One last point I missed in my previous post -- big files. Many Perl
functions and/or libraries expect to do everything in RAM. This becomes
a problem when I want to compress, encrypt, save, and checksum 14 GiB
system drive images using a live drive and a Perl program in computers
without large amounts of memory (64 GiB?). So, I ended up writing a
Perl program that is truly a script -- it formulates and invokes
gzip(1), ccrypt(1), md5sum(1), and sha256sum(1) commands as required.
Traditional Bourne shell scripts are ideally suited for such tasks, and
the syntax can be cleaner than Perl in simple cases.
On 2020-07-19 03:46, Ajith R wrote:
I seem to recall that he puts Perl at the top of the
heap, and notes that Perl compatible regular expressions (PCRE) are
available via libraries in other programming languages.
Thanks for confirming that I didn't make a wrong choice. Programs that claim to
use PCRE don't support everything that PERL does.
My use of regular expressions is primarily via grep/ egrep and Perl. I
am curious to see what Erlang offers.
I wanted to clean many documents (Wikipedia dump) to analyse the Malayalam
content. As I was not comfortable with scripting, I was looking for some
prorgam that could remove the foreign language text from the files. As, I could
find none that could do the job, I had to use a Perl script with the line below
(among others)
s/[^\p{Block: Malayalam}\p{Block: Basic_Latin}\p{Block:
General_Punctuation}\s]//g; # remove characters outside the specified unicode
blocks.
As of now, the simple substitute command of perl is sufficient for my
requirements. Even that one command appears powerful compared to others.
That sounds like you need lexing and parsing, followed by your desired
processing. In simple cases, Perl regular expressions can accomplish
two or three of those tasks. But as complexity grows, you will need
more and more code. Writing a lexer/ parser in any language is a
non-trivial task. I have used the Perl 'LWP' library to parse HTML 4
pages, but I have not tried to parse HTML 5 and/or Wikipedia pages. I
would look for a library:
https://metacpan.org/search?q=parse
https://metacpan.org/search?q=wikipedia
I spent a little time with Raku (formerly Perl 6). AIUI improving
regular expressions was a design goal of Perl 6, and features were added
specifically for parsing. There is a book dedicated to the subject. I
have a friend who put some time into this area, and he seemed impressed:
https://www.apress.com/us/book/9781484232279
David
