hugo_new_post.sh

#!/usr/bin/env bash
# Time-stamp: <2017-04-21 17:44:04 kmodi>
# https://discuss.gohugo.io/t/auto-generate-file-name-based-on-title/4648/2?u=kaushalmodi

# This script uses the unofficial strict mode as explained in
# http://redsymbol.net/articles/unofficial-bash-strict-mode
#
# Also checks have been done with www.shellcheck.net to have a level of
# confidence that this script will be free of loopholes.. or is it? :)

set -euo pipefail # http://redsymbol.net/articles/unofficial-bash-strict-mode
IFS=$'\n\t'

h="
┌─────────────────────────────────────────────────────────────────────────────────┐
│                                                                                 │
│ Script        : hugo_new_post.sh                                                │
│ Description   : Script to generate a hugo post file using the title specified   │
│                 at the command line.                                            │
│ Usage Example : hugo_new_post.sh -b \"/home/$USER/hugo/myblog\" -t \"My New Post\"  │
│                                                                                 │
│                                                                                 │
│ OPTIONS                                                                         │
│                                                                                 │
│   -b|--blogpath <path> : Root of hugo blog. (Mandatory argument)                │
│                           --blogpath \"/home/\$USER/hugo/myblog\"                  │
│                                                                                 │
│   -t|--title <string>  : Title string of the post. (Mandatory argument)         │
|                          Use double quotes if the title contains spaces.        │
│                           --title \"My New Post\"                                 │
│                                                                                 │
│   -s|--section <dir>    : Sub-directory in the 'content/' dir where the post    │
│                          should be created. (Default: posts)                    │
│                           --section \"blog\"                                      │
│                                                                                 │
│                                                                                 │
│   -h|--help            : Show this help                                         │
│   -d|--debug           : Debug mode                                             │
│                                                                                 │
└─────────────────────────────────────────────────────────────────────────────────┘"

blog_path=""
section="posts"
title=""
extra_args=""
here=$(pwd)
fext=".md"                     # default file extension is .md

debug=0
help=0

while [ $# -gt 0 ]
do
    case "$1" in
        "-b"|"--blogpath" ) shift
                            blog_path="$1";;
        "-t"|"--title" ) shift
                         title="$1";;
        "-s"|"--section" ) shift
                           section="$1";;
        "-o"|"--org" ) fext=".org";;
        "-m"|"--md" ) fext=".md";;

        "-d"|"--debug" ) debug=1;;
        "-h"|"--help" ) help=1;;
        * ) extra_args="${extra_args} $1";;
    esac
    shift # expose next argument
done

if [[ ${debug} -eq 1 ]]
then
    echo "blog path  = ${blog_path}"
    echo "title      = ${title}"
    echo "sub dir    = ${section}"
    echo "extra args = ${extra_args}"
fi

main () {
    # Remove leading and trailing whitespace from ${title}
    title=$(echo "${title}" | sed -r -e 's/^[[:space:]]*//' -e 's/[[:space:]]*$//')

    # Using ${title}
    #  - Replace '&' with "and", and '.' with "dot".
    #  - Then lower-case the whole title.
    #  - Then replace all characters except a-z, 0-9 and '-' with spaces.
    #  - Then remove leading/trailing spaces if any.
    #  - Then replace one or more spaces with a single hyphen.
    # For example, converts "This, That \& Other!" to "this-that-and-other.md"
    # (Note that we need to escape & with \ above, in the shell.)
    fname=$(echo "${title}" \
                | sed -r -e 's/&/ and /g' \
                      -e 's/\./ dot /g' \
                      -e 's/./\L\0/g' \
                      -e 's/[^a-z0-9-]/ /g' \
                      -e 's/^[[:space:]]*//g' \
                      -e 's/[[:space:]]*$//g' \
                      -e 's/[[:space:]]+/-/g');
    fpath="${blog_path}/content/${section}/${fname}${fext}"

    if [[ ${debug} -eq 1 ]]
    then
        echo "fname      = ${fname}"
        echo "fpath      = ${fpath}"
    fi

    # Create the new post
    # Need to first cd to the hugo blog root dir
    cd "${blog_path}" || exit
    hugo new "${section}/${fname}${fext}" "${extra_args}"

    # Replace the title in TOML front matter with ${title}, and add slug
    tmp_file="/tmp/${USER}_hugo_post"
    cp -f "${fpath}" "${tmp_file}"
    sed -r -e 's/^(\s*title = ).*/\1"'"${title}"'"/' \
        -e 's/^(\s*title = .*)/\1\nslug = "'"${fname}"'"/' \
        "${tmp_file}" > "${fpath}"
    rm -f "${tmp_file}"

    # Insert summary divider at the end of the file
    echo -e '\n<!--more-->' >> "${fpath}"

    # Convert front-matter from TOML to org-mode style for .org content
    if [[ ${fext} == ".org" ]]
    then
        cp -f "${fpath}" "${tmp_file}"
        # - Delete the TOML +++ markers
        # - Change 'title = "foo"' to '#+TITLE: foo', and similar for other
        #   front-matter variables.
        # - Delete '[""]' from lines (like empty categories in TOML representation)
        # - Replace '<!--more-->' to '# more'
        sed -r -e "/^\+\+\+/d" \
            -e "s/^([a-z]+) =/#+\U\1\E:/" \
            -e "s/: \"(.*)\"/: \1/" \
            -e "s/\[\"\"\]//" \
            -e "s/^<!--(more)-->/# \1/" \
            "${tmp_file}" > "${fpath}"
        rm -f "${tmp_file}"
    fi

    # Go back to the directory from where you launched this script
    cd "${here}" || exit

    # Open the file in EDITOR with cursor placed above the summary divider line
    # containing "<!--more-->" (for .md) or "# more" (for .org)
    line_num=$(wc -l "${fpath}" | awk '{ print $1 }') # last line -- contains 'more' divider
    line_num=$((line_num-2)) # get 2 lines above 'more'
    open_file_cmd="${EDITOR} +${line_num} ${fpath} &"
    if [[ ${debug} -eq 1 ]]
    then
        echo "last line     = ${line_num}"
        echo "open file cmd = ${open_file_cmd}"
    fi
    eval "${open_file_cmd}"
}

help () {
    echo "${h}"
}

if [[ ${help} -eq 1 ]]
then
    help
    exit 0
elif [[ -z ${blog_path} || -z ${title} ]]
then
    echo "Error: Both '-b' and '-t' are mandatory arguments"
    help
    exit 1
else
    main
    exit 0
fi