c# Regex to remove single characters and orphaned spaces when input is unknown and can contain multiple words

Question

This is almost similar to this OR condition in Regex and many others close ...

I have an OCR Program that is reading labels off of pictures some of the bits cause some small errors with single characters in odd places but all the labels will have at least 2 letters and any wrong letters will be space padded at least trailing maybe leading

GIVEN :

m Rose
a a m a this test b c z ^ @
k This Bigger k
Great m z
One Big Good Word This IS About AS LRG Possible and good one

DESIRED :

Rose
this test
This Bigger
Great
One Big Good Word This IS About AS LRG Possible and good one

How do I get rid of the odd ball singles in c# I have been trying for hours with single and multiple Regex.Replace but am getting nowhere

str = Regex.Replace(str2, @"([0-9a-zA-Z]{1}) ([0-9a-zA-Z]{2,100})?","$2", RegexOptions.Multiline);

gets close but truncates a letter and space between words so "Open Hours" is "OpeHours" happy to replace with spaces then another line to get rid of them ..just not getting the words multiple words out since the lengths and occurrences are random and my regex skill is average at best, just seems there should be a one liner for this without having to split and reassemble.

...after regex for a reason.. I know could loop through the string and look for spaces before and after or other string voodoo ways ...

The reason is : i run across this all the time and have never come up with a good solution that is a single regex , multiple passes sure. — mxdog, Jul 18 '22 at 22:21
Perhaps match with `\w{2,}(?:\s+\w{2,})*` https://regex101.com/r/MGLDIE/1 — The fourth bird, Jul 18 '22 at 23:18
@Flydog57 don't care these are labels not sentences...just need the strays gone — mxdog, Jul 20 '22 at 22:30

score 1 · Accepted Answer · answered Jul 18 '22 at 23:44

1

try this .(?= )|(?<= ). |^. | .$:

str = Regex.Replace(str2, @" .(?= )|(?<= ). |^. | .$","", RegexOptions.Multiline);

answered Jul 18 '22 at 23:44

yassinMi

707
2
9

This works too except for some odd reason left a leading and trailing space on Great .. going to accept this as answer it is a little simpler then Wiktor's. Thanks much now i feel silly how easy it looks for the time i put in .. – mxdog Jul 20 '22 at 23:14

score 1 · Answer 2 · answered Jul 19 '22 at 13:30

1

You can use

text = Regex.Replace(text, @"(?:\b\w\b|[^\w\r\n])+", " ")

See the regex demo.

Details:

(?:\b\w\b|[^\w\r\n])+ - one or more sequences of
- \b\w\b - a single word char word
- | - or
- [^\w\r\n] - any char other than a word char, or CR / LF.

answered Jul 19 '22 at 13:30

Wiktor Stribiżew

607,720
39
448
563

This mostly works but leave leading and trailing spaces on the word groups, where there are some, but like said those i can deal with thanks ... – mxdog Jul 20 '22 at 23:11

c# Regex to remove single characters and orphaned spaces when input is unknown and can contain multiple words

2 Answers2