Remove entries found in file2 from file1 using sed, awk

[sandeep t]

Hi,
I have a very large XML file(Lets say file1.xml) and I have another XML file(file2.xml) which will contain some entries same as file1.xml
My purpose is to remove those common entries from file1.xml and store it to file3.xml

I can't sort the file. How can I do it?
Pleas help!!

[burschik]

Why can't you use "sort"?

[piwener]

I don't really understand your purpose.
Can you post an example.
I think you should use a scripting language like Perl or PHP to do it.

[sandeep t]

I can't use the sort because it's an installer project file and I do not want to screw it up by sorting it.
Let's take an example of text files:

file1.txt contains:

line1
line2
line3
line4
line5

file2.txt contains:

line3
line4

My purpose is to have a file3.txt which should contain:

line1
line2
line5

[ghostdog74]

show samples of your xml files.

[piwener]

for your simple example above, you can do:

grep -v -f file1.txt file2.txt > file3.txt

[sandeep t]

Thanks, It works for text files, but does not seem to work for xml files.

file1.xml

<fileEntry mountPoint="559" file="${compiler:TOOLS_ROOT}/lib/embed.lib" overwrite="4" shared="false" mode="644" uninstallMode="0" />
<fileEntry mountPoint="559" file="${compiler:ICU_ROOT}/lib/icuin.lib" overwrite="4" shared="false" mode="644" uninstallMode="0" />
<fileEntry mountPoint="559" file="${compiler:ICU_ROOT}/lib/icuuc.lib" overwrite="4" shared="false" mode="644" uninstallMode="0" />
<fileEntry mountPoint="559" file="${compiler:CXX_ROOT}/lib/mscvcch.lib" overwrite="4" shared="false" mode="644" uninstallMode="0" />

file2.xml

<fileEntry mountPoint="559" file="${compiler:TOOLS_ROOT}/lib/embed.lib" overwrite="4" shared="false" mode="644" uninstallMode="0" />
<fileEntry mountPoint="559" file="${compiler:CXX_ROOT}/lib/mscvcch.lib" overwrite="4" shared="false" mode="644" uninstallMode="0" />

[piwener]

Hello,

for your tiny xml example above, it still works perfectly.
I tried it!

grep -v -f file2.xml file1.xml > file3.xml

[scm]

man comm will tell you all you need to know.