Howto use StringRegExpReplace() to ins CRLF into XML files?

[rudi]

Hi.

I'd like to beak lines in a given XML file after each "</some value>", so that it's easy to read with ANY editor, e.g. Notepad.

Can't I use StringRegExpReplace() to do so? StringRegExp() DOES correctly catch the substing where I want to insert a CRLF muttley

#include <array.au3>

$string='<Value1>"some value"</Value1><Value2>"Another Value goes here"</Value2>'
$SplitThis=StringRegExp($string,"\</.*(\>\<)",3)
_ArrayDisplay($SplitThis)
StringRegExpReplace($string,"\</.*(\>\<)","\>" & @CRLF & "\<")
MsgBox(0,"Error: " & @error & ", Ext: " & @extended,$string) ; it's still "one line" :(

Regards, Rudi.

[JFee]

I think you're making this hard on yourself

StrinRegExpReplace($string, "(</(?:.*?)>)[^{" & @CRLF & "}]", "\1" & @CRLF)

Not tested but should work. I'll test when I get to my dev machine

[rudi]

I think you're making this hard on yourself

StrinRegExpReplace($string, "(</(?:.*?)>)[^{" & @CRLF & "}]", "\1" & @CRLF)

Hm. Let me ask step by step what you are doing there?

1.)

"(</"

so "<" doesn't need to be escaped, right? (\<)

2.)

(?: ....)

from the help file: "Non captureing group". If I got it correctly, it has to match, but "doesn't show up" anywhere in the resulting array?

(?:.*) = any character, 0 to whatever occurences show up. And what for is the trailing "?" (?:.*?)

3.)

[^ ....

opening a class definition, allowing all BUT the listed items? ("[" opening the class, "^" negation )

Is that to prevent another insertion of @CRLF in case it's already a CRLF in there?

4.)

[^{" & @CRLF & "}]

from the "{" I've lost you: All I know for "{...}" is to use them to define the "number of occurences", like {1,} or {4,9}

RegEx is really powerfull, but sometimes hard to read muttley

And, BTW: What is wrong, or too complicated in the RegEx I've tried? (I don't get it)

Regards, Rudi.

[JFee]

You are absolutely right on all of your remarks.... the {'s are not supposed to be there muttley

StrinRegExpReplace($string, "(</(?:.*?)>)[^" & @CRLF & "]", "\1" & @CRLF)

And in your original one you were replacing just the end of the string, not the entire </tag>

[rudi]

Hi.

This is still not working, no replaces happen muttley

That's what I want to replace:

<section1>SectName<value1>name1</value1><value2>V2</value2></section1><section2>...

These red marked is what I want to "reg-ex-look-for"

The underlined ">" are the closing ones of "</some name>"

They shall be replaced with ">" & @crlf

Howto?

Regards, Rudi.

[smashly]

Hi.

This is still not working, no replaces happen muttley

That's what I want to replace:

<section1>SectName<value1>name1</value1><value2>V2</value2></section1><section2>...

These red marked is what I want to "reg-ex-look-for"

The underlined ">" are the closing ones of "</some name>"

They shall be replaced with ">" & @crlf

Howto?

Regards, Rudi.

Hi, is this what you mean?

$string='<Value1>"some value"</Value1><Value2>"Another Value goes here"</Value2>'
$SRER = StringRegExpReplace($string,"(\><)", ">" & @CRLF & "<")
MsgBox(0, "Replacements: " & @extended, $SRER)

Edited July 23, 2008 by smashly

[rudi]

Hi, is this what you mean?

$string='<Value1>"some value"</Value1><Value2>"Another Value goes here"</Value2>'
$SRER = StringRegExpReplace($string,"(\><)", ">" & @CRLF & "<")
MsgBox(0, "Replacements: " & @extended, $SRER)oÝ÷ Ûú®¢×¢×±iËeÉ«­¢+ÙMÑÉ¥¹IáÁIÁ± ÀÌØíÍÑÉ¥¹°ÅÕ½Ðì Ðì±Ðì¤ÅÕ½Ðì°ÅÕ½ÐìÐìÅÕ½ÐìµÀì
I1µÀìÅÕ½Ðì±ÐìÅÕ½Ðì¤ìÑ¡ÅÕ½ÐìÀäÈìÅÕ½Ðì¥Ì¹½ÐÉÅÕ¥É

This will replace ANY occurence of "><" with ">" & @CRLF & "<"

But I want do do so ONLY if the closing ">" belongs to a "XML-Close", like "</{name}>"

Regards, Rudi.

[rudi]

Hi, is this what you mean?

