modify the response format?

[leoborlot]

Hi guys. Any one know where the output template is set?
I want to try modify the output format to rescue and convert in JSON.

I want to modify this string format:

root@h01:/home/ubuntu# docker exec mailserver bash setup email list
* [email protected] ( 9.0K / 15G ) [0%]
    [ aliases -> [email protected] ]

[polarathene]

setup email list calls this script.

This is the part that will iterate through each line of postfix-accounts.cf to output the list results:

docker-mailserver/target/bin/listmailuser

Lines 21 to 26 in 811a769

	local ENTRY_TO_DISPLAY
	while read -r LINE; do
	ENTRY_TO_DISPLAY=$(_format_list_item "${LINE}")
	echo -e "* ${ENTRY_TO_DISPLAY}\n"
	done < <(_get_valid_lines_from_file "${DATABASE}")

Account, optional quota storage and aliases are then formatted here, with the final part concatenating into a single multi-line string per item:

docker-mailserver/target/bin/listmailuser

Lines 32 to 43 in 811a769

	local MAIL_ACCOUNT
	MAIL_ACCOUNT=$(echo "${LINE}" | cut -d'|' -f1)
	local WITH_QUOTA
	WITH_QUOTA=$(_quota_show_for "${MAIL_ACCOUNT}")
	local WITH_ALIASES
	WITH_ALIASES=$(_alias_list_for_account "${MAIL_ACCOUNT}")
	local ACCOUNT_ENTRY="${MAIL_ACCOUNT}"
	[[ -n ${WITH_QUOTA} ]] && ACCOUNT_ENTRY+=" ${WITH_QUOTA}"
	[[ -n ${WITH_ALIASES} ]] && ACCOUNT_ENTRY+="\n [ aliases -> ${WITH_ALIASES} ]"

Quota support queries Dovecot via the doveadm command, and parses that stdout response to get the data we're interested in:

docker-mailserver/target/bin/listmailuser

Lines 48 to 77 in 811a769

	function _quota_show_for() {
	local MAIL_ACCOUNT=${1}
	[[ ${ENABLE_QUOTAS} -ne 1 ]] && return 0
	local QUOTA_INFO
	# Matches a line where the 3rd column is `type='STORAGE'` - returning the next three column values:
	IFS=' ' read -r -a QUOTA_INFO <<< "$(doveadm quota get -u "${MAIL_ACCOUNT}" | tail +2 | awk '{ if ($3 == "STORAGE") { print $4" "$5" "$6 } }')"
	local CURRENT_SIZE SIZE_LIMIT PERCENT_USED
	# Format the extracted quota storage columns:
	CURRENT_SIZE="$(_bytes_to_human_readable_size "${QUOTA_INFO[0]}")"
	SIZE_LIMIT="$(_bytes_to_human_readable_size "${QUOTA_INFO[1]}")"
	PERCENT_USED="${QUOTA_INFO[2]}%"
	echo "( ${CURRENT_SIZE} / ${SIZE_LIMIT} ) [${PERCENT_USED}]"
	}
	function _bytes_to_human_readable_size() {
	# `-` represents a non-applicable value (eg: Like when `SIZE_LIMIT` is not set):
	if [[ ${1:-} == '-' ]]; then
	echo '~'
	# Otherwise a value in KibiBytes (1024 bytes == 1k) is expected (Dovecots internal representation):
	elif [[ ${1:-} =~ ^[0-9]+$ ]]; then
	# kibibytes to bytes, converted to approproate IEC unit (eg: MiB):
	echo $(( 1024 * ${1} )) | numfmt --to=iec
	else
	_exit_with_error "Supplied non-number argument '${1:-}' to '_bytes_to_human_readable_size()'"
	fi
	}

Alias support is a bit tricky, using grep and sed commands to find aliases in postfix-aliases.cf for the account:

docker-mailserver/target/bin/listmailuser

Lines 79 to 102 in 811a769

	# Returns a comma delimited list of aliases associated to a recipient (ideally the recipient is a mail account):
	function _alias_list_for_account() {
	local GREP_OPTIONS
	local MAIL_ACCOUNT=${1}
	# postfix-virtual.cf sample lines:
	#
	# [email protected] [email protected]
	# [email protected] [email protected],[email protected]
	# [email protected] [email protected],[email protected],[email protected]
	# [email protected] [email protected],[email protected]
	GREP_OPTIONS=(
	--ignore-case
	--extended-regexp
	-e "\s${MAIL_ACCOUNT}($|,)" # match first and second sample line
	-e ",${MAIL_ACCOUNT}($|,)" # match third and fourth sample line
	"${DATABASE_VIRTUAL}"
	)
	if grep --quiet --no-messages "${GREP_OPTIONS[@]}"; then
	grep "${GREP_OPTIONS[@]}" | awk '{print $1;}' | sed ':a;N;$!ba;s/\n/, /g'
	fi
	}

---

Alternatively, you could parse the output of setup email list?

If you do rewrite a JSON output, you may have some luck contributing it via PR as an alternative setup email list --json, but it must correctly list the data.

[leoborlot]

First now, I resolve with RE, break line in '*', and use RE to split the components, and solve.
Next time I will try to build this job and create a more solid output.
RE is good but, a . on the road break all.

If any one want to try my soluttion in python, this is the code. (the variables was in pt-br, next time bring more international)

  try:
       command = 'bash setup email list'
       result = execute_docker_command(command)
       partes = result.split('* ')
       padrao = r'(\S+) \(\s*([\d\.]+[KMG]?|~) / ([\d\.]+[KMG]?|~) \) \[([\d\.]+)%\](?:\s+\[ aliases -> (\S+(?:, \S+)*) \])?'
       dados = []

       for parte in partes[1:]:  # Ignorar a primeira parte vazia após a divisão
           correspondencia = re.findall(padrao, parte, re.M)
           if correspondencia:
               email, usado, total, porcentagem, aliases = correspondencia[0]
               aliases = {'aliases': aliases.split(', ')} if aliases else {}

               # Crie um dicionário com os dados extraídos
               item = {
                   'dominio': re.search(r'@(.+)', email).group(1) if email else None,
                   'email': email,
                   'usado': usado,
                   'total': total,
                   'porcentagem': porcentagem,
                   **aliases  # Adicione o dicionário de aliases, se existir
               }
               dados.append(item)
               

This is the output in JSON