textutils 0.5.12 → 0.5.13
Sign up to get free protection for your applications and to get access to all the features.
@@ -24,6 +24,30 @@ module TextUtils
|
|
24
24
|
|
25
25
|
new_address_line
|
26
26
|
end
|
27
|
+
|
28
|
+
|
29
|
+
# todo/fix: add _in_adr or _in_addr to name - why? why not?
|
30
|
+
|
31
|
+
def find_city_for_country( country_key, address )
|
32
|
+
|
33
|
+
return nil if address.blank? # do NOT process nil or empty address lines; sorry
|
34
|
+
|
35
|
+
lines = address.split( '//' )
|
36
|
+
|
37
|
+
if country_key == 'at' || country_key == 'de'
|
38
|
+
# first line strip numbers (assuming zip code) and whitespace
|
39
|
+
line1 = lines[0]
|
40
|
+
line1 = line1.gsub( /\b[0-9]+\b/, '' ) # use word boundries (why? why not?)
|
41
|
+
line1 = line1.strip
|
42
|
+
|
43
|
+
return nil if line1.blank? # nothing left sorry; better return nil
|
44
|
+
|
45
|
+
line1 # assume its the city
|
46
|
+
else
|
47
|
+
nil # unsupported country/address schema for now; sorry
|
48
|
+
end
|
49
|
+
end
|
50
|
+
|
27
51
|
|
28
52
|
end # module AddressHelper
|
29
53
|
end # module TextUtils
|
@@ -34,5 +34,26 @@ module TextUtils
|
|
34
34
|
end
|
35
35
|
|
36
36
|
|
37
|
+
def find_grade( text ) # NB: returns ary [grade,text] / two values
|
38
|
+
grade = 4 # defaults to grade 4 e.g *** => 1, ** => 2, * => 3, -/- => 4
|
39
|
+
|
40
|
+
text = text.sub( /\s+(\*{1,3})\s*$/ ) do |_| # NB: stars must end field/value
|
41
|
+
if $1 == '***'
|
42
|
+
grade = 1
|
43
|
+
elsif $1 == '**'
|
44
|
+
grade = 2
|
45
|
+
elsif $1 == '*'
|
46
|
+
grade = 3
|
47
|
+
else
|
48
|
+
# unknown grade; not possible, is'it?
|
49
|
+
end
|
50
|
+
'' # remove * from title if found
|
51
|
+
end
|
52
|
+
|
53
|
+
[grade,text]
|
54
|
+
end
|
55
|
+
|
56
|
+
|
57
|
+
|
37
58
|
end # module ValueHelper
|
38
59
|
end # module TextUtils
|
@@ -5,6 +5,9 @@
|
|
5
5
|
class ValuesReader
|
6
6
|
|
7
7
|
include LogUtils::Logging
|
8
|
+
|
9
|
+
include TextUtils::ValueHelper # e.g. includes find_grade()
|
10
|
+
|
8
11
|
|
9
12
|
def initialize( path, more_values={} )