$string='<Value1>"some value"</Value1><Value2>"Another Value goes here"</Value2>'
$SRER = StringRegExpReplace($string,"(\><)", ">" & @CRLF & "<")
MsgBox(0, "Replacements: " & @extended, $SRER)oÝ÷ Ûú®¢×¢×±iËeÉ«­¢+ÙMÑÉ¥¹IáÁIÁ± ÀÌØíÍÑÉ¥¹°ÅÕ½Ðì Ðì±Ðì¤ÅÕ½Ðì°ÅÕ½ÐìÐìÅÕ½ÐìµÀì
I1µÀìÅÕ½Ðì±ÐìÅÕ½Ðì¤ìÑ¡ÅÕ½ÐìÀäÈìÅÕ½Ðì¥Ì¹½ÐÉÅÕ¥É

This will replace ANY occurence of "><" with ">" & @CRLF & "<"

But I want do do so ONLY if the closing ">" belongs to a "XML-Close", like "</{name}>"

Regards, Rudi.

[rudi]

Hi.

I've got now, that with "()" I get groups returned that can be referenced later in the replace part with "\1", "\2", ...

So my thoughts now are to use "(</.*)" to match the start of closing XML tags. Making it "(/.*?)" matches not the largest, but the (my English) smaller part.

Using a regEx "(</.*?)(>)" upon the string "</closing tag>" results in \1 = "</closing tag", and \2 =">", and as expected only the closing XML tags are matching -- when using StringRegExp(). But I fail to add a @CRLF following the closing ">" with StringRegExpReplace()

; trying to insert a @CRLF after closing XML tags...

#include <array.au3>

$string="<value1>Val1</value1><Sect2>Sect2<value2.1>Text</value2.1><value2.2>text</value2.2></Sect2>"
;                    ^^^^^^^^^ match one                 ^^^^^^^^^ match 2       ^^^^^^^^^^^^^^^^^^^^ 3 + 4
$regEx="(</.+?)(>)" 
$RegExReplace="\1\2" & @CRLF
MsgBox(0,"Original string:",$string & @LF & "Search expression: '" & $regEx & "'" & @LF & "Replace Expression:'" & $RegExReplace & "'")

; stringregexp *DOES* catch what I want:
$Result=StringRegExp($string,$regEx,3)
_ArrayDisplay($Result,"Four matches, split apart, as expected")

; how to replace this "catch" with stringregexpreplace() ??
StringRegExpReplace($string, $regEx, $RegExReplace)

MsgBox(0,"no @CRLFs inserted :(",$string)

What is my mistake?

Regards, Rudi.

Edited July 23, 2008 by rudi

[weaponx]

You probably shouldn't be brute force formatting XML, it could get seriously corrupted. Especially when namespaces and CDATA areas come into play. You should use the XML UDF along with a stylesheet.

Something like this:

http://skew.org/xml/stylesheets/reindent/

-or-

<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
     <xsl:output method="xml" version="1.0" encoding="utf-8" indent="yes"/>
 </xsl:stylesheet>

You might have to search around for an XSL file that works for your needs.

Edited July 23, 2008 by weaponx

[rudi]

You probably shouldn't be brute force formatting XML, it could get seriously corrupted.

OK. This is just about VIEWING an XML file, that is "10 pages in two lines". I know that I can do so in IE as well now.

Especially when namespaces and CDATA areas come into play. You should use the XML UDF along with a stylesheet.

Something like this:

http://skew.org/xml/stylesheets/reindent/

-or-

<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
     <xsl:output method="xml" version="1.0" encoding="utf-8" indent="yes"/>
 </xsl:stylesheet>

You might have to search around for an XSL file that works for your needs.

Thanks for you reply, URL, sample.

muttley How to use these XSL files ? I honestly have no clue...

Just to improve my knowlege upon StringRegExpReplace(), (no matter if it's wise to do so) -- do you see why my approach didn't work as I expected?

Regards, Rudi.

[Squirrely1]

I have an includes file that apparently comes with the latest beta of AutoIt. It has a function for this:

_XMLTransform ( oXMLDoc, Style = "",szNewDoc="" )

I don't think you have to know much about xsl to just apply the transformation to your xml.

I have tried the paste-a-string method and weaponx is right - and this method kind of defeats the (supposed) simplicity of working with xml.

[rudi]

Hi.

> [new function in the current beta]

Thanks, I'll get the latest beta and try that one.

BTW: Maybe somebody can point out the mistake in the StringRegExReplace() syntax I have used? I'd like to understand what I missed muttley

Regards, Rudi.

[weaponx]

I would just do this:

$string='"some value""Another Value goes here"'
$string = StringReplace($string, ">", ">" & @CRLF)
$string = StringReplace($string, "<", @CRLF & "<")
$string = StringReplace($string, @CRLF & @CRLF, @CRLF)
ConsoleWrite($string)

[Squirrely1]

Rudi - once you learn, then teach me - I couldn't regExp my way out of a wet paper bag.

