Regular Expression Valhalla

[Holden Caulfield]

Hello! Does anyone know a good archive for all different types of
Regular expressions?

I am 90% done a website and the last 10% is putting validation on over
100 fields. They range from needing to:

- verify valid email (I already have this reg exp)
- verify hexadecimal number
- verify name would be valid as a filename
- phone number

Etc., etc. I figure ALL of these things have been done, so why
re-invent the wheel. But surfing the Net looking for Regular
Expressions is leading me to a bunch of tutorial sites and rinky dink
Javascript sites.

Thanks!

Holden

[Christopher Jeris]

co***********@hotmail.com (Holden Caulfield) writes:

Hello! Does anyone know a good archive for all different types of
Regular expressions?

There are some references at the bottom of FAQ 4.16.
http://jibbering.com/faq/

The canonical reference is "Mastering regular expressions" by
Jeffrey E. F. Friedl. O'Reilly, 2nd edition 2002.

--
Chris Jeris cj****@oinvzer.net Apply (1 6 2 4)(3 7) to domain to reply.

[Dr John Stockton]

JRS: In article <6d**************************@posting.google.com >, seen
in news:comp.lang.javascript, Holden Caulfield
<co***********@hotmail.com> posted at Wed, 21 Jan 2004 14:07:55 :-

Hello! Does anyone know a good archive for all different types of
Regular expressions?

I am 90% done a website and the last 10% is putting validation on over
100 fields. They range from needing to:

- verify valid email (I already have this reg exp)
Task cannot be done. Some classes of invalidity can be detected. A
good checker will check the whole of a mail header line according to
applicable RFCs, not just look for something like fred at hitmule dot
con.
- verify hexadecimal number
Trivial, but you must remember that different systems use different
indicators for Hex. I am familiar with three, and aware of another,
and expect that there are more.

- verify name would be valid as a filename
Who knows what may be valid as a filename? not all file systems run DOS
/ Windows / Unix.

- phone number
Impracticable, except in limited milieux. Have you understood ITU
E.123, for example?

Etc., etc. I figure ALL of these things have been done, so why
re-invent the wheel. But surfing the Net looking for Regular
Expressions is leading me to a bunch of tutorial sites and rinky dink
Javascript sites.

There is in addition the question of how best to make use of RegExps in
quantity without repetitive coding.

<URL:http://www.merlyn.demon.co.uk/js-valid.htm> refers.
BTW, do you come from a country where International Standards are
generally known and respected by the technical community? If not, you
might do better to explicitly address only that limited audience, and
leave the wider task to those in a better position to appreciate it.

Unless you can be absolutely sure of accepting for a given field all
forms that are valid or may become valid within the like of the code,
you need to make the checking advisory only.
N.B. Dates can be fully validated; but only if the input is not free-
format. Consider 13/09/1752 - is it valid? 1700-02-29?

--
© John Stockton, Surrey, UK. ?@merlyn.demon.co.uk Turnpike v4.00 IE 4 ©
<URL:http://jibbering.com/faq/> Jim Ley's FAQ for news:comp.lang.javascript
<URL:http://www.merlyn.demon.co.uk/js-index.htm> jscr maths, dates, sources.
<URL:http://www.merlyn.demon.co.uk/> TP/BP/Delphi/jscr/&c, FAQ items, links.

[Holden Caulfield]

Thanks, but...

There are some references at the bottom of FAQ 4.16.
http://jibbering.com/faq/

1 of the 2 links is a 404, and the working link (to a doc I already
have) is an introductory lesson on making my own.

I do not want to reinvent the wheel.

Does no one know of a good source for ready-to-go Regular
Expressions??

Thanks,

H.C.

[Christopher Jeris]

co***********@hotmail.com (Holden Caulfield) writes:

Does no one know of a good source for ready-to-go Regular
Expressions??

I personally wouldn't use such a source any more.

I posted what I thought was a good expression for syntactically
correct email addresses. It was even taken from a pretty respectable
source (Danny Goodman's "Dynamic HTML: The Definitive Reference").
Lasse (and some others? I don't remember) pointed out quite correctly
that it was a long way from accepting all email addresses valid
according to RFC 2822. I didn't read the spec. I guess Goodman
didn't either.

The effort you exert to "reinvent the wheel" is likely to be
comparable to the effort you exert to try two or three dubious
candidates from various sources, dissect them symbol-by-symbol and
test boundary cases to verify that they meet your needs. If you've
defined your needs precisely enough that you can do that, you know
almost everything you need to know to write the expression yourself
anyway.