|
10
13
|
@path = path
|
@@ -68,7 +71,7 @@ class ValuesReader
|
|
68
71
|
if line =~ /^[a-z][a-z0-9.]*[a-z0-9]:/
|
69
72
|
# NB: every additional line is one value e.g. city:wien, etc.
|
70
73
|
# allows you to use any chars
|
71
|
-
logger.debug " multi-line record - add key-value
|
74
|
+
logger.debug " multi-line record - add key-value »#{line}«"
|
72
75
|
|
73
76
|
more_cols.unshift( line.dup ) # add value upfront to array (first value); lets us keep (optional) tags as last entry; fix!! see valuereaderEx v2
|
74
77
|
next
|
@@ -85,18 +88,17 @@ class ValuesReader
|
|
85
88
|
|
86
89
|
|
87
90
|
### guard escaped commas (e.g. \,)
|
88
|
-
line = line.gsub( '\,', '
|
91
|
+
line = line.gsub( '\,', '♣' ) # use black club suit/=shamrock char for escaped separator
|
89
92
|
|
90
93
|
## use generic separator (allow us to configure separator)
|
91
|
-
line = line.gsub( ',', '
|
94
|
+
line = line.gsub( ',', '♦') # use black diamond suit for separator
|
92
95
|
|
93
96
|
## restore escaped commas (before split)
|
94
|
-
line = line.gsub( '
|
95
|
-
|
97
|
+
line = line.gsub( '♣', ',' )
|
96
98
|
|
97
|
-
logger.debug "line:
|
99
|
+
logger.debug "line: »#{line}«"
|
98
100
|
|
99
|
-
values = line.split( '
|
101
|
+
values = line.split( '♦' )
|
100
102
|
|
101
103
|
# pass 1) remove leading and trailing whitespace for values
|
102
104
|
|
@@ -107,14 +109,14 @@ class ValuesReader
|
|
107
109
|
|
108
110
|
values = values.select do |value|
|
109
111
|
if value =~ /^#/ ## start with # treat it as a comment column; e.g. remove it
|
110
|
-
logger.debug " removing column with value
|
112
|
+
logger.debug " removing column with value »#{value}«"
|
111
113
|
false
|
112
114
|
else
|
113
115
|
true
|
114
116
|
end
|
115
117
|
end
|
116
118
|
|
117
|
-
logger.debug " values:
|
119
|
+
logger.debug " values: »#{values.join('« »')}«"
|
118
120
|
|
119
121
|
|
120
122
|
### todo/fix: allow check - do NOT allow mixed use of with key and w/o key
|
@@ -141,6 +143,15 @@ class ValuesReader
|
|
141
143
|
|
142
144
|
attribs = {}
|
143
145
|
|
146
|
+
## check title_col for grade (e.g. ***/**/*) and use returned stripped title_col if exits
|
147
|
+
grade, title_col = find_grade( title_col )
|
148
|
+
|
149
|
+
# NB: for now - do NOT include default grade e.g. if grade (***/**/*) not present; attrib will not be present too
|
150
|
+
if grade == 1 || grade == 2 || grade == 3 # grade found/present
|
151
|
+
logger.debug " found grade #{grade} in title"
|
152
|
+
attribs[:grade] = grade
|
153
|
+
end
|
154
|
+
|
144
155
|
## title (split of optional synonyms)
|
145
156
|
# e.g. FC Bayern Muenchen|Bayern Muenchen|Bayern
|
146
157
|
titles = title_col.split('|')
|
@@ -153,7 +164,7 @@ class ValuesReader
|
|
153
164
|
if key_col == '<auto>'
|
154
165
|
## autogenerate key from first title
|
155
166
|
key_col = TextUtils.title_to_key( titles[0] )
|
156
|
-
logger.debug " autogen key
|
167
|
+
logger.debug " autogen key »#{key_col}« from title »#{titles[0]}«, textutils version #{TextUtils::VERSION}"
|
157
168
|
end
|
158
169
|
|
159
170
|
attribs[ :key ] = key_col
|
data/lib/textutils/version.rb
CHANGED
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: textutils
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.5.
|
4
|
+
version: 0.5.13
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,11 +9,11 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2013-05-
|
12
|
+
date: 2013-05-14 00:00:00.000000000 Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: logutils
|
16
|
-
requirement: &
|
16
|
+
requirement: &83466980 !ruby/object:Gem::Requirement
|
17
17
|
none: false
|
18
18
|
requirements:
|
19
19
|
- - ~>
|
@@ -21,10 +21,10 @@ dependencies:
|
|
21
21
|
version: '0.5'
|
22
22
|
type: :runtime
|
23
23
|
prerelease: false
|
24
|
-
version_requirements: *
|
24
|
+
version_requirements: *83466980
|
25
25
|
- !ruby/object:Gem::Dependency
|
26
26
|
name: rdoc
|
27
|
-
requirement: &
|
27
|
+
requirement: &83466720 !ruby/object:Gem::Requirement
|
28
28
|
none: false
|
29
29
|
requirements:
|
30
30
|
- - ~>
|
@@ -32,10 +32,10 @@ dependencies:
|
|
32
32
|
version: '3.10'
|
33
33
|
type: :development
|
34
34
|
prerelease: false
|
35
|
-
version_requirements: *
|
35
|
+
version_requirements: *83466720
|
36
36
|
- !ruby/object:Gem::Dependency
|
37
37
|
name: hoe
|
38
|
-
requirement: &
|
38
|
+
requirement: &83466490 !ruby/object:Gem::Requirement
|
39
39
|
none: false
|
40
40
|
requirements:
|
41
41
|
- - ~>
|
@@ -43,7 +43,7 @@ dependencies:
|
|
43
43
|
version: '3.3'
|
44
44
|
type: :development
|
45
45
|
prerelease: false
|
46
|
-
version_requirements: *
|
46
|
+
version_requirements: *83466490
|
47
47
|
description: textutils - Text Filters, Helpers, Readers and More
|
48
48
|
email: webslideshow@googlegroups.com
|
49
49
|
executables: []
|