in reply to Arrays with CGI perl help

SuicideJunkie and ikegami have provided answers to your nominal question but even with their corrections, I'm not quite sure why your test script produces any content renderable in the browser, since it lacks several elements required by html standards, notably <html>, <head>, <title>, </title>,</head> and <body>:

The CGI.pm doc (available on your local machine at perldoc CGI.pm) provides a simple example of (one way) to do so (and throws in a recommended-but-not-required DTD for free):

use CGI qw/:standard/;
print header,
start_html('A Simple Example'),
...

which will provide, for rendering:

Content-Type: text/html; charset=ISO-8859-1

<!DOCTYPE html
        PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
         "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-US" xml:lang="en-U
+S">
<head>
<title>A Simple Example</title>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1
+" />
</head>
<body>

[download]

Note that the print header, took care of the required Content-Type: text/html; charset=ISO-8859-1-followed-by-two-newlines. Since you didn't use print header your manual print of the Content-Type is NOT redundant (if interested in the issue, you can read numerous nodes here where the problem was coding in such a way as to emit Content-Type... twice).

If you chose to avoid print header, you MUST supply the <html>, <head>, <title>, title-content, </title>, </head><c> and <c><body>. (IMO, the best reason to eschew print header,... is to allow you to select another DTD and add additional meta data.)

Update: edited html to fix cut'n'paste error which doubled "<!DOCTYPE html" - thanks ikegami!

Replies are listed 'Best First'.
Re^2: Arrays with CGI perl help (More fundamental issues)
by DarthWavy (Novice) on Jun 25, 2009 at 10:21 UTC
    Browsers will attempt to render even if the code is not to standard, in fact no actual html tags are needed whatsoever to get text to display.
[reply]
      Indeed. In fact, early versions of Internet Explorer would render plain text or even binary data as HTML if the first few bytes appeared to contain a '<' character, no matter what content-type the server specified :-|

      -- Time flies when you don't know what you're doing
[reply]
Re^2: Arrays with CGI perl help (More fundamental issues)
by ikegami (Patriarch) on Jun 25, 2009 at 18:13 UTC

    You have been greatly misinformed. Under strict HTML4,

    • You failed mention that the DOCTYPE decleration is required.
    • No, the <html> tag is optional.
    • No, the <head> tag is optional.
    • While the title element is required (and its start and end tags are required), browsers cope with its absence.
    • No, </head> tag is optional.
    • No, <body> tag is optional.

    Of course, the OP gave no indication he was interested in issuing strict HTML4.

[reply]
[d/l]
[select]

      Given your cite, I stand corrected on my statement re the DOCTYPE being optional; I can't even recall which 'authority' put that notion in my head.

      However, I would continue to hold that the <head>...</head> is required (*1 qualifier below) because <title>...</title> is required and §7.4 of w3c's html 4 documentation says:

      "Every HTML document must have a TITLE element in the HEAD section." (emphasis in the original at 7.4.3)

      *1 And yes, that directly contradicts the w3c cite you provided ...suggesting that:

      1. w3c needs to get its act together
        or
      2. I'm mis-reading something, or failing to distinguish properly between its "normative" and "informative" documentation.

      So, while I'm uncertain which citation governs here, I think each of us has an arguably sound basis for our assertions (at least re <head>...<head>).

      Moving on to into calculations of how many angels can dance on the head of a pin...

      §7.3, "The HTML element," appears to me to be directive and thus to offer some support for my notion that the <html ...> is required:

      "After document type declaration, the remainder of an HTML document is contained by the HTML element."

      But, I must also concede that the quote above does not quite offer explicit contradiction of the earlier indication that the <html> element is optional... and I cannot re-find the (vividly/spuriously?) remembered 'authority' which -- as I recall it -- asserted that tags like <p>, <h#>, <ol>, <table>, <div> and company can appear only inside a <body>...</body> pair.

      That said, I belatedly noticed (and only after your msg re the doubled <!DOCTYPE> for which, thanks!) that CGI.pm produced an XHTML <!DOCTYPE>, making the example pretty much irrelevant to the thrust of my comment ((or irrelevant, confusing and stupid: ie, I forgot to include -no_xhtml in the use CGI line).

[reply]
[d/l]
[select]

        I would continue to hold that the <head>...</head> is required (*1 qualifier below) because <title>...</title>

        You seem to be confusing elements and tags.

        A HEAD element must be present in an HTML element, and the HEAD element must be the HTML element's first child. A HEAD element will be created in the DOM, and any content required by the HEAD element must be present as well. I think you agree with this.

        The representation of the element in the document is a separate matter. Neither tags of the HEAD element (<head> and </head>) are required. Due to the requirement of the presence of a HEAD element, the element is still present even if the tags are omitted.

        Just like the closing of the P element is implied in <div><p></div>, the opening and closing of the HEAD element is implied in <html><title>Title</title><body>

        Update: Improved wording.

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom