nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

treebuilder as_HTML method - quoting numeric attrs

OliStreek

I'm using HTML::TreeBuilder to parse and modify html tables. When I'm finished, I call the as_HTML() method to write out the code. However, numeric attributes are written out without quotes, for example:
<table border=0 cellpadding=4 cellspacing=0 summary="Dimensions">
when I want:
<table border="0" cellpadding="4" cellspacing="0" summary="Dimensions">

Does anyone know of a way to instruct the as_HTML method to quote numeric values as well as non numeric? If the HTML was always strictly XHTML compliant, I'd use an XML parser, but it isn't. I could use regexes afterwards, but I'd rather do all the work with the parser if possible.

Thanks,
Oli

Find more posts tagged with

Comments

Migrateduser

Could you post a section of the code that calls the TreeBuilder? I'm particularly interested to see how the TB is constructed.

Dave

Current Environment(s):
(1) TS 6.1 SP1 on W2K3
(2) TS 6.1 SP1 on W2K
(3) TS 5.5.2 SP2 on Win2K

OliStreek

Here it is...

my $root;
...
$root = HTML::TreeBuilder->new();

$root->parse("<html>$string</html>");
$root->eof();

I had tried
$root = HTML::TreeBuilder->new('quote_attr' => 1);
but no joy.

Cheers

Migrateduser

Uhh, okay, sorry. Where is $string coming from? In other words, what's the value of $string going into that function?

Dave

Current Environment(s):
(1) TS 6.1 SP1 on W2K3
(2) TS 6.1 SP1 on W2K
(3) TS 5.5.2 SP2 on Win2K

Dwayne

I'm guessing that HTML::TreeBuilder doesn't support the functionality that you seek. From the HTML::Element perldocs (which HTML::TreeBuilder uses:

$h->starttag() or $h->starttag($entities): Returns a string representing the complete start tag for the elrment. I.e., leading "<", tag name, attributes, and trailing ">". Attributes values that don't consist entirely of digits are surrounded with double-quotes, and appropriate characters are encoded.

No mention of an ability to override that behavior that I can find.

--

Current project: TS 5.5.2/6.1 W2K

Migrateduser

What I was wondering is what if the values were fed in with escaped quotes? For instance,



$string = qq^<table border=\"0\">^;

... etc., etc.

Would TB interpret the quotes as part fo the attribute then?

Dave

P.S. THANKS for the pointer on associating the JS event with added replicants. That's the post I was looking for.

Current Environment(s):
(1) TS 6.1 SP1 on W2K3
(2) TS 6.1 SP1 on W2K
(3) TS 5.5.2 SP2 on Win2K

Dwayne

Even if it did, the resulting HTML would probably be spit out something like this:

<table border="&quote;0&quote;">

(or whatever the HTML entity is for double quote. I can't remember right now)

--

Current project: TS 5.5.2/6.1 W2K

OliStreek

Dave,
Sorry, I meant to mention that - $string is the html fragment from a VisualFormat field that contains the table html. And I also tried what you suggested with the escaped quotes, but it just escapes the whole string including the \" characters.

Dwayne,
Thanks for your pointer on the Element, I think you must be right - this just isn't something I can do this way. I am surprised there's no way to override this behavior though.

Oh well, I guess I either have to use regexes or look at using the HTML: Smiley Tongue

rettyPrinter. Before I decide which to use, has anyone had a bad experience of the PrettyPrinter? I think this solution is cleaner than using regexes but obviously it will reformat all the html, which means more possible side effects (and more testing for me).

Migrateduser

I'm just curious -- what do you have against regexes? I love them! What's your concern about using them to parse your entire HTML string?

Dave

Current Environment(s):
(1) TS 6.1 SP1 on W2K3
(2) TS 6.1 SP1 on W2K
(3) TS 5.5.2 SP2 on Win2K

Dwayne

Well, the as_HTML method is a convenience function. You could "roll your own" using methods of HTML::Element. More work, of course, but do-able. Specifically, look at tag(), content_list() and all_attr()

--

Current project: TS 5.5.2/6.1 W2K

OliStreek

I don't have anything against regexes as such! This method that performed all our processing on these tables used to be written entirely with regexes, but it got very complicated. We were adding different classes to table rows and cells based on their position in the table, then adding id and header tags, etc. It's all possible, but it got to the point where reading it into a tree and manipulating it that way was easier to understand and modify.

So given that we've made this change, it would be nice to be able to implement this last bit in a consistent manner!

Johnny

are you playing with the attributes usingTreeBuilder?

because I noticed the same problem when changing attributes using numbers.

This is how I got around it:



$table->attr('cellpadding', " 8 ");

note the spaces between the number and the quotes.
It writes it directly out like this; I haven't seen any browsers complain about it.



<table cellpadding=" 8 " >

John Cuiuli