Eliminate an element used throughout document, but retain text it's wrapped in

Question

Hi,

We're trying to eliminate an element used throughout our document in Arbortext, but retain the text it's wrapped in.

The issue is nearly resolved following the solution provided in this topic:

Solved: Re: Remove <P> element from a table - PTC Community

This solution works for our issue, but only for tags in tables. Our target tag could have any parent. I have tried modifying this line...

if (oid_name(oid_parent($arr[$i])) == "entry")

...replacing "entry", but had little luck.

Here's the script as it currently exists:

function remove_text_tags()
{
local oid = oid_first_tag()
local cnt
local i = 1

cnt = oid_find_children(oid, $arr, "text")
for ($i=1; $i<=$cnt; $i+=1)
{
if (oid_name(oid_parent($arr[$i])) == "entry")
{
goto_oid($arr[$i],-1)
dt
}
}
}

Thanks for any help!

Jeff

ClayHelberg · Accepted Answer

If you don't care what the parent is, you just want to remove them all, then why not simply remove the parent test altogether?function remove_text_tags() { local oid = oid_first_tag() local cnt local i = 1 cnt = oid_find_children(oid, $arr, "text") for ($i=1; $i<=$cnt; $i+=1) { goto_oid($arr[$i],-1) dt }}

GarethOakes · Answer

Hi Jeff,

It looks like you are trying to flatten out the change tracking elements in your document. What you need is a "tree walker" function. That is so common it is even covered in the PTC documentation: https://support.ptc.com/help/arbortext/r8.2.1.0/en/#page/Program%2Fprogrammer_ref%2Ftraversing_and_printing_a_document_structure.html

EDIT: Oh it looks like the getElementsByTagName() might even do the tree walk for you? https://support.ptc.com/help/arbortext/r8.2.1.0/en/#page/Program%2Fprogrammer_ref%2Fusing_getelementsbytagname.html

I believe that searching through all the .acl files that ship with the product will produce more examples of this type of algorithm too. Anywhere it is doing oid_find_children() is probably a good candidate, and especially where that function calls itself recursively.

Basically, the recursive tree walker algorithm for your usage is something like this:

1. Set document root node as context node

2. List the child nodes for the context node

3. For each child node, set it as the context node and go to step #2 (this is the recursive part), unless...

4. ... if only child text nodes remain then check the context node, is it the <text> element with status="changed"?

4.a. if the node is one to remove, then delete it

4.b. if the node doesn't match then leave it be and keep going

I don't have a working copy of Arbortext right now so I can't check further but hopefully that keeps you moving. The problem with the code you pasted is it is not recursive. It only works on one level of child nodes.

Sign up

Please use your PTC eSupport account.

Welcome to the PTC Community

Please use your PTC eSupport account.

Scanning file for viruses.

This file cannot be downloaded