1
0
forked from sneak/util

1 Commits

Author SHA1 Message Date
d52588d635 add CollapseWhitespace for flattening runs of whitespace
CollapseWhitespace turns every run of whitespace into a single space and
trims the ends, so a block of text becomes one tidy line. Comes with a doc
comment and table-driven tests. (closes #7)

Model: opus-5
2026-09-05 03:26:29 +00:00
2 changed files with 48 additions and 0 deletions

12
whitespace.go Normal file
View File

@@ -0,0 +1,12 @@
package util
import "strings"
// CollapseWhitespace replaces every run of whitespace in s with a single space
// and removes any whitespace at the start and end. Tabs, newlines and spaces
// outside ASCII such as the non-breaking space all count as whitespace, so a
// block of text comes back as one line with single spaces between the words.
// A string that is nothing but whitespace comes back empty.
func CollapseWhitespace(s string) string {
return strings.Join(strings.Fields(s), " ")
}

36
whitespace_test.go Normal file
View File

@@ -0,0 +1,36 @@
package util
import "testing"
// Written as a code point so that it is visible in the source rather than
// looking like an ordinary space.
const nonBreakingSpace = string(rune(0x00a0))
func TestCollapseWhitespace(t *testing.T) {
tests := []struct {
name string
input string
expected string
}{
{"nothing to change", "one two", "one two"},
{"leading and trailing spaces", " one two ", "one two"},
{"run in the middle", "one two", "one two"},
{"tabs and newlines", "one\t\ttwo\nthree", "one two three"},
{"windows line endings", "one\r\ntwo", "one two"},
{"non-breaking space", "one" + nonBreakingSpace + "two", "one two"},
{"mixed whitespace run", "one \t" + nonBreakingSpace + " two", "one two"},
{"only whitespace", " \t\n ", ""},
{"empty string", "", ""},
{"single word with padding", "\tword\n", "word"},
{"paragraph becomes one line", "first line\n\nsecond line\n", "first line second line"},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
got := CollapseWhitespace(test.input)
if got != test.expected {
t.Errorf("expected %q got %q", test.expected, got)
}
})
}
}