python regular expression replace

Question

I'm trying to change a string that contains substrings such as

the</span></p>
<p><span class=font7>currency

to

the currency

At the line break is CRLF

The words before and after the code change. I only want to replace if the second word starts with a lower case letter. The only thing that changes in the code is the digit after 'font'

I tried:

p = re.compile('</span></p>\r\n<p><span class=font\d>([a-z])')
res = p.sub(' \1', data)

but this isn't working

How should I fix this?

In addition to @JBernardo your question is unclear. Such as? So what else? — FailedDev, Oct 15 '11 at 23:48
@JBernardo: what would you suggest, especially to replace only if the second word starts with a lower case letter? — foosion, Oct 15 '11 at 23:56
How about putting `python strip html` into the search box and pressing enter? The first hit will get you [this nifty anser](http://stackoverflow.com/questions/753052/strip-html-from-strings-in-python/925630#925630). — ekhumoro, Oct 16 '11 at 00:05
@ekhumoro: because the string is littered with other html that I don't want to strip — foosion, Oct 16 '11 at 00:13
@heltonbiker: the compile statement is not matching anything in the string. The string has many occurrences of substrings in the form indicated. — foosion, Oct 16 '11 at 00:14

score 1 · Accepted Answer · answered Oct 16 '11 at 00:00

1

Use a lookahead assertion.

p = re.compile('</span></p>\r\n<p><span class=font\d>(?=[a-z])')
res = p.sub(' ', data)

answered Oct 16 '11 at 00:00

Peter Graham

11,323
7
40
42

That does not seem to find the substring. p.match(data) returns nothing – foosion Oct 16 '11 at 00:17
@foosion: `match` searches from the start of the string. Try `p.search(data)`. – Peter Graham Oct 16 '11 at 00:27

FailedDev · Answer 2 · 2011-10-16T00:25:28.483

1

This :

result = re.sub("(?si)(.*?)</?[A-Z][A-Z0-9]*[^>]*>.*</?[A-Z][A-Z0-9]*[^>]*>(.*)", r"\1 \2", subject)

Applied to :

the</span></p>
<p><span class=font7>currency

Produces :

the currency

Although I would strongly suggest against using regex with xml/html/xhtml. THis generic regex will remove all elements and capture any text before / after to groups 1,2.

edited Oct 16 '11 at 00:25

answered Oct 16 '11 at 00:02

FailedDev

26,680
9
53
73

The reason the html seems reversed is that you're seeing the end of one paragraph and the beginning of another. That regex seems very complex – foosion Oct 16 '11 at 00:20
@foosion Well this regex will work regardless of any elements you throw in it. – FailedDev Oct 16 '11 at 00:26

score 1 · Answer 3 · answered Oct 16 '11 at 00:03

1

I think you should use the flag re.DOTALL, which means it will "see" nonprintable characters, such as linebreaks, as if they were regular characters.

So, first line of your code would become :

p = re.compile('</span></p>..<p><span class=font\d>([a-z])', re.DOTALL)

(not the two unescaped dots instead of the linebreak).

Actually, there is also re.MULTILINE, everytime I have a problem like this one of those end up solving the problem.

Hope it helps.

answered Oct 16 '11 at 00:03

heltonbiker

26,657
28
137
252

That does not seem to find the substring. p.match(data) returns nothing. – foosion Oct 16 '11 at 00:16

python regular expression replace

3 Answers3