Re to replace & with name entity

Question

I have to replace & with its name entity or decimal entity from input string, but input string may contains other name and decimal entities in with & will present.

Code:

import re
text =' At&T, " < I am > , At&T  so  &#60; &lt; &  & '

#- Get all name entities and decimal entities.
replace_tmp = re.findall("&#\d+;|&[a-z]+;", text)

#- Replace above values from tempvalues.
tmp_dict = {}
count = 1
for i in replace_tmp:
    text = text.replace(i, "$%d$"%count)
    tmp_dict["$%d$"%count] = i
    count += 1


#- Replace & with &amp;
text = text.replace("&", "&amp;")

#- Replace tempvalues values with original.
for i in tmp_dict:
    text = text.replace(i, tmp_dict[i])

print text

Final Output: At&T, " < I am > , At&T so < < & &

But Can I get regular expression which directly does above thing?

Final line in py file:

value = re.sub(r'&(?!(#[0-9]+;|[a-zA-Z]+;))', '&', value).replace("<", "<").replace(">", ">").replace('"', """)

Possible duplicate of [How do I perform HTML decoding/encoding using Python/Django?](http://stackoverflow.com/questions/275174/how-do-i-perform-html-decoding-encoding-using-python-django) — Peter Wood, Apr 08 '16 at 05:53
See particularly [this answer](http://stackoverflow.com/a/7088472/1084416) — Peter Wood, Apr 08 '16 at 05:54
@VivekSable use negative lookahead as indicated in my answer, its better to use predefined function though — riteshtch, Apr 08 '16 at 05:56
@PeterWood i think OP doesn't need that (as shown in his Final Output text) — riteshtch, Apr 08 '16 at 06:00
no, I am getting correct output from above code, I am looking if there is re method which replace `&` with `&` in one or two lines — Vivek Sable, Apr 08 '16 at 06:08
This string we put in attribute value of tag. We can directly replace `<` , '>` and `"` with its name entities.. — Vivek Sable, Apr 08 '16 at 06:10

mkHun · Answer 1 · 2016-04-08T05:53:06.727

1

Use string substitution with negative look ahead.

import re
text =' At&T, " < I am > , At&T  so  &#60; &lt; &  & '

text = re.sub(r'&(?![\w\d#]+?;)',"&amp;",text)
print text

edited Apr 08 '16 at 05:53

answered Apr 08 '16 at 05:38

mkHun

5,891
8
38
85

`&` can be any where, this is only one test case – Vivek Sable Apr 08 '16 at 05:47

riteshtch · Accepted Answer · 2016-04-08T05:53:35.347

1

>>> import re
>>> re.sub(r'&(?!(#[0-9]+;|\w+;))', '&amp;', ' At&T, " < I am > , At&T  so  &#60; &lt; &  & ')
' At&amp;T, " < I am > , At&amp;T  so  &#60; &lt; &amp;  &amp; '

You can use negative look ahead assertion for \w+; (for eg:  ) and #[0-9]+; (for #60;).

Therefore the regex is:

&(?!(#[0-9]+;|\w+;)) negative look ahead assertion ensures there is neither #[0-9]+; nor \w+; ahead of &

You could also use [a-zA-Z]+; instead of \w+;

edited Apr 08 '16 at 05:53

answered Apr 08 '16 at 05:38

riteshtch

8,629
4
25
38

Thanx, looking to same. – Vivek Sable Apr 08 '16 at 06:14
1

`re.sub(r'&(?!(#[0-9]+;|[a-zA-Z]+;))', '&', '''& At&T, " < I am &>& , At&T "so" < < & & ''').replace("<", "<").replace(">", ">").replace('"', """)` this is my final py line – Vivek Sable Apr 08 '16 at 06:22

Re to replace & with name entity

2 Answers2