most complete xml package for Python?

[metaperl]

Without even checking them all out, I'm thinking the Amara XML Toolkit
must be the most feature-packed. The developers are readily available
on IRC for support and they seem to make regular releases.

As a meld3 user, I have been using ElementTree under the hood, but was
dismayed to find out that you can't find elements by attribute and I
am finding it difficult to remove elements.

So I am out shopping for a new friend to grok XML.

Input welcome.

[Gerard Flanagan]

On Mar 13, 1:43 pm, "metaperl" <[e-mail]:

Without even checking them all out, I'm thinking the Amara XML Toolkit
must be the most feature-packed. The developers are readily available
on IRC for support and they seem to make regular releases.

As a meld3 user, I have been using ElementTree under the hood, but was
dismayed to find out that you can't find elements by attribute and I
am finding it difficult to remove elements.

Hardly in competition with Amara or lxml, but I wrote this extension
to ElementTree which handles attributes to a limited extent:

http://gflanagan.net/site/python/utils/elementfilter/

(Be aware that parts of the filter string are 'eval-ed')

Gerard

[Stefan Behnel]

metaperl wrote:

Without even checking them all out, I'm thinking the Amara XML Toolkit
must be the most feature-packed.

Please check out lxml before you post FUD like this.

http://codespeak.net/lxml/

Regards,
Stefan

[Paul Boddie]

On 14 Mar, 16:15, Stefan Behnel <[e-mail]:

metaperl wrote:

Without even checking them all out, I'm thinking the Amara XML Toolkit
must be the most feature-packed.

Please check out lxml before you post FUD like this.

http://codespeak.net/lxml/

Calm down, Stefan! The inquirer must have got the impression of
Amara's superiority from somewhere; all you need to do is to discover
which resource gave that impression and balance the situation out. In
any case, it's an opinion: there's no fear involved, and the only
uncertainty and doubt are in the inquirer's mind, which I suppose is
why they posted the inquiry.

Paul

[Steve Holden]

Paul Boddie wrote:

On 14 Mar, 16:15, Stefan Behnel <[e-mail]:

>metaperl wrote:

>>Without even checking them all out, I'm thinking the Amara XML Toolkit
must be the most feature-packed.

Please check out lxml before you post FUD like this.

http://codespeak.net/lxml/

Calm down, Stefan! The inquirer must have got the impression of
Amara's superiority from somewhere; all you need to do is to discover
which resource gave that impression and balance the situation out. In
any case, it's an opinion: there's no fear involved, and the only
uncertainty and doubt are in the inquirer's mind, which I suppose is
why they posted the inquiry.

Although the phrase "without even checking them all out" hardly lends
credence to the OP's assertion of Amara's superiority, and does tend to
support a hypothesis involving some ulterior motive (or would if less
ingenuously done).

regards
Steve
--
Steve Holden [phone] [phone]
Holden Web LLC/Ltd http://www.holdenweb.com
Skype: holdenweb http://del.icio.us/steve.holden
Blog of Note: http://holdenweb.blogspot.com
See you at PyCon? http://us.pycon.org/TX2007

[John Machin]

On Mar 15, 3:45 am, Steve Holden <[e-mail]:

Paul Boddie wrote:

On 14 Mar, 16:15, Stefan Behnel <[e-mail]:

metaperl wrote:
Without even checking them all out, I'm thinking the Amara XML Toolkit
must be the most feature-packed.
Please check out lxml before you post FUD like this.

>http://codespeak.net/lxml/

Calm down, Stefan! The inquirer must have got the impression of
Amara's superiority from somewhere; all you need to do is to discover
which resource gave that impression and balance the situation out. In
any case, it's an opinion: there's no fear involved, and the only
uncertainty and doubt are in the inquirer's mind, which I suppose is
why they posted the inquiry.

Although the phrase "without even checking them all out" hardly lends
credence to the OP's assertion of Amara's superiority, and does tend to
support a hypothesis involving some ulterior motive (or would if less
ingenuously done).

Given keywords like "Amara" and "Elementtree" and past history, it
looked to me like a troll of one kind trying to incite a troll of
another kind to pop out from under the bridge and chew on his
ankle :-)

[metaperl]

On Mar 14, 5:34 pm, "John Machin" <[e-mail]:

>
Given keywords like "Amara" and "Elementtree" and past history, it
looked to me like a troll of one kind trying to incite a troll of
another kind to pop out from under the bridge and chew on his
ankle :-)

Well, I'm not a troll. And I am now even less impressed with
ElementTree. It does not preserve the document but reforms it in
certain cases.

