Thanks for this really complete answer. I really appreciated it. Thanks for your investment in it. I learnt a lot thanks to it.
Boyd Stephen Smith Jr. wrote:
> On Friday 14 May 2010 12:52:45 Merciadri Luca wrote:
>
>> "Boyd Stephen Smith Jr." <b...@iguanasuicide.net> writes:
>>
>>> On Friday 14 May 2010 12:04:42 Merciadri Luca wrote:
>>>
>>>> I have many text files (actually .tex files) which contain some
>>>> sequence or regexp (it depends on the files) that I would like to
>>>> remove. Is there a commandline/GUI for doing this massive edit?
>>>>
>>> (sed -i -e "s/$regexp//" "$file") for a single file. (GNU sed only.)
>>>
>>> (find $dir -type f -exec sed -i -e "s/$regexp//" {} \;) for all files in
>>> a directory.
>>>
>> I am using the second command. The problem is that, for one set of
>> files (that I have selected, no problem for this), I have to use a
>> really simple expression: I need to find all the occurences of
>> `\paragraph{}' and replace them with nothing (i.e. with `'). I know
>> regexps, but replacing `$regexp' with `\paragraph{}' gives error
>> messages. Any idea? Thanks.
>>
>
> First you need a (basic) regular expression (BRE) that matches
> "\paragraph{}".
> The '\' is a BRE special character, so it needs to be escaped. Also, the
> "{}"
> is a bit troublesome with find/-exec, so we will match it using the construct
> "[{][}]".
>
> The definitive documentation for regular expression is the Single UNIX
> Specification, Version 3 -- Base Definitions, Chapter 9. I don't actually
> like (man 7 regex) for this.
>
> This gives us the regex "\\paragraph[{][}]". Now, we need to get that
> regular
> expression to sed. (find $dir -type f -exec sed -i -e
> "s/\\paragraph[{][}]//"
> {} \;) won't work, since during Quote Removal, one of the '\'s are dropped
> and
> neither find nor sed "sees" it.
>
> The shell does a *lot* of processing to the text you type before it reaches
> the command you are invoking. Single UNIX Specification, Version 3 -- Shell
> and Utilities, Chapter 2 is the core documentation, but some shells are much
> more featureful.
>
> We can either use (find $dir -type f -exec sed -i -e 's/\\paragraph[{][}]//'
> {} \;) OR--my preference--(regex='\\paragraph[{][}]'; find $dir -type f -exec
> sed -i -e "s/$regex//" {} \;) to make sure sed gets that important '\'.
>
> Also, I left it out, but you may want the "g" flag to the "s"ubstitute
> command
> in sed. Otherwise, only one occurrence of the regex will get eliminated per
> line.
>
--
Merciadri Luca
See http://www.student.montefiore.ulg.ac.be/~merciadri/
I use PGP. If there is an incompatibility problem with your mail
client, please contact me.
First deserve, then desire.
signature.asc
Description: OpenPGP digital signature
