Numbers get in the way more often than you would think. A scraped list of product names with item codes stuck on the front. A transcript with timestamps before every line. A column of “1. Apples”, “2. Pears” where the numbering has to go before the list is reused. In all of these, the words are the data and the digits are noise.
Paste the text above and the numbers are gone from the result.
Two modes
“Whole numbers” removes a number as a unit, including the commas and decimal point inside it. So $1,299.50 becomes just $, not $,., and 3.14159 disappears completely. This is the right choice nearly every time.
“Single digits only” removes each digit separately and leaves any separators behind. The same price becomes $,.. It exists for odd cases where you want to strip digits out of the middle of words, for instance turning abc123def into abcdef without caring what the digits formed.
Either way, numbers inside words go. A name like Room 12B becomes Room B, and COVID19 becomes COVID.
What stays behind
Currency signs, percent signs, slashes in dates, and the hyphens in a phone number are all left in place because they are not digits. A date like 12/03/2026 ends up as //. If that mess matters, take the punctuation out with remove punctuation in a second pass, or use remove special characters, which keeps only letters if you ask it to.
The “tidy double spaces” option closes the gaps left where numbers used to be. Without it, “Order 66 shipped” becomes “Order shipped” with two spaces in the middle.
Number words are not numbers
“Five” and “twenty” stay, because they are words. Going the other way, if you need digits turned into words, number to words does that, and if you need to keep only the numbers and throw away everything else, the number extractor is the opposite of this tool and also adds them up for you.
One use that surprises people
Cleaning text before a duplicate line check. If a list has entries like “Smith 1”, “Smith 2”, “Smith 3”, they all look different. Strip the numbers, tidy the spaces, and the duplicates line up.
A worked example
Say you have scraped a product list where every entry carries a code: “Blue ceramic mug 4471”, “Steel water bottle 0932”, “Oak coaster set 2210”. Removing numbers with the tidy option on gives “Blue ceramic mug”, “Steel water bottle” and “Oak coaster set”, ready to compare against another catalogue. The same approach strips list numbering such as “1. Apples” and “2. Pears”, although a line that begins with a year or a quantity, like “3 eggs”, loses that figure as well, so look over lines where the number was part of the meaning.