Maybe to make it easier for interested persons to check out Rick's incubator
regex.cls, three links:
* Java "regular expression" introduction/tutorial as regex.cls started out
with it in mind, also a
quite nice tutorial:
<https://docs.oracle.com/javase/tutorial/essential/regex/>
* The regex.cls package and testcases showing how to use it:
<https://sourceforge.net/p/oorexx/code-0/HEAD/tree/incubator/regex/>
o An overview about the regex.cls (structure deduced from directives):
<http://wi.wu.ac.at/rgf/rexx/tmp/docs.tmp/incubator/regex/htmldocs/oorexxdoc.html>
* The ooRexx test unit framework, if anybody wants to run the regex test
cases:
<https://sourceforge.net/p/oorexx/code-0/HEAD/tree/test/trunk/>
---rony
On 08.06.2020 14:24, Rick McGuire wrote:
>
>
> On Mon, Jun 8, 2020 at 8:06 AM Rony G. Flatscher <rony.flatsc...@wu.ac.at
> <mailto:rony.flatsc...@wu.ac.at>> wrote:
>
> On 08.06.2020 13:39, Rick McGuire wrote:
>> My incubator version has named captures, and also has a new feature I've
>> not seen other regex
>> packages, reusable named patterns. For example, one of the standard
>> patterns is a URL:
>> pattern that has named elements like protocol, port, domain name, etc.
>> Rather than having to
>> copy and paste one of these patterns into your own, you can refer to
>> them by name. I'm a bit
>> stalled on the design of user pattern libraries at the moment, but this
>> is a very powerful
>> concept.
>>
>> While a regex can be incorporated into the PARSE instruction, it would
>> only be in a very
>> simplistic way. My implementation has a parsing context class that
>> supports everything the
>> PARSE instruction does and quite a bit more beyond what PARSE can do.
>
> Went into incubator/regex and loaded the programs: *wow*!
>
> Overwhelming from the functionality that is there already (currently
> there are 87 classes in
> regex.cls)!!
>
> Would it be already possible to parse xml elements (supplying the tag
> name and having the
> parser parse everything to the matching end tag, independently whether
> the tag name gets
> recursively used) with a single pattern?
>
> If such a thing is possible with a regex, then it should be possible with
> this version.
>
> What would you suggest to study first in order to understand what you
> have created so far?
>
> No docs written for this at all. This was originally based on the Java
> version, so it follows that
> for most of the patterns. Probably the best place to start would be to look
> at what the test cases
> are doing.
>
> Rick
>
>
>
> ---rony
>
>
>>
>> On Mon, Jun 8, 2020 at 7:16 AM Rony G. Flatscher <rony.flatsc...@wu.ac.at
>> <mailto:rony.flatsc...@wu.ac.at>> wrote:
>>
>> Just an interesting mail thread from a mainframe Rexx mailing list.
>>
>> Are there any ongoing projects in this area? I seem to remember that
>> Rick started to work
>> on a
>> regular expression package with name captures, and independently,
>> Erich mentioned to make the
>> regular expression features of the latest C++ standard available
>> (not sure whether there
>> would be
>> something like name captures available).
>>
>> There was an idea of incorporating such a feature also into the
>> PARSE keyword statement,
>> if some
>> (human-oriented) easy syntax would be suggested for it (which is
>> dependent on the feature-set
>> implemented in such an extension library, of course, so a little bit
>> of a chicken-egg
>> problem there).
>>
>> ---rony
>>
>>
>> -------- Forwarded Message --------
>> Subject: Re: PCRE2 10.35 for z/OS
>> Date: 7 Jun 2020 07:05:37 -0700
>> From: Paul Gilmartin <paulgboul...@aim.com
>> <mailto:paulgboul...@aim.com>>
>> Reply-To: TSO REXX Discussion List <tso-r...@vm.marist.edu
>> <mailto:tso-r...@vm.marist.edu>>
>> Organization: None
>> Newsgroups: bit.listserv.tsorexx
>> References: <1271218775.429747.1591498456491....@mail.yahoo.com
>> <mailto:1271218775.429747.1591498456491....@mail.yahoo.com>>
>> <1271218775.429747.1591498456...@mail.yahoo.com
>> <mailto:1271218775.429747.1591498456...@mail.yahoo.com>>
>>
>> <bl0pr05mb5156c5ee0273a84de9f867f599...@bl0pr05mb5156.namprd05.prod.outlook.com
>>
>> <mailto:bl0pr05mb5156c5ee0273a84de9f867f599...@bl0pr05mb5156.namprd05.prod.outlook.com>>
>>
>> On 2020-06-07, at 04:56:19, Seymour J Metz wrote:
>> >
>> > For those not familiar with Perl, its regex syntax includes named
>> captures, which
>> properly used can make it much easier to write maintainable code.
>> PCRE is much more
>> powerful the the regexen in, e.g., oorexx.
>> >
>> Ooh! Sorta like SNOBOL4! Does the Rexx API provide named captures
>> to Rexx variables (another case of quoting symbol names?)
>>
>> > ________________________________________
>> > From: Ze'ev Atlas
>> > Sent: Saturday, June 6, 2020 10:54 PM
>> >
>> > Hi AllI would like to encourage all of you to use the PCRE2
>> library with the new and
>> improved Rexx API. In some previous conversation somebody had
>> mentioned the package (in
>> a previous release and much more primitive state, and pointed to
>> some issues. The
>> current release is an up to date version and, basically, perfected
>> the Rexx API. Again.
>> please try it.As opposed to other mentioned products which are not
>> Perl compatible, this
>> one is Perl Compatible and Perl has the golden standard of Regular
>> Expression usage.
>> > Ze'ev Atlas
>>
>> -- gil
>>
_______________________________________________
Oorexx-devel mailing list
Oorexx-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/oorexx-devel