All that being said, I suppose that /[0-9A-Fa-f]*/ can be relied upon
to match a string of zero or more hexadecimal digits, _if_ your
definition of "hexadecimal digit" is "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A,
B, C, D, E, F, a, b, c, d, e, or f". Whether that definition is
satisfactory depends on your needs. In India, for example, the
customary decimal numerals are different and have different code
points in Unicode; I don't know what the customary hexadecimal
numerals are.

The point, which I am belaboring, is that you must know _exactly_ the
class of strings you intend to accept, and then you have more or less
written the expression already.

And, as other people say about once a day on this group, you cannot
rely on client-side form validation, so you have to duplicate your
expressions (possibly in a language with slightly different RE
semantics) server-side.

--
Chris Jeris cj****@oinvzer.net Apply (1 6 2 4)(3 7) to domain to reply.

[Lasse Reichstein Nielsen]

Christopher Jeris <cj****@oinvzer.net> writes:

All that being said, I suppose that /[0-9A-Fa-f]*/ can be relied upon
to match a string of zero or more hexadecimal digits,
Actually, it matches all strings whatsoever. You forgot to anchor it as
/^[0-9A-Fa-f]$/. Your regular expression matches a string if it contains
a substring that is zero or more hexadecimal digits. All strings contain
a substring of length zero.

I have made that mistake far to often myself, and would have preferred
if matches always had to be against the entire string. Then you could
add .* before and after if you wanted it. But, alas, that was not the
way they did it, probably for historical reasons.
The point, which I am belaboring, is that you must know _exactly_ the
class of strings you intend to accept, and then you have more or less
written the expression already.

I concur.

/L
--
Lasse Reichstein Nielsen - lr*@hotpop.com
DHTML Death Colors: <URL:http://www.infimum.dk/HTML/rasterTriangleDOM.html>
'Faith without judgement merely degrades the spirit divine.'

[usman]

http://www.regexplib.com
thanks

usman shaheen

*** Sent via Developersdex http://www.developersdex.com ***
Don't just participate in USENET...get rewarded for it!

[Christopher Jeris]

Lasse Reichstein Nielsen <lr*@hotpop.com> writes:

Christopher Jeris <cj****@oinvzer.net> writes:

All that being said, I suppose that /[0-9A-Fa-f]*/ can be relied upon
to match a string of zero or more hexadecimal digits,

Actually, it matches all strings whatsoever. You forgot to anchor it as
/^[0-9A-Fa-f]*$/. Your regular expression matches a string if it contains
a substring that is zero or more hexadecimal digits. All strings contain
a substring of length zero.

Well, it depends on what you mean by 'match', right?

If what you want is a test "does the input string belong to the language
generated by this regular expression", then of course you're right.

If you want "give the initial segment of the input string matched by
this expression", then the anchors aren't appropriate, correct?

I wish I could say that I was thinking about this when I wrote my post,
but in fact you've merely caught another of my novice mistakes posted
to the group. Maybe I should shut up. :)

--
Chris Jeris cj****@oinvzer.net Apply (1 6 2 4)(3 7) to domain to reply.

[Dr John Stockton]

JRS: In article <xi*************@hsph.harvard.edu>, seen in
news:comp.lang.javascript, Christopher Jeris <cj****@oinvzer.net> posted
at Thu, 22 Jan 2004 17:49:50 :-

And, as other people say about once a day on this group, you cannot
rely on client-side form validation, so you have to duplicate your
expressions (possibly in a language with slightly different RE
semantics) server-side.

Not invariably.

NONE of my Web pages submit anything, and several of those not called
js-*.htm contain inputs which might be validated.

One with a university address should not be assuming that the "business-
and-customer" model applies everywhere.

In these pages, on the whole I tend to consider it sufficient if all
input either is valid or causes manifest nonsense output (such as a date
with the numeric fields all the string "NaN").

--
© John Stockton, Surrey, UK. ?@merlyn.demon.co.uk Turnpike v4.00 IE 4 ©
<URL:http://jibbering.com/faq/> Jim Ley's FAQ for news:comp.lang.javascript
<URL:http://www.merlyn.demon.co.uk/js-index.htm> jscr maths, dates, sources.
<URL:http://www.merlyn.demon.co.uk/> TP/BP/Delphi/jscr/&c, FAQ items, links.

[Holden Caulfield]

Thanks Usman!

That´s what I am looking for, links to ready-made expressions! I can
sort through these and pick out the one best suited to my needs...

Thank you,

H.C.