I am new to regexps, can someone help me in getting a regex for parsing the tag
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
with all the possiblities?
Asked
Active
Viewed 2,678 times
1
-
TAG – vaibhav Jan 21 '11 at 20:41
-
4**All** the possibilities? Don't try to do this with a regular expression. You can get away with them for HTML that fits a template, but for generic parsing, you need a real HTML parser. – Quentin Jan 21 '11 at 20:42
-
3http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags/1732454#1732454 – Reid Jan 21 '11 at 20:43
-
@David @Reid Since – phihag Jan 21 '11 at 20:44
-
1A meta that specifies the character encoding *should* only occur once. – Quentin Jan 21 '11 at 20:45
-
@David Dorward Precisely, and regular expressions are therefore as applicable as searching for the string " – phihag Jan 21 '11 at 20:46
-
My point was that "should" and "will" are not the same thing, I have stumbled across documents with multiple, contradictory encoding declarations. (And that's just limiting this to ``, if you actually want to figure out what the encoding of a document is, then there are half a dozen places you need to look.) – Quentin Jan 21 '11 at 20:47
-
1@phihag still, there are endless variations in case, in attribute order, in other attributes.... – Pekka Jan 21 '11 at 20:48
1 Answers
2
To cover "all the possibilities", you really should be using HTML 5's Determining the character encoding rules. These aren't expressible as a regular expression.
There is an open source Java implementation of it in validator.nu
If you insist on using a regular expression, then this will probably cover most cases where the encoding it declared with a meta element (it won't, for instance, cover XML declarations). It is however, dirty, makes some assumptions that are usually (but may not always be) right and I do not recommend it.
/<meta[^>]+charset=['"]?(.*?)['"]?[\/\s>]/i
Quentin
- 914,110
- 126
- 1,211
- 1,335