Regex for quoting unquoted XML attributes

Question

Edit: The 100% correct theory is that you don't want to do this at all. However I have accepted the answer that helped the most.

So I'm being given ugly XML from a client that promises to fix it. In the meantime I need to clean it up myself. I'm looking for a regex to use in Java to add quotes around unquoted attributes. The general case is better, but so far it is only one attribute that is broken so the regex can specifically refer to "attr1". The value of the attribute is unknown, so I can't include that in the search.

<tag attr1 = VARIABLETEXT>
<tag attr1 = "VARIABLETEXT">not quoted</tag>
<tag attr1 = VARIABLETEXT attr2 = "true">
<otherTag>buncha junk</otherTag>
<tag attr1 = "VARIABLETEXT">"quoted"</tag>

Should turn into

<tag attr1 = "VARIABLETEXT">
<tag attr1 = "VARIABLETEXT">not quoted</tag>
<tag attr1 = "VARIABLETEXT" attr2 = "true">
<otherTag>buncha junk</otherTag>
<tag attr1 = "VARIABLETEXT">"quoted"</tag>

EDIT: Thank you very much for telling me not to do what I'm trying to do. However, this isn't some random, anything goes XML, where I'll run into all the "don't do it" issues. I have read the other threads. I'm looking for specific help for a specific hack.

vi filename.xml; :%s/attr1 = false/attr1 = "false"/g ... There is also gVim for Windows. — Dave Jarvis
– Dave Jarvis, Commented Feb 11, 2010 at 17:51
If its only temporary why dont you just use a cleaning/validation library to preprocess it? — prodigitalson
– prodigitalson, Commented Feb 11, 2010 at 17:52
One question: How can you tell in VARIABLETEXTattr2 where to split? Is it the fact that the next attribute starts with attr? Or a uppercase/lowercase switch? — Tim Pietzcker
– Tim Pietzcker, Commented Feb 11, 2010 at 20:38
Sorry Tim, there was supposed to be a space between VARIABLETEXT and attr2. — Instantsoup
– Instantsoup, Commented Feb 11, 2010 at 20:58

Community · Accepted Answer · 2017-05-23 12:00:17Z

4

Do not use regex to fix/parse/process markup languages. Read here why.

Use a forgiving parser like tidy to read and fix the document in a few easy steps. There is a Java library (jtidy) you can use.

edited May 23, 2017 at 12:00

CommunityBot

11 silver badge

answered Feb 11, 2010 at 17:52

Tomalak

339k68 gold badges547 silver badges635 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

prodigitalson Over a year ago

Thank you for that thread reference. It made life worth living.

Instantsoup Over a year ago

Yeah I've read that. Can anyone just help me with the regex without preaching?

Tomalak Over a year ago

No, I'm sorry. Because there is no way to get it 100% right, there is always some weird corner case. Why is using a parser not an option?

Instantsoup Over a year ago

I'll settle for 89% right then. Thanks for the parser idea. It's not not an option. I just don't have the time to do it right right now, which is why I came here for regex help.

codekiln Over a year ago

If you suggest using tidy you should suggest the configuration options that would let OP achieve what (s)he wants

Tim Pietzcker · Accepted Answer · 2010-02-11 21:24:00Z

2

OK, given your constraints, you could:

Search for

<tag attr1\s*=\s*([^" >]+)

and replace with

<tag attr1 = "\1"

So, in Java, that could be (according to RegexBuddy):

String resultString = subjectString.replaceAll("<tag attr1\\s*=\\s*([^\" >]+)", "<tag attr1 = \"$1\"");

EDIT: Simplified regex a bit more.

edited Feb 11, 2010 at 21:24

answered Feb 11, 2010 at 20:46

Tim Pietzcker

337k59 gold badges520 silver badges572 bronze badges

1 Comment

Instantsoup Over a year ago

Sorry, there is definitely a space between the variable text and attr2.

Autumn Skye · Accepted Answer · 2024-12-17 02:40:43Z

This solution wraps the first occurrence of an unquoted attribute value, even it is in between other properly-quoted attributes (or the first or last attribute):

<a id="a2" href=https://twitter.com/nlm_nih class="ff">

becomes:

<a id="a2" href="https://twitter.com/nlm_nih" class="ff">

    final String SPACE = " \r\n";
    final String ATTNAME_PATTERN = "[a-z]+(?:[-][a-z]+)*";

    // Remove any spaces before and after = (simplifies next regex)
    String wrappedAtts = targetHtml.replaceAll("[" + SPACE + "]*=[" + SPACE + "]*", "=");

    wrappedAtts = wrappedAtts.replaceAll("([<][a-z]+(?:[" + SPACE + "]+" + ATTNAME_PATTERN + "[=][\"][^\"]*[\"])*)[" + SPACE + "]+(" + ATTNAME_PATTERN + ")=([^\"][^" + SPACE + "]+)", "$1 $2=\"$3\"");

If you need to handle multiple occurrences in a tag, just put that last line in a loop and iterate until you don't find any more.

Collectives™ on Stack Overflow

Regex for quoting unquoted XML attributes

3 Answers 3

5 Comments

1 Comment

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

5 Comments

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related