Shell script to extract t() function translations in drupal custom module and generate .pot file

generate_pot.sh

#!/bin/bash

# Shell script to extract $this->t() function translations in drupal 8+ custom module and generate .pot file
# Extract from .php .module .js .twig from current custom module folder
# Exclude .min.js for js libraries
# Put this in your module folder, make it executable : chmod +x generate_pot.sh , run it
# Feel free to improve it 

# Define the regex patterns for each file type
php_t_regex="\$this->t\(\s*'([^']*(?:@[a-zA-Z0-9_]+[^']*)*)'\s*(?:,\s*\[[^\]]*\])?\s*\)"
twig_t_regex="{{\s*['\"]([^'\"]*(?:@[a-zA-Z0-9_]+[^'\"]*)*)['\"]\s*\|?\s*t\s*(?:\(\s*\{(?:[^\}]*)\}\s*\)|\(\s*\{[^\}]*\}?\s*\))?(?:\.\s*t\(\s*\{(?:[^\}]*)\}\s*\))?\s*}}"
twig_trans_regex="{{\s*['\"]([^'\"]*)['\"]\s*\|\s*trans\s*}}|{%\s*trans\s*%}\s*((?:(?!{%\s*endtrans\s*%}).)*)\s*{%\s*endtrans\s*%}|{%\s*trans\s*%}\s*((?:(?!{%\s*plural\s*[^%]*%\s*).)*)\s*{%\s*plural\s*([^%]*)%\s*}\s*((?:(?!{%\s*endtrans\s*%}).)*)\s*{%\s*endtrans\s*%}"
js_regex="Drupal\.t\(\s*['\"](.+?)['\"]\s*\)"
js_regex_a="Drupal\.t\(\s*['\"](.+?)['\"],\s*{\s*['\"]@[^'\"]*['\"]:\s*[^,}]*[,\s}]"
js_regex_p="Drupal\.formatPlural\(\s*[^,]*\s*,\s*['\"](.+?)['\"]\s*,\s*['\"](.+?)['\"]\s*\)"


# Initialize the .pot file
output_file="translations.pot"

# Function to list all files used to extract translations
function list_files() {
  file_type=$1
  name_pattern=$2
  find . -type f -name "$name_pattern" -print0 | sort -z | xargs -0 -I {} echo "# {}"
}

# Get the custom module name from the .info.yml file
module_name=$(grep -oP "name:\s*\K[^\n]+" *.info.yml | tr '[:lower:]' '[:upper:]' | tr ' ' '_' | sed 's/_$//')

# Get the current date and time
current_date=$(date +"%Y-%m-%d %H:%M%z")

# Write the header to the .pot file
{
  echo "# \$Id$"
  echo "#"
  echo "# LANGUAGE translation of Drupal 9 module"
  echo "# Copyright YEAR NAME <EMAIL@ADDRESS>"
  echo "# Generated from files:"
  list_files "php" "*.php"
  list_files "php" "*.module"
  list_files "twig" "*.twig"
  list_files "js" "*.js" "! -name *.min.js"
  echo
  echo "# xx translation of $module_name"
  echo
  echo "\"Project-Id-Version: $module_name v1\\n\""
  echo "\"POT-Creation-Date: $current_date\\n\""
  echo "\"PO-Revision-Date: $current_date\\n\""
  echo "\"Last-Translator: NAME <EMAIL@ADDRESS>\\n\""
  echo "\"Language-Team: LANGUAGE <EMAIL@ADDRESS>\\n\""
  echo "\"MIME-Version: 1.0\\n\""
  echo "\"Content-Type: text/plain; charset=utf-8\\n\""
  echo "\"Content-Transfer-Encoding: 8bit\\n\""
  echo "\"Plural-Forms: nplurals=2; plural=(n > 1);\\n\""
  echo "\"X-Generator: Custom script v.01\\n\""
  echo
  echo "msgid \"\""
  echo "msgstr \"\""
  echo
} > "$output_file"

# Helper function to extract the strings and append them to the .pot file
function extract_strings() {
  file_type=$1
  regex=$2
  name_pattern=$3
  find . -type f -name "$name_pattern" ! -name "*.min.js" -print0 | while IFS= read -r -d '' file; do
    matches=$(perl -ne "print \"\$1\n\" while (/$regex/g);" "$file" | sort -u)
    if [ -n "$matches" ]; then
      echo -e "# ${file}\n# $(printf '%0.s-' $(seq 1 ${#file}))\n" >> "$output_file"
      echo "$matches" | while read -r match; do
        echo -e "msgid \"$match\"\nmsgstr \"\"\n" >> "$output_file"
      done
    fi
  done
}

# Helper function to remove duplicate msgid entries
function remove_duplicates() {
  awk '
    /^msgid/ {
      if (!msgid[$0]++)
        print
      else
        getline; getline
    }
    /^msgstr/ { print }
    !/^(msgid|msgstr)/ { print }
  ' "$output_file" > "$output_file.tmp" && mv "$output_file.tmp" "$output_file"
}

# Extract strings from each file type
extract_strings "php" "$php_t_regex" "*.php"
extract_strings "php" "$php_t_regex" "*.module"
extract_strings "twig" "$twig_t_regex" "*.twig"
#extract_strings "twig" "$twig_trans_regex" "*.twig"
extract_strings "js" "$js_regex" "*.js"
#extract_strings "js" "$js_regex_a" "*.js"
#extract_strings "js" "$js_regex_p" "*.js"

# Remove duplicate lines except for those that start with 'msgstr ""', except empty lines \n and line starting with #
remove_duplicates

# Remove double empty lines
sed -i '/^$/N;/^\n$/D' "$output_file"

Updated regex and extract_strings() function

Manage duplicate entry of the final result

change first msgid msgstr position