How to regex made up of numbers surrounded by alphabets & special chars

[Jack Bush]

Hi All,

I need your advice on coming up with a regex that matches the following list of numbers potentially surrounded by a combination of alphabets and special characters:

1
1&2/4
106&108
1-3
1-6/247
1.06/38-52
1.08/12
1/1
1/1-3
1/13-15
1/1a
106a/1
10f/15-19
11a
12/49b
12a/19a
159-161
15b/1-3
24/K
39(2)
403/3-7a
409/1a
42/A
48-48a
77c/1-3
?/12a
?/96-100
A22/335
Cg10/3
L07/10-16
Pg05/11-15
S/2-4

Below are the regexes that I have come up with:

[\w\/\-.?()&] – worked but doesn’t insist having numbers in between.
\w*[\/\-.?()&]\d+?[\/\-.?()&]\w* - does insist on having numbers in between the alphabets and special characters but does not match every line.
\w*\W*\d+?\W*\w* - does insist on having numbers in between the string and matches every line.

Can someone confirm whether the last regex is correct? Otherwise, what would the correct regex be?

This is the address number of a public property sales report.

Thanks a lot,

Jack

[fred rosenberger]

haven't really looked at your examples, but I have to ask one question...

Is it a requirement that you find all the possible matches in one pass? Might it be easier to have several regexes, each of which match a (possibly distinct) subset of what you want, then you combine them later?

[Harsha Smith]

This may work. Please try.

[()0-9&A-Za-z/.&?-]+

[Jack Bush]

Thanks to both Fred & Harsha for your suggestions.

Yes, it is a requirement to identify that it is a property number from its sequence (first number) as well as ensuring that it consists of at least one digit. Below is an example of what this input data is made up of:

“South Yarra 5/106 Toorak Rd W 2 br u $566,263 SP HS South Yarra”

The first 2 words is the district name which can start from 1 – 3 letters or more. As a result, it is difficult to determine which position will the property number be at, and whether the value found is a property number (first number) or number of bed room (second number). Would a lookahead regex such as one starting from the last word of the district name (i.e. [A-Za-z]) looking for the next string that is made up of digits and potentially other characters? I couldn’t come up with a lookaround solution still for this.

[()0-9&A-Za-z/.&?-]+ does work but it also picked up words with letters only such as district (before the number) and Street (after the number) names, so it is not correct. I need a regex that make digits mandatory while the rest are optional.

Any idea on how this could be achieved?

Thanks again,

Jack

[Winston Gutkowski]

Jack Bush wrote:Yes, it is a requirement to identify that it is a property number from its sequence (first number) as well as ensuring that it consists of at least one digit. Below is an example of what this input data is made up of:

“South Yarra 5/106 Toorak Rd W 2 br u $566,263 SP HS South Yarra”

The first 2 words is the district name which can start from 1 – 3 letters or more...

Actually, it isn't. From what I can see from your original data (the one you posted in your last thread), the district ends with the first word
that contains a number as it's first character (which seems, pretty consistently, to be the start of an address).

That word seems to contain some combination of the following:
1. A house/building number, which may be suffixed with a letter (eg, '43a').
2. A range of building numbers, separated by a hyphen (-).
3. A suite or apartment number + plus a building number (or range), separated by a forward slash (/).

I reckon Fred's right. Once you've identified the first word of the address, I think you might be better off breaking down the possibilities,
maybe with String.split(), and then using individual regexes to validate/extract the actual numbers.

BTW, as far as I can see, the document is also consistent about having all that "numeric stuff" in a single word (ie, no spaces),
so if you don't need to actually parse the contents, you could simply use "[0-9][^ ]*" to get the whole word.

Winston

[Harsha Smith]

I need a regex that make digits mandatory while the rest are optional.

This the regex

((?=.*\\d)([()&A-Za-z/.?-]*).*)

[nick kaushik]

& while we are at it could you guys also tell me regex for fishing out strings from text such that it doesn't contain any space,newline,tabs,etc & the full stops.....
i know split method fish out line feeds & whitespace with "//s" but what about the dot.
eg.
i am doing Java.
output should be:
i
am
doing
Java

[Jack Bush]

Hi All,

Apologies for the late reply but I have applied Fred & Winston’s suggestion of splitting each word out and match them with their respective regexes in sequences, which is working out finally. I am still at an experimental stage but it is looking good so far.

Also thank you to Hersha for kept coming up with great regex gems.

I will close this thread now since the question has been answered.

Thank you again,
Jack