#!/usr/bin/env bash

## NPR to RSS

topics=(
    1003 # National
    1004 # World
    1006 # Business
    1007 # Science
    1013 # Education
    1017 # Economy
    1019 # Technology
    1008 # Arts & Life
)

tmptex="$(mktemp).tex"
trap "rm -f $tmptex" EXIT
tmppdf="/tmp/NPR-$(date "+%Y-%m-%d").pdf"
site="https://text.npr.org"

function tex_escape {
    sed 's|{|\{|g' \
    | sed 's|}|\}|g' \
    | sed 's|~|\\textasciitilde{}|g' \
    | sed 's|\\|\\textbackslash{}|g' \
    | sed 's|\^|\\textasciicircum{}|g' \
    | sed 's|\#|\\\#|g' \
    | sed 's|\$|\\\$|g' \
    | sed 's|%|\%|g' \
    | sed 's|&|\\&|g' \
    | sed 's|_|\_|g'
}

function bodyfix {
    sed 's|.*&middot; |<p>|g'
}

function to_tex {
    pandoc --standalone -f html -t latex | tail -n +50 | head -n -2 \
        | sed 's|\\rule{3in}|\\rule{1in}|g'
}

function html_encode {
    sed 's/&/\&amp;/g; s/</\&lt;/g; s/>/\&gt;/g; s/"/\&quot;/g; s/'"'"'/\&#39;/g'
}

function get_topic_stories {
    tid=$1
    turl="$site/t.php?tid=$tid"
    tpage="$(curl --silent "$turl")"
    ttitle="$(echo "$tpage" | sed '13p;d' | awk -F'Topic: ' '{print $2}' | awk -F'<' '{print $1}' | html_encode)"

    readarray sids < <(echo "$tpage" | sed '13p;d' | sed 's/<li>/\n<\/li>/g' | awk -F'sId=' '{print $2}' | awk -F'[&>]' '{print $1}' | grep -v '^$')

    for (( i=0; i<${#sids[@]}; i++ ))
    do
        sid="$(echo ${sids[$i]})"
        spage="$(curl --silent "$site/s.php?sId=$sid")"
        stitle="$(echo "$spage" | grep '<p>' | sed '3p;d' | sed 's|<p>||g' | sed 's|</p>||g' | html_encode)"
        sbyline="$(echo "$spage" | grep '<p>' | sed '4p;d' | sed 's|<p>||g' | sed 's|</p>||g' | html_encode)"
        sbody="$(echo "$spage" | bodyfix | grep '<p>' | tail -n +5 | head -n -1)"

        if [[ "$(echo "$byline" | cut -d ' ' -f 1 )" != "By" ]]
        then
            sbyline="Program Error"
        fi

        # If there isn't much of a body, skip it
        if [[ 5 -gt $(echo "$sbody" | wc -l) ]]
        then
            continue
        fi

        # Put it into RSS
        echo "<item>"
        echo "<title>$stitle</title>"
        echo "<content:encoded>$(echo "$sbody" | base64)</content:encoded>"
        echo "<pubDate>Tue, 02 Oct 2018 19:20:00 -0400</pubDate>"
        echo "<author>not-used@example.com ($sbyline)</author>"
        echo "<category>$ttitle</category>"
        echo "</item>"
    done
}

# Bulid the document
cat <<EOF
<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
    <title>NPR</title>
    <link>http://this-is-a-link-to-the-channel-page</link>
    <description>Description of the channel</description>
    <language>en</language>
    <copyright>Copyright 2018 NPR</copyright>
    <generator>RSS to LaTeX</generator>
    <managingEditor>d@visr.me (Davis Remmel)</managingEditor>
    <!-- Does this one even matter? -->
    <lastBuildDate>Tue, 02 Oct 2018 19:20:00 -0400</lastBuildDate>
EOF

# Items in the middle
for tid in ${topics[@]}
do
    get_topic_stories $tid
done

# Footer
cat <<EOF
</channel>
</rss>
EOF