Remove links from text file

Question

how can I remove links from a raw html text? I've got:

Foo bar <a href="http://www.foo.com">blah</a> bar foo

and want to get:

Foo bar blah bar foo

afterwards.

Is it from a text file, with a handful of links, or is it fully generic html? If the latter and you just want something quick and cheap, look into `w3m -dump` or `lynx -dump`. If you want a repeatable or configurable tool, Brian's answer is right, find an HTML parser for the environment you want to use. — sarnold, Jul 04 '10 at 23:16
@spinon - he's using "SED" [Stream Editor] - UNIX... @Marko... putting REGEX at the beginning of his question won't solve his problem — Chase Florell, Jul 04 '10 at 23:16
yeah I figured it was something I wasn't familiar with. Because surely that wasn't a made up tag. I obviously know nothing about that. :) — spinon, Jul 04 '10 at 23:18
@Marko: You shouldn't necessarily prefix your question title with something — that's what tags are for. — icktoofay, Jul 04 '10 at 23:50

score 2 · Answer 1 · edited May 23 '17 at 12:18

2

You're looking to parse HTML with regexps, and this won't work in all but the simplest cases, since HTML isn't regular. A much more reliable solution is to use an HTML parser. Numerous exist, for many different languages.

edited May 23 '17 at 12:18

Community

1
1

answered Jul 04 '10 at 23:11

Brian Agnew

268,207
37
334
440

This is a pretty simple case though. You're not parsing the HTML so much as stripping a specific string (``) and any strings matching a specific pattern (``) from a block of text. This type of manipulation is precisely what regexp is designed for. None of aspects of HTML which make it _non-regular_ come into play (i.e. anchors can't be nested within one another, and we don't care about any other tags). – Lèse majesté Jul 05 '10 at 01:25
But do you care about tags in comments etc.? – Brian Agnew Jul 05 '10 at 07:06

danlei · Answer 2 · 2010-07-04T23:36:00.153

2

sed -re 's|<a [^>]*>([^<]*)</a>|\1|g'

But Brian's answer is right: This should only be used in very simple cases.

edited Jul 04 '10 at 23:36

answered Jul 04 '10 at 23:23

danlei

14,121
5
58
82

score 0 · Answer 3 · answered Jul 04 '10 at 23:12

0

try with:

sed -e 's/<a[^>]*>.*<\/a>//g' test.txt

answered Jul 04 '10 at 23:12

patrick

6,533
7
45
66

This would produce "Foo bar bar foo" instead of "Foo bar blah bar foo" for the example in question. See danlei's solution for the correct version. – Bolo Jul 04 '10 at 23:28

score 0 · Answer 4 · answered Jul 04 '10 at 23:47

0

$ echo 'Foo bar <a href="http://www.foo.com">blah</a> bar foo' | awk 'BEGIN{RS="</a>"}/<a href/{gsub(/<a href=\042.*\042>/,"")}1'

Foo bar blah bar foo

answered Jul 04 '10 at 23:47

ghostdog74

327,991
56
259
343

Remove links from text file

4 Answers4