[rudi]

@weaponx

I would just do this:

$string='"some value""Another Value goes here"'
$string = StringReplace($string, ">", ">" & @CRLF)
$string = StringReplace($string, "<", @CRLF & "<")
$string = StringReplace($string, @CRLF & @CRLF, @CRLF)
ConsoleWrite($string)

Yes, that's fine. The one I'm INTERESTED in, interested to UNDERSTAND RegEx in gereral and StringRegExpReplace() especially, is this one:

"Add a @CRLF behind each closing XML tag".

Green = the beginning of a "closing tag"

Blue = the tag name. (don't touch these, just to "catch" the right ">" Chars)

Red = the final ">" of a closing tag.

So I expected this behaviour:

RegEx = "(</)" should catch the beginning of closing tags = "</", right?

RegEx = "(</.*)" should match from the beginning, largest possible match?

RegEx = "(</.*?)" should match the smallest match in stead of the largest?

RegEx = "(>)" should match the final ">". That's the one to add a @CRLF behind it.

So finally:

RegEx = "(</.*?)(>)" will return as "\1" the "green+blue" and as "\2" the "read".

When adding a :? to the first group, then just the 2nd one will be "\1" ??

RegEx = "(:?</.*?)(>)"

Where is my mistake? Or how to use this with StringRegExpReplace()?

Regards, Rudi.

[weaponx]

I still think you are trying to over-simplify something that is not simple.

$xml = '<?xml version="1.0" encoding="ISO-8859-1"?><bookstore><book><title lang="eng">Harry Potter</title><price>29.99</price></book><book><title lang="eng">Learning XML</title><price>39.95</price></book></bookstore>'
$new_xml = StringRegExpReplace($xml, '<.+?>',"$0" & @CRLF)
ConsoleWrite($new_xml & @CRLF)

<?xml version="1.0" encoding="ISO-8859-1"?>
 <bookstore>
 <book>
 <title lang="eng">
 Harry Potter
 </title>
 <price>
 29.99
 </price>
 
 </book>
 <book>
 <title lang="eng">
 Learning XML
 </title>
 <price>
 39.95
 </price>
 
 </book>
 
 </bookstore>

This leaves extra whitespace. Next you will replace double carriage returns, this takes 3 passes. I don't think this can be done in a single pass correctly.

Edited July 24, 2008 by weaponx

[rudi]

I still think you are trying to over-simplify something that is not simple.

This final question was not to really use this, but to see, what's my mistake.

$new_xml = StringRegExpReplace($xml, '<.+?>',"$0" & @CRLF)

And there I CAN see it now:

1.) I can catch the closing XML tags in *one* RegEx group.

2.) the first group match is referenced through "$0" (or "\0") and NOT through "\1", as I tried to use.

Thanks, Rudi.

[dhaley77]

2.) the first group match is referenced through "$0" (or "\0") and NOT through "\1", as I tried to use.

I know this is kinda old, but just in case someone is searching.....

$0 doesn't actually match the first group, it matches the entire search string (like "&" in UNIX/POSIX/etc regex)

In your case, you don't need "()" to catpure anything.

Here's how I'd do this with regexp:

$xml = '<?xml version="1.0" encoding="ISO-8859-1"?><bookstore><book><title lang="eng">Harry Potter</title><price>29.99</price></book><book><title lang="eng">Learning XML</title><price>39.95</price></book></bookstore>'

$new_xml = StringRegExpReplace($xml, "</[^>]*>", "\0" & @CRLF)

ConsoleWrite($new_xml)

This will output:

<?xml version="1.0" encoding="ISO-8859-1"?><bookstore><book><title lang="eng">Harry Potter</title>
<price>29.99</price>
</book>
<book><title lang="eng">Learning XML</title>
<price>39.95</price>
</book>
</bookstore>

Also, if you were trying to add a @CRLF before the end tag (like in weaponx's example), you'd only need two passes. The second pass would just be needed to strip out the extra @CRLF's. Since this is only two passes and the patterns aren't complex, I went ahead and nested the StringRegExpReplace commands (the output from the inner command is the input to the outer command).

Here's the code modified to add the @CRLF before and after the end tag and strip then strip the extra's:

$xml = '<?xml version="1.0" encoding="ISO-8859-1"?><bookstore><book><title lang="eng">Harry Potter</title><price>29.99</price></book><book><title lang="eng">Learning XML</title><price>39.95</price></book></bookstore>'

$new_xml = StringRegExpReplace(StringRegExpReplace($xml, "</[^>]*>", @CRLF & "\0" & @CRLF), @CRLF & @CRLF, @CRLF)

ConsoleWrite($new_xml)

Output:

<?xml version="1.0" encoding="ISO-8859-1"?><bookstore><book><title lang="eng">Harry Potter
</title>
<price>29.99
</price>
</book>
<book><title lang="eng">Learning XML
</title>
<price>39.95
</price>
</book>
</bookstore>

dan