) |
---|
69 | col_type = in_column_types[ col - 1 ] |
---|
70 | type_cast = "%s(%s)" % ( col_type, col_name ) |
---|
71 | type_casts.append( type_cast ) |
---|
72 | |
---|
73 | col_str = ', '.join( cols ) # 'c1, c2, c3, c4' |
---|
74 | type_cast_str = ', '.join( type_casts ) # 'str(c1), int(c2), int(c3), str(c4)' |
---|
75 | assign = "%s = line.split( '\\t' )" % col_str |
---|
76 | wrap = "%s = %s" % ( col_str, type_cast_str ) |
---|
77 | skipped_lines = 0 |
---|
78 | first_invalid_line = 0 |
---|
79 | invalid_line = None |
---|
80 | lines_kept = 0 |
---|
81 | total_lines = 0 |
---|
82 | out = open( out_fname, 'wt' ) |
---|
83 | |
---|
84 | # Read and filter input file, skipping invalid lines |
---|
85 | code = ''' |
---|
86 | for i, line in enumerate( file( in_fname ) ): |
---|
87 | total_lines += 1 |
---|
88 | line = line.rstrip( '\\r\\n' ) |
---|
89 | if not line or line.startswith( '#' ): |
---|
90 | skipped_lines += 1 |
---|
91 | if not invalid_line: |
---|
92 | first_invalid_line = i + 1 |
---|
93 | invalid_line = line |
---|
94 | continue |
---|
95 | try: |
---|
96 | %s |
---|
97 | %s |
---|
98 | if %s: |
---|
99 | lines_kept += 1 |
---|
100 | print >> out, line |
---|
101 | except: |
---|
102 | skipped_lines += 1 |
---|
103 | if not invalid_line: |
---|
104 | first_invalid_line = i + 1 |
---|
105 | invalid_line = line |
---|
106 | ''' % ( assign, wrap, cond_text ) |
---|
107 | |
---|
108 | valid_filter = True |
---|
109 | try: |
---|
110 | exec code |
---|
111 | except Exception, e: |
---|
112 | out.close() |
---|
113 | if str( e ).startswith( 'invalid syntax' ): |
---|
114 | valid_filter = False |
---|
115 | stop_err( 'Filter condition "%s" likely invalid. See tool tips, syntax and examples.' % cond_text ) |
---|
116 | else: |
---|
117 | stop_err( str( e ) ) |
---|
118 | |
---|
119 | if valid_filter: |
---|
120 | out.close() |
---|
121 | valid_lines = total_lines - skipped_lines |
---|
122 | print 'Filtering with %s, ' % cond_text |
---|
123 | if valid_lines > 0: |
---|
124 | print 'kept %4.2f%% of %d lines.' % ( 100.0*lines_kept/valid_lines, total_lines ) |
---|
125 | else: |
---|
126 | print 'Possible invalid filter condition "%s" or non-existent column referenced. See tool tips, syntax and examples.' % cond_text |
---|
127 | if skipped_lines > 0: |
---|
128 | print 'Skipped %d invalid lines starting at line #%d: "%s"' % ( skipped_lines, first_invalid_line, invalid_line ) |
---|