web2pdf

#!/bin/bash
# web2pdf       Print a webpage to pdf file
#
# Usage: web2pdf URL [pdffile] [encoding]
#
# Args:
#   URL         input web address to convert content
#   pdffile     file-name for output pdf. Default is to auto-generate from
#               input URL
#   encoding    test encoding for for desired output.
#               Text encoding for the input content is auto-detected, and
#               default encoding for output is defined as:
#               - same as input encoding if confidence is 1
#               - ascii/TRANSLIT if confidence is below 1
#               User-define encoding ($3) supersedes the default behaviour
#
# yaswant.pradhan
# -----------------------------------------------------------------------------
set -e

# html2ps config file see (http://user.it.uu.se/~jan/html2psug.html)
CONFIG_FILE='~/.html2psrc'

usage() { echo "usage: $(basename $0) URL [pdffile] [encoding]" ; }
[[ $# -lt 1 ]] && usage && exit 1

webf=$(basename $1)
TMP_HTML="$webf"
OUT_PDF="${webf%.*}.pdf"

[[ -f "$TMP_HTML" ]] && TMP_HTML=tmp_"$webf"
[[ $# -ge 2 ]] && OUT_PDF=$2
if [[ -f "$OUT_PDF" ]]; then
  read -rp "A file $OUT_PDF already exists in path. Overwrite [y|n]? " opt
  [[ "$opt" != y ]] && exit
fi


## html2ps (1.0b5) cant deal with a) secure protocols, b) limited encoding opts
## best is to wget the file and and use gnu iconv to convert to ascii encoding
## https://unix.stackexchange.com/questions/141539
wget -nd -q $1 -O $TMP_HTML \
  || { rc=$?; rm $TMP_HTML; echo -e "ERROR $rc: wget $1 failed."; exit $rc ;}


## autodetect input charset and convert to ascii or perhaps iso-8859-1
encoding=($(chardetect $TMP_HTML |cut -d" " -f 2,5))
confidence=$(echo "${encoding[1]} == 1" |bc)
[[ $confidence -eq 1 ]] && DEF_ENC=${encoding[0]} || DEF_ENC='ascii//TRANSLIT'
TO_ENC=${3:-$DEF_ENC}


## set base URL to embed images
[[ ${1: -1} == '/' ]] && BASE=$1 || BASE=${1%/*}
echo "${0##*/}: BASE_URL=$BASE"


## update html encoding to plain text if input encoding confidence is below 1
## convert to ps | convert tp pdf
iconv -f ${encoding[0]} -t $TO_ENC $TMP_HTML | \
html2ps -b $BASE/ -f $CONFIG_FILE -nUH 2>/dev/null | \
ps2pdf - $OUT_PDF


## print summary
echo "${0##*/}: writing $OUT_PDF (encoded from:$encoding to:$TO_ENC)"


## Clean-up
rm -f $TMP_HTML

## Display output
evince $OUT_PDF