Python: differences in search for regex and replace via regex

Question

I'm writing a small script for auto-correcting interwiki translation links with pywikibot. I look for the existing links and want to rewrite them in a standard format with links to all pages.

The text I'm looking for looks either somewhat like

{{Trad|EN=Under Spring|FR=Sources Interdites|DE=Verbotene Quellen}}

or multiline somewhat like

{{Trad
|DE=Urwurzeln
|EN=Prime Roots
|ES=Raíces Primarias
|FR=Primes Racines
|RU =Изначальных Корней
|H  = 
|palette=primes
}}

I manage to find these two instances in a wiki page source via

reg_strg = '{{trad([\w\s\|\=]*)}}'
rex = re.search(reg_strg, text, re.IGNORECASE | re.MULTILINE)

That yields me the heart of the templates like (for the 1st case)

|EN=Under Spring|FR=Sources Interdites|DE=Verbotene Quellen

and similarily as multiline-string for the 2nd.

However, I now use the same reg_strg in a replace command, it fails to do any replacement, text remains unmodified, new_strg is created from what was read to constitute a replacement string. But the result is independent of whether new_strg is a multi-line string or just a simple "flobberigoo"

text = re.sub(reg_strg, new_strg, text, re.IGNORECASE | re.MULTILINE)

So obviously there is some difference between re.search and re.sub - however I fail to find that in the documentation (even though I am aware of the difference between re.search and re.match, I understood it such, that re.sub should behave like the first).

What do I miss? How can I replace the mentioned regex which I find in the pages with a string?

For completeness sake, this is the complete function including the debug prints:

def replace_translation_template(self, text, translations):
    """
    @param text The page text to look through
    @param translations dictionary of translations
    """
    reg_strg = '{{trad([\w\s\|\=]*)}}'
    rex = re.search(reg_strg, text, re.IGNORECASE | re.MULTILINE)

    print("Replacing:")
    try:
        print(rex.group(1))
        strgs = rex.group(1).split('|')
        print(strgs)
        new_strg = ""
        for lang,pagename in translations.items():
            if pagename is None:
                pagename = ""
            new_strg += '|' + lang.upper() + '=' + pagename + '\n'

        #print("New_strg: ", new_strg)

        for lang in translations.keys():
            for (n,str) in enumerate(strgs):
                if lang.upper() in str:
                    strgs.pop(n)

        for s in strgs:
            if len(s) > 2:
                new_strg += '|' + s + '\n'
        print(new_strg, '\n')
        print('\n with \n \n')
        text = re.sub(reg_strg, new_strg, text, re.IGNORECASE | re.MULTILINE)
        print(text)

    except:
        print("no text matched:", rex)