text_clean 0.2.0 → 0.2.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/ext/text_clean/text_clean.cc +21 -9
- data/lib/text_clean/version.rb +1 -1
- data/spec/text_clean_spec.rb +31 -13
- metadata +1 -1
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: bb04963f984de57939786cb051eb1096f04fa416
|
4
|
+
data.tar.gz: 1f811f6b548b3e267f962d327b63bb9df8aef61d
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 7f9e02f59b7900ea3ff595fd0f07a344a70913a9a1f0ee484fe46b02e4459ee1a749fc554c220fe0573563467f96aaa9a2fb7c32fee64bf2b3e106be8cd0e3da
|
7
|
+
data.tar.gz: 77f6299a203f3129269ca29fa817dcbb935eaa2c1a55229eab24927e9ad10dad6b4533f5c3f1222ea2a878dd52fe382a59f07e8b0e82720808fce0f2071da9a5
|
@@ -27,13 +27,13 @@ size_t text_clean_cstr(char* text, long len, char line_sep)
|
|
27
27
|
char* read;
|
28
28
|
char* write = text;
|
29
29
|
uint8_t just_added_space = true, // prevent prefix spaces
|
30
|
-
|
30
|
+
just_added_line_sep = false;
|
31
31
|
for (read = text; read < eos; read++) {
|
32
32
|
char c = *read;
|
33
33
|
if (c >= 'A' && c <= 'Z') {
|
34
34
|
// Change upper case to lowercase
|
35
35
|
c += 32;
|
36
|
-
} else if (c == '\t' || c == ',' || c == '&' || c == '/') {
|
36
|
+
} else if (c == '\t' || c == '_' || c == ',' || c == '&' || c == '/') {
|
37
37
|
// Change inconsequential punctuation to spaces (i.e. all count as whitespace)
|
38
38
|
c = ' ';
|
39
39
|
} else if (c == '?' || c == '!' || c == ':' || c == ';') {
|
@@ -45,11 +45,11 @@ size_t text_clean_cstr(char* text, long len, char line_sep)
|
|
45
45
|
if (c == '-') {
|
46
46
|
// double dash?
|
47
47
|
if (*(read + 1) == '-') {
|
48
|
-
if (!just_added_space) {
|
48
|
+
if (!just_added_space && !just_added_line_sep) {
|
49
49
|
*write++ = ' ';
|
50
50
|
read++;
|
51
51
|
just_added_space = true;
|
52
|
-
|
52
|
+
just_added_line_sep = false;
|
53
53
|
}
|
54
54
|
} else {
|
55
55
|
// scan ahead to see if this hyphen is at the end of the line
|
@@ -68,20 +68,32 @@ size_t text_clean_cstr(char* text, long len, char line_sep)
|
|
68
68
|
}
|
69
69
|
}
|
70
70
|
}
|
71
|
-
} else if (c == '.' && !
|
71
|
+
} else if (c == '.' && !just_added_line_sep) {
|
72
|
+
// look-behind and see if this is an abbreviation
|
73
|
+
if (write - text >= 2) {
|
74
|
+
char a = *(write - 2);
|
75
|
+
char b = *(write - 1);
|
76
|
+
|
77
|
+
// we're just checking for single-letter abbrevs, so see if 2-chars-behind is whitespace
|
78
|
+
if (a == ' ' || a == '.' || a == '\n' || a == '\t') {
|
79
|
+
*write++ = '.';
|
80
|
+
continue;
|
81
|
+
}
|
82
|
+
}
|
83
|
+
|
72
84
|
// erase space before period
|
73
85
|
if (just_added_space) write--;
|
74
86
|
*write++ = line_sep;
|
75
|
-
|
87
|
+
just_added_line_sep = true;
|
76
88
|
just_added_space = false;
|
77
|
-
} else if ((c == ' ' || c == '\n') && !just_added_space && !
|
89
|
+
} else if ((c == ' ' || c == '\n') && !just_added_space && !just_added_line_sep) {
|
78
90
|
*write++ = ' ';
|
79
91
|
just_added_space = true;
|
80
|
-
|
92
|
+
just_added_line_sep = false;
|
81
93
|
} else if (c == '\'' || (c >= 'a' && c <= 'z')) {
|
82
94
|
*write++ = c;
|
83
95
|
just_added_space = false;
|
84
|
-
|
96
|
+
just_added_line_sep = false;
|
85
97
|
}
|
86
98
|
}
|
87
99
|
// erase space at end of text
|
data/lib/text_clean/version.rb
CHANGED
data/spec/text_clean_spec.rb
CHANGED
@@ -19,12 +19,34 @@ describe TextClean do
|
|
19
19
|
clean_eq("is\t it so?", "is it so\n")
|
20
20
|
end
|
21
21
|
|
22
|
+
context "separates words" do
|
23
|
+
it "with double hyphen" do
|
24
|
+
clean_eq("good--do it", "good do it")
|
25
|
+
end
|
26
|
+
|
27
|
+
it "with newlines" do
|
28
|
+
clean_eq("a\nb\nc", "a b c")
|
29
|
+
end
|
30
|
+
|
31
|
+
it "with underscores" do
|
32
|
+
clean_eq("short_title", "short title")
|
33
|
+
end
|
34
|
+
end
|
35
|
+
|
36
|
+
it "compacts whitespace around double hyphen" do
|
37
|
+
clean_eq("good -- do it", "good do it")
|
38
|
+
end
|
39
|
+
|
40
|
+
it "does not add whitespace when double hypen occurs after end of sentence" do
|
41
|
+
clean_eq("ok.--maybe", "ok\nmaybe")
|
42
|
+
end
|
43
|
+
|
22
44
|
it "treats [,/&] as whitespace" do
|
23
45
|
clean_eq("a,bb&cc/d", "a bb cc d")
|
24
46
|
end
|
25
47
|
|
26
48
|
it "treats [;:!?] as sentence separators" do
|
27
|
-
clean_eq("x;y?z!:q", "x
|
49
|
+
clean_eq("x;y?z!:q", "x|y|z|q", "|")
|
28
50
|
end
|
29
51
|
|
30
52
|
it "joins hyphenated words at line end" do
|
@@ -35,18 +57,6 @@ describe TextClean do
|
|
35
57
|
clean_eq("satis- \t \nfaction", "satisfaction")
|
36
58
|
end
|
37
59
|
|
38
|
-
it "treats a double hyphen as word separator" do
|
39
|
-
clean_eq("good--do it", "good do it")
|
40
|
-
end
|
41
|
-
|
42
|
-
it "compacts whitespace around double hyphen" do
|
43
|
-
clean_eq("good -- do it", "good do it")
|
44
|
-
end
|
45
|
-
|
46
|
-
it "treats newlines as word separator" do
|
47
|
-
clean_eq("a\nb\nc", "a b c")
|
48
|
-
end
|
49
|
-
|
50
60
|
it "ignores numbers" do
|
51
61
|
clean_eq("123abc", "abc")
|
52
62
|
end
|
@@ -58,4 +68,12 @@ describe TextClean do
|
|
58
68
|
it "keeps apostrophes" do
|
59
69
|
clean_eq("dad's", "dad's")
|
60
70
|
end
|
71
|
+
|
72
|
+
it "keeps abbreviated single letters" do
|
73
|
+
clean_eq("a mr t. sawyer", "a mr t. sawyer")
|
74
|
+
end
|
75
|
+
|
76
|
+
it "keeps abbreviated single letters that follow other abbreviations without spaces" do
|
77
|
+
clean_eq("a mr t.j. sawyer", "a mr t.j. sawyer")
|
78
|
+
end
|
61
79
|
end
|