[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Improvements to gnuapl

From: Dr . Jürgen Sauermann
Subject: Re: Improvements to gnuapl
Date: Tue, 23 Feb 2021 13:01:38 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.10.0

Hi Blake,

You're correct. Another problem in HTML is unquoted attribute values in HTML tags.

I should have said "One can use ⎕XML for decoding HTML pages and the like as long as
they obey the fundamental XML encoding rules".

I believe it would be possible to make ⎕XML tolerate some of these HTML quirks,
but I wonder if it is worth the effort.

Best Regards,

On 2/22/21 10:11 PM, Blake McBride wrote:
Some of those "optional" end tags are not optional at all.  It's not HTML if it's there.  For example:

<br></br>    is not HTML.


On Mon, Feb 22, 2021 at 1:21 PM Dr. Jürgen Sauermann <mail@jürgen-sauermann.de <mailto:mail@j%C3%BCrgen-sauermann.de>> wrote:


    as far as I understand it, HTML has almost the same format as XML
    (the main difference being optional end tags in
    HTML which are mandatory in XML. I would assume that ⎕XML can do
    the decoding of common web interfaces
    like the REST API or other XML based queries quite well. Fetching
    of the data can be done with ⎕FIO[32 ff.] so
    the combination of them should almost do the job.

    Best Regards,

    On 2/22/21 4:13 PM, Elias Mårtenson wrote:
    This could be quite useful when collecting data from a web site.
    For example, pull in a table of numbers from a Wikipedia page.
    Google Docs has this feature already and it can be quite useful.


    On Mon, 22 Feb 2021 at 22:26, Chris Moller <moller@mollerware.com
    <mailto:moller@mollerware.com>> wrote:

        Sounds like another native function!  :-)

        Maybe after I finish my current project...

        On 2/22/21 5:26 AM, Hans-Peter Sorge wrote:

        I would modify the data model and/or process graph or use an
        adequate programming language.
        In my opinion, having to rely on data content to control
        program flow is 'costly'.
        (My be one reason too, that APL has no language specific
        regular expressions).

        My highest priority for APL would be the mapping between an
        apl name and a file,
        directory, a db-table, a spread sheet or an editor instance.

        APL was designed to contain code and data in a 'closed'
        Those days data entry was done by human nature - into the
        work space.
        Nowadays I get the data very likely from somewhere outside
        of the workspace.
        ⍎ ')host' and piping are already a big help here.

        But for example analyzing a web page, that is being done
        faster in python.
        Having a proper infrastructure in APL, like
        *page ← ⎕curl '...url...' **
        **page['head';'link' ] *
        could return all link tags. - just dreaming:-)

        However - please no if/then/else

        Best Regards

        Am 20.02.21 um 19:59 schrieb Christian Robert:
        well I saw the new thrends aka Quad-XML, Quad-JSON,
        Quad-FFT and so on

        but I think thoses will never be used in real life or quite

        I really think that Juergen should be looking at


        :for var :in array

        :while condition:

        :until condition

        this will eases newcommers to the language.

        I know that APL goal is to do a whole "program" in one or
        two lines of code...
        but the language must accomodate newcommers.

        I asked for that several years ago (may me 8 or 10 years)

        Juergen ansewered at that time "this can be done" but I
        wont yet

        well my principal next improvements wish list is

        my real though,


reply via email to

[Prev in Thread] Current Thread [Next in Thread]