<script></script>

gets rewritten as

<script />

which leads to problems when embedding Dojo Rich Text Editors.

[Marc 'BlackJack' Rintsch]

In <11**********************@n76g2000hsh.googlegroups .com>, metaperl
wrote:

Well, I'm not a troll. And I am now even less impressed with
ElementTree. It does not preserve the document but reforms it in
certain cases.

<script></script>

gets rewritten as

<script />

which leads to problems when embedding Dojo Rich Text Editors.

So those editors have some problem with XML. I'm not impressed
by such code. ;-)

It's like complaining about:

In [100]: '\x61'
Out[100]: 'a'

It's the same information, just a different representation. Both are
within the specs.

Ciao,
Marc 'BlackJack' Rintsch

[John Machin]

On Mar 15, 9:04 am, "metaperl" <[e-mail]:

On Mar 14, 5:34 pm, "John Machin" <[e-mail]:

Given keywords like "Amara" and "Elementtree" and past history, it
looked to me like a troll of one kind trying to incite a troll of
another kind to pop out from under the bridge and chew on his
ankle :-)

Well, I'm not a troll. And I am now even less impressed with
ElementTree. It does not preserve the document but reforms it in
certain cases.

<script></script>

gets rewritten as

<script />

which leads to problems when embedding Dojo Rich Text Editors.

As BlackJack has pointed out, that's syntactically equivalent. If that
bothers you, best you don't check out what it does with namespace
prefixes.

[Paul Boddie]

On 14 Mar, 17:45, Steve Holden <[e-mail]:

>
Although the phrase "without even checking them all out" hardly lends
credence to the OP's assertion of Amara's superiority, and does tend to
support a hypothesis involving some ulterior motive (or would if less
ingenuously done).

It's an opinion being expressed, qualified with evidence to show that
it's weakly substantiated. Hardly a reason to "call FUD" or to suspect
an astroturfing campaign - FUD being one of the most readily used
terms (and consequently one of the most overused or inappropriately
used terms) when people disagree in the technology domain, anyway.

If the inquirer has it wrong, a gentle process of education is
necessary, although it should be noted that the process does involve
listening to their account of what they are trying to achieve rather
than telling them that they are stupid for using a particular tool or
other.

Paul

[Diez B. Roggisch]

metaperl schrieb:

On Mar 14, 5:34 pm, "John Machin" <[e-mail]:

>Given keywords like "Amara" and "Elementtree" and past history, it
looked to me like a troll of one kind trying to incite a troll of
another kind to pop out from under the bridge and chew on his
ankle :-)

Well, I'm not a troll. And I am now even less impressed with
ElementTree. It does not preserve the document but reforms it in
certain cases.

<script></script>

gets rewritten as

<script />

which leads to problems when embedding Dojo Rich Text Editors.

Beside the fact that this is a problem of browsers (which won't help you
much, as you have to deal with this somehow, blaming browsers alone
isn't solving problems I'm just too aware of), I wonder: which XML
framework _does_ preserve such tags?

I can't imagine a way of doing so as a DOM node is a DOM node. I could
only create a custom renderer for HTML to circumvene that specific
pitfall - but then, you could also create a post-rendering filter
regex-based that expands the script-tags

Diez

[Paul Boddie]

Diez B. Roggisch wrote:

metaperl schrieb:

Well, I'm not a troll. And I am now even less impressed with
ElementTree. It does not preserve the document but reforms it in
certain cases.

<script></script>

gets rewritten as

<script />

which leads to problems when embedding Dojo Rich Text Editors.

Beside the fact that this is a problem of browsers (which won't help you
much, as you have to deal with this somehow, blaming browsers alone
isn't solving problems I'm just too aware of), I wonder: which XML
framework _does_ preserve such tags?

I can't remember exactly how I solved this within an XML/XSLT-heavy
Java-based framework, mostly to satisfy Internet Explorer if I
remember correctly, but it does lead to some fairly inelegant hacks.
Something like this might work:

<script><![CDATA[]]></script>

I note that such things are mentioned here (the first search result
for "CDATA" when I tried):

http://www.w3schools.com/xml/xml_cdata.asp

Paul

P.S. I'm off to fix CDATA support in libxml2dom, but I'm sure lxml
supports CDATA. Not sure about Amara, though. ;-)

[metaperl.etc]

On Mar 15, 5:45 pm, "Paul Boddie" <[e-mail]:

>
I can't remember exactly how I solved this within anXML/XSLT-heavy
Java-based framework,

I just put a single space between:

<script</script>

and then it kept the XML container-style as opposed to single-tag.