Migrate Neorg files to org-mode (.norg -> .org)

Migrate norg to org.md

Migrate neorg files to orgmode files

I wanted to switch from nvim-neorg to emacs orgmode. The major hurdle for switching was not setting up orgmode, but migrating my thousands of files that I still had in the .norg format.

The script contained here is what I used to migrate all of my files at scale. In the test files you can see how different features will be transformed by the script. Enjoy.

migrate-norg-to-org

#! /bin/sh

if [ "$1" = "--help" ] || [ "$1" = "-h" ] || [ "$1" = "help" ] || [ -z "$1" ]; then
  cat <<EOF
$ migrate-norg-to-org DIR
Transform all .norg files in the directory into .org files.

Parameters:
  \$1: dir to use this on

Flags
  --remove-originals: delete migrated .norg files successful processing

Example:
  $ migrate-norg-to-org .
  $ migrate-norg-to-org ~/org --remove-originals
EOF
  exit
fi

dir="$1"
removeOriginals=0
if [ "$2" = "--remove-originals" ]; then
  printf "Are you sure that you want to remove the orignals after the migration? (y/N): "
  read ans
  if [ "$ans" = "y" ] || [ "$ans" = "Y" ]; then
    removeOriginals=1
  else
    echo "Aborting."
    exit 0
  fi
fi

transform_file_contents() {
  block=
  ol_n=

  if [ "$(wc -l </dev/stdin)" -eq 0 ]; then # one line in the file not caught by IFS=
    cat /dev/stdin
  fi

  cat /dev/stdin | while IFS= read -r line; do
    line=$( echo "$line" |
      sed "s|- (x) |- [X] |; s|- (-) |- [-] |; s|- (_) |- [_] |; s|- (.) |- [ ] |" | # checkboxes & states
      sed "s|~ (x) |~ [X] |; s|~ (-) |~ [-] |; s|~ (_) |~ [_] |; s|~ (.) |~ [ ] |" | # checkboxes & states
      sed 's/{\(.\+\)}\[\(.\+\)\]/[[\1][\2]]/; s/{\(.\+\)}/[\1]/' | # urls
      sed "s|\`<->\`|<->|g" # personal idiocracy
    )

    if [ "$line" = "@document.meta" ]; then
      block=meta
      continue
    fi

    if echo "$line" | grep "^ *> " >/dev/null; then
      if [ "$block" != "quote" ]; then
        echo "#+BEGIN_QUOTE"
        block=quote
      fi
      echo "$line" | sed 's|^ *> ||'
      continue
    fi

    if echo "$line" | grep "^ *@code" >/dev/null; then
      block_type=$(echo "$line" | grep -Po "@code \K.+")
      if [ -z "$block_type" ]; then
        block=code
        echo "#+BEGIN_EXAMPLE"
      else
        block=$block_type
        echo "#+BEGIN_SRC ${block}"
      fi
      continue
    fi

    if echo "$line" | grep "^ *@end" >/dev/null; then
      if [ "$block" = "meta" ]; then
        block=
        continue
      elif [ "$block" = "code" ]; then
        echo "#+END_EXAMPLE"
      else
        echo "#+END_SRC ${block}"
      fi
      block=
      continue
    fi

    # meta tags
    if [ "$block" = "meta" ]; then
      if echo "$line" | grep "^title" >/dev/null; then
        title="$(echo "$line" | grep -Po "^title: \K.+")"
        echo "#+TITLE: $title"
      elif echo "$line" | grep "^created" >/dev/null; then
        date="$(echo "$line" | grep -Po "^created: \K.+")"
        echo "#+DATE: $date"
      elif echo "$line" | grep "^deadline" >/dev/null; then # individual tag
        date="$(echo "$line" | grep -Po "^deadline: \K.+")"
        echo "#+DEADLINE: $date"
      elif echo "$line" | grep "^_deadline" >/dev/null; then
        date="$(echo "$line" | grep -Po "^_deadline: \K.+")"
        echo "#+_DEADLINE: $date"
      fi # discard every other meta tag
      continue
    fi

