[TriLUG] Linux From Scratch

[Cristóbal Palmer via TriLUG]

On Tue, Mar 4, 2025, at 14:02, via TriLUG wrote:
> Speaking of signal-to-noise ratio, here's a little bash script 

I write for two reasons. The first is to iterate on your script to make it more cross-platform, and the second is to share an example URL that highlights (for me) the biggest flaw in the approach.

cmp at oof ~ % cat textsig.sh
#!/usr/bin/env bash
while [ ! -z $1 ]; do
	htmlpage=$(curl --disable -sSL "$1")

	if [ $? -ne 0 ]; then
		echo "curl error"
		exit 1
	fi
	htmlbytes=$(echo "$htmlpage" | wc -c)
	echo "html bytes: $htmlbytes"
	txtbytes=$(echo "$htmlpage" | html2text | wc -c)
	echo "text bytes: $txtbytes"
	echo -n "Signal ratio for "
	# some platforms don't support 'echo -e' so we
	# use printf here
	printf "\033[1m\033[4m%s\033[0m\n" "${1}"
	echo " $(echo "$txtbytes / $htmlbytes" | bc -l)"
	unset htmlpage txtbytes htmlbytes
	shift
done
cmp at oof ~ % ./textsig.sh https://www.flickr.com/photos/nasa2explore/
html bytes:   588455
text bytes:     4589
Signal ratio for https://www.flickr.com/photos/nasa2explore/
 .00779838730234257504


Warmly,
CMP