conditionally replace an HTML tag in a large file

[lothario]

We receive this large file which is generated as a HTML page.
So anybody can view it with a web brower.
A few lines from this file ...

<tr><td><font color="#555555" size="3" face="Tahoma, Arial, Verdana, Helvetica">  1999&nbsp; <font color="#cc6600" size="5">&larr;</font> &nbsp;<font color="#0066ff" size="4">03</font>-27 23:36:50 &rarr; <font color="#9932cc" size="2">left</font> &nbsp; &rarr; &nbsp; <font color="#800000">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Stub:</font> <font color="#0066ff" size="2">star</font><font color="#bb6633" size="2"> start</font> Block to negative <font color="#228b22" size="3">&nbsp; Orange </font>        </font></td></tr>  

<tr><td><font color="#555555" size="3" face="Tahoma, Arial, Verdana, Helvetica">  1998&nbsp; <font color="#cc6600" size="5">&larr;</font> &nbsp;<font color="#0066ff" size="4">03</font>-27 23:36:48 &rarr; <font color="#9932cc" size="2">right</font> &nbsp; &rarr; &nbsp; <font color="#800000">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Stub:</font> <font color="#0066ff" size="2">star</font><font color="#9933cc" size="2"> return</font> from negative <font color="#228b22" size="3">&nbsp; Green </font>        </font></td></tr>  

<tr><td><font color="#555555" size="3" face="Tahoma, Arial, Verdana, Helvetica">  1997&nbsp; <font color="#cc6600" size="5">&larr;</font> &nbsp;<font color="#0066ff" size="4">03</font>-27 23:36:46 &rarr; <font color="#9932cc" size="2">up</font> &nbsp; &rarr; &nbsp; <font color="#800000">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Stub:</font> <font color="#0066ff" size="2">star</font><font color="#9933cc" size="2"> return</font> from negative <font color="#228b22" size="3">&nbsp; Yellow </font>        </font></td></tr>

Note:
Each <tr><td> ... </td></tr> is on a single line.
This file has hundreds of lines.

We need to make a small change to this file so it is easier to read.
If the <tr><td> ... </td></tr> line has the word "Yellow" then
we want to replace the '<td>' tag
with '<td bgcolor="#FFFF00">'
so that the line with with word "Yellow" has a yellow background.

How can I do this with a bash shell script?
I can add the shell script to the cron job that delivers this HTML file.

[japafi]

Search with your favorite search engine for perl, search & replace, backreferences.

Quick (and most likely not working) suggestion would be something like

perl -pi -e 's/<tr><td>(.*)Yellow/<tr><td bgcolor="#FFFF00">$1/g' [inputfile.ext]

[giulix]

awk '/Yellow/ {sub(/<td>/, "<td bgcolor=\"#FFFF00\">")} {print}' oldfile.html >newfile.html

[lothario]

Originally Posted by giulix

awk '/Yellow/ {sub(/<td>/, "<td bgcolor=\"#FFFF00\">")} {print}' oldfile.html >newfile.html

Thanks, that worked perfectly.

[briand]

of course it did!

just remember: if awk is the "swiss army knife" of text processing, then perl is the "swiss army chainsaw".

[giulix]

This must be the Swiss safety-pin, then:

sed -i '/Yellow/ s/<td>/<td bgcolor="#FFFF00">/' file.html

It overwrites the input file, though...

[jpollard]

Just remember that the pattern /Yellow/ will replace EVERY occurrence of the word Yellow, even when it does NOT occur in a "<font color=..." tag.

[giulix]

Actually, it applies the substitution to all lines that contain the word Yellow. It doesn't touch "Yellow" at all

[jpollard]

Dang.. missed that it didn't use any fields...