    if [ -n "$block" ]; then
      if [ "$block" = "quote" ]; then
        if echo "$line" | grep "^ *$" >/dev/null; then
          echo "#+END_QUOTE"
          block=
          echo ""
        else
          echo "$line" | sed 's|^ *||'
        fi
        continue
      else # code blocks, don't manipulate them
        echo "$line"
        continue
      fi
    fi

    line=$( echo "$line" |
      sed "s|\`|~|g" # inline code
    )

    if echo "$line" | grep -P "^ *~+ " >/dev/null; then # ol
      contents=$(echo "$line" | grep -Po "~+ \K.+")
      if echo "$line" | grep -P "~{3}" >/dev/null; then
        echo "    + ${contents}" | sed "s|\`|~|g" # inline code
      elif echo "$line" | grep -P "~{2}" >/dev/null; then
        echo "  - ${contents}" | sed "s|\`|~|g" # inline code
      elif echo "$line" | grep -P "~{1}" >/dev/null; then
        if [ -z "$ol_n" ]; then
          ol_n=1
        else
          ol_n=$(($ol_n +1))
        fi
        echo "${ol_n}. ${contents}" | sed "s|\`|~|g" # inline code
      fi
      continue
    fi

    if echo "$line" | grep -P "^ *-+ " >/dev/null; then # ul
      contents=$(echo "$line" | grep -Po "\-+ \K.+")
      if echo "$line" | grep -P "\-{4}" >/dev/null; then
        echo "      + ${contents}"
      elif echo "$line" | grep -P "\-{3}" >/dev/null; then
        echo "    - ${contents}"
      elif echo "$line" | grep -P "\-{2}" >/dev/null; then
        echo "  + ${contents}"
      elif echo "$line" | grep -P "\-{1}" >/dev/null; then
        echo "- ${contents}"
      fi
      continue
    fi

    echo "$line" | # indentation is not removed, because I'm not checking if part of ul/ol
      sed "s|\`|~|g" # inline code

    if [ -z "$line" ]; then
      ol_n= # reset
    fi
  done

  # not covered
  # - strike-through
  # - file ending in a quote
}

transform_file() {
  norgFile="$1"
  filenameNoExt="${norgFile%.*}"
  orgFile="${filenameNoExt}.org"

  echo "Processing: $norgFile"
  transform_file_contents <"$norgFile" >"$orgFile"
}

matched_files=$(mktemp -u)
match_files() {
  if [ ! -e "$matched_files" ]; then
    find "$dir" -type f -name "*.norg" >"$matched_files"
  fi
  cat "$matched_files"
}

match_files | while read file; do
  printf "\n" >> "$file"
  transform_file "$file"
done

if [ "$removeOriginals" -eq 1 ]; then
  echo "Removing orignals"
  match_files | while read file; do
    rm "$file"
  done
fi

test file.norg

@document.meta
title: test file
created: 2024-11-08
updated: 2024-11-08
@end

* H1
  - ( ) /Todo/
  - (-) *In progress*
  - (x) `Done`
  - (_) -Deleted-

  - ul
  -- nested
  --- 3 deep

  > quote me this
    quote me that
    quote me that

  No longer quoting

** H2
   {http://localhost:1313/now/}
   {http://localhost:1313/now/}[outside]
   ~ {http://localhost:1313/now/}[inside]
   ~~ ul
      some text
   ~~ `nested`
   ~~~ (x) l3
   ~ no2
   ~~ ( ) no1 `<->`

   ~ reset

*** H3
    @code javascript
    console.log(`${1+1}`)
    @end

    @code
    123
    @end
***** H5

test file.org

test file

H1

• [ ] Todo
• [-] In progress
• [X] Done
• [_] -Deleted-
• ul
  • nested
    • 3 deep

quote me this quote me that quote me that

No longer quoting

H2

[http://localhost:1313/now/] outside

1. inside
   • ul some text
   • nested
     • [X] l3
2. no2
   • [ ] no1 <->
3. reset

H3

console.log(`$[1+1]`)

123

H5