Replace spaces

Shell Programming and Scripting
1

Hi guys, so I have another issue. Can I use sed to replace spaces in a string or variable with %20

I am having trouble with using curl on URL's containing spaces

Thanks!

2

post sample input and output record

3 
string=$( printf "%s\n" "$string" | sed 's/ /%20/g' )

Or. with bash or ksh93:

string=${string// /%20}
4

this is a sample of the code I have to get the file "GetSeries.php.xml" from thetvdb.com

series_name="green hornet"

curl -s -o /Users/rtipton/Desktop/GetSeries.php.xml "http://thetvdb.com/api/GetSeries.php?seriesname=$series_name"

When I run this, I get a bunch of results in the xml for tv shows including "green" in the title. Curl doesn't include anything after the first space in the URL.

I'm not sure what output I could provide in this instance, let me know if I can give more

Thanks again Vid

5

Hey thanks cfajohnson, that did the trick, the second suggestion works perfectly!

Thanks again for being so helpful guys, I really appreciate it.

6

you can try something like this

 
series_name=`echo "$series_name"|sed 's/ /%20/g'`
7

Nice this works as well. Too many good options here! You guys are good, thanks again!

Rob

8

If your shell support parameter expansion this is the best solution.

9

Not necessarily.

Writing a portable script may be more important than writing the most efficient one.

10

Hey cfa!

string=${string// /%20}

This ended up working in Mac OS X but not in ubuntu hardy, I was getting an error so I ended up using vid's solution:

series_name=`echo "$series_name"|sed 's/ /%20/g'`

This works just fine on ubuntu hardy.

Do you know a lot about sed? I used vid's solution for extracting a number in between <seriesid></seriesid> in an xml file and it worked fine in Mac OS X but when I tried it in ubuntu hardy the output is blank, no number, nothing.

mind taking a look? Here is the sample XML

<Data>

<Series>
<seriesid>73545</seriesid>
<language>en</language>
<SeriesName>Battlestar Galactica (2003)</SeriesName>
<banner>graphical/73545-g11.jpg</banner>

<Overview>blah blah blah blah</Overview>
<FirstAired>2003-12-01</FirstAired>
<IMDB_ID>tt0407362</IMDB_ID>
<zap2it_id>SH710749</zap2it_id>
<id>73545</id>
</Series>
</Data>

and here is the solution given to me by vid to extract "73545" and set it to a variable

variable=`sed -ne 's/\(^\<seriesid\>\)\(.*\)\(\<\/seriesid\>\)/\2/p' filename|head -1`
11

If you can give me the definition of portable maybe I'll reevaluate my concept's of efficiency :cool:

12

It will work on Ubuntu if you specify that it is to be executed by bash. Use a shebang as the very first line of the file:

#!/bin/bash

(Check that bash is indeed in /bin.)

I avoid sed for anything beyond a simple search and replace. You get much more understandable (and thus easier-to-debug) code with awk.

awk -F '[<>]' '/^<seriesid>/ { print $3 }' filename
13

A portable script will run on any modern (i.e., POSIX) system. That means using only POSIX shell syntax and features and standard POSIX/Unix commands.