Minidom should not pretty-print inside text elements

[mitar]

BPO	37442
Nosy	@mitar

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = None
created_at = <Date 2019-06-28.18:04:51.410>
labels = ['expert-XML']
title = 'Minidom should not pretty-print inside text elements'
updated_at = <Date 2019-06-28.18:04:51.410>
user = 'https://github.com/mitar'

bugs.python.org fields:

activity = <Date 2019-06-28.18:04:51.410>
actor = 'mitar'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['XML']
creation = <Date 2019-06-28.18:04:51.410>
creator = 'mitar'
dependencies = []
files = []
hgrepos = []
issue_num = 37442
keywords = []
message_count = 1.0
messages = ['346846']
nosy_count = 1.0
nosy_names = ['mitar']
pr_nums = []
priority = 'normal'
resolution = None
stage = None
status = 'open'
superseder = None
type = None
url = 'https://bugs.python.org/issue37442'
versions = ['Python 3.6']

Linked PRs

• gh-81623: Do not add whitespace to significant content when pretty-printing #156660

[mitar]

Inside text elements both in HTML and SVG white-space is significant and introduces differences in how things are rendered. By default in general all white-space is collapsed into one space and then this is rendered, adding additional text content.

I observed this while working with SVG which can have content like:

<text><tspan>foo</tspan><tspan>bar</tspan></text>

After pretty-printing it with minidom, and white-space collapsing, the following is what is input to SVG rendering:

<text> <tspan>foo</tspan> <tspan>bar</tspan> </text>

And space between "foo" and "bar" is now visible and while before it was one word to the user, now it is shown as two.

Related issue: mozman/svgwrite#58

I think pretty-printing not add whitespace inside text elements.

[Andrej730]

Still an issue for me with svgwrite because of that you basically either avoid pretty-printing or find some hack to remove the text inside tag - e.g. if you know font-size for children tags you can do: font-size: 0 on text.

[serhiy-storchaka]

PR #156660 stops pretty-printing from adding whitespace inside an element which is marked with xml:space="preserve" or which contains text, in both xml.dom.minidom.toprettyxml() and xml.etree.ElementTree.indent().

toprettyxml() also takes into account the content model declared in the DTD: only white space in element content is ignorable (XML 1.0, 3.2.1). With such declaration the reported case is no longer changed:

>>> src = ("<!DOCTYPE text [<!ELEMENT text (#PCDATA|tspan)*><!ELEMENT tspan (#PCDATA)>]>"
...        "<text><tspan>foo</tspan><tspan>bar</tspan></text>")
>>> parseString(src).documentElement.toprettyxml(indent="  ")
'<text><tspan>foo</tspan><tspan>bar</tspan></text>\n'

Without a DTD nothing in the document says that the whitespace between the tspan elements is significant, and xml:space="preserve" is the only way to express it. ElementTree has no such information at all.