Implemented support for Aruba Business DNS API (#7042)

* Add support for arubabusiness api

* Fix formatting

* record names are always converted to lowercase

* Docs

* remove leftover unconditional authentication call
lowercase
urlencoded body + x-www-form-urlencoded content-type
cleanup header variables
cleanup typos
grammar

* Strengthen _ab_rest failure checks
Properly process parallel lists in _ab_dns_record_id
Remove hard fails when a txt record already exists

* fix json parsing

* Fix formatting

---------

Co-authored-by: Manwe-Sulimo <blarghonaut@gmail.com>
This commit is contained in:
Manwe Sulimo
2026-07-02 07:06:41 +02:00
committed by GitHub
parent b039ff3087
commit cb40b47a4e

490
dnsapi/dns_arubabusiness.sh Normal file
View File

@@ -0,0 +1,490 @@
#!/usr/bin/env sh
# shellcheck disable=SC2034
dns_arubabusiness_info='ArubaBusiness
Site: business.aruba.it
Docs: github.com/acmesh-official/acme.sh/wiki/dnsapi#dns_arubabusiness
Options:
AB_Key Your ArubaBusiness API Key
AB_User Your account user
AB_Pass Your account password
'
#
# A word of warning: as of this writing, api.arubabusiness.it only supports oauth authentication using the "password" grant type.
# If you are REALLY sure you want to use it, it would be wise set up a dedicated technical user without administrative privileges
#
ARUBABUSINESS_API='https://api.arubabusiness.it'
######## Public functions ########
#
# Usage: dns_arubabusiness_add _acme-challenge.www.domain.com aaaabbbbcccc111122223333
#
# Add a new TXT record whose name and value match the given domain and value
#
# Variables
# _full_domain: $1 - the name of the TXT record
# _txt_value: $2 - the value of the TXT record
# _body
# dns_details
# domain_id
# dns_record_id
# response
#
dns_arubabusiness_add() {
_full_domain=$1
_txt_value=$2
if ! _ab_authenticate; then
return 1
fi
if ! _ab_domain_id "$_full_domain"; then
return 1
fi
if _ab_dns_record_id "$_full_domain" "$_txt_value" "$dns_details"; then
# This is very unlikely, but allow the process to use the existing record
_info "A TXT record with name: $_full_domain and value: $_txt_value already exists (id: $dns_record_id)"
return 0
fi
_body="{ \"IdDomain\": $domain_id, \"Type\": \"TXT\", \"Name\": \"$_full_domain\", \"Content\": \"\\\"$_txt_value\\\"\" }"
_debug "Adding TXT record with name: $_full_domain and value: $_txt_value"
if ! _ab_rest POST "api/domains/dns/record" "$_body" || ! _contains "$response" "DomainId"; then
_err "Failed to add TXT record with name: $_full_domain"
return 1
fi
_info "Sleeping 10 seconds to let ArubaBusiness do its magic"
_sleep 10
# Refresh dns details and check that the record was really added
if ! _ab_dns_details "$root_domain"; then
return 1
fi
if ! _ab_dns_record_id "$_full_domain" "$_txt_value" "$dns_details"; then
# This should never happen
_err "The TXT record with name: $_full_domain was not set"
_err "Please check that the dns records are clean"
return 1
fi
_info "Added TXT record with id: $dns_record_id"
return 0
}
#
# Usage: dns_arubabusiness_rm _acme-challenge.www.domain.com aaaabbbbcccc111122223333
#
# Remove the TXT record whose name and value match the given domain and value
#
# Variables
# _full_domain: $1 - the name of the TXT record
# _txt_value: $2 - the value of the TXT record
# dns_details
# dns_record_id
#
dns_arubabusiness_rm() {
_full_domain=$1
_txt_value=$2
if ! _ab_authenticate; then
return 1
fi
if ! _ab_domain_id "$_full_domain"; then
return 1
fi
if ! _ab_dns_record_id "$_full_domain" "$_txt_value" "$dns_details" || [ -z "$dns_record_id" ]; then
_err "Could not retrieve the record id for: $_full_domain"
return 1
fi
_debug "Deleting TXT record: $dns_record_id"
if ! _ab_rest DELETE "api/domains/dns/record/$dns_record_id" || ! _contains "$response" "DomainId"; then
_err "Failed to delete TXT record: $dns_record_id"
return 1
fi
_info "Deleted TXT record: $dns_record_id"
return 0
}
######## Private functions ########
#
# Usage: _ab_domain_id _acme-challenge.www.domain.com
#
# Split the input domain into subdomain + root domain and get the id of the root domain
#
# Variables
# _full_domain: $1 - the domain whose root needs to be extracted
# _domain_sections
# _current_index
# _candidate_subdomain
# _candidate_domain
# sub_domain
# root_domain
# domain_id
# dns_details: a json containing all dns records registered on the root domain
#
# Example
# _get_root _acme-challenge.www.domain.com
#
# Should return
# sub_domain=_acme-challenge.www
# root_domain=domain.com
# domain_id=123123123123
# dns_details="{JSON_CONTENT}"
#
_ab_domain_id() {
_full_domain=$1
_info "Attempting to retrieve root domain details for: $_full_domain"
_domain_sections=$(_math "$(printf "%s" "$_full_domain" | tr '.' '\n' | wc -l)" + 1)
if [ "$_domain_sections" -lt 1 ]; then
_err "Invalid input $_full_domain"
return 1
fi
_current_index=1
while true; do
_candidate_subdomain=$(if [ "$_current_index" = "1" ]; then printf ""; else printf "%s" "$_full_domain" | cut -d . -f 1-"$(_math "$_current_index" - 1)"; fi)
_candidate_domain=$(printf "%s" "$_full_domain" | cut -d . -f "$_current_index"-"$_domain_sections")
if ! _ab_dns_details "$_candidate_domain"; then
_debug2 "Could not fetch dns details for: $_candidate_domain"
_current_index=$(_math "$_current_index" + 1)
# Fail if there are no candidates left
if [ "$_current_index" -gt "$_domain_sections" ]; then
_err "Could not determine the root domain for: $_full_domain"
return 1
fi
else
sub_domain="$_candidate_subdomain"
root_domain="$_candidate_domain"
# Extract the domain id, which is an integer and contains no commas
domain_id="$(printf "%s" "$dns_details" | _egrep_o '"Id":[^,]*' | _head_n 1 | cut -d : -f 2 | tr -d ' "')"
if [ -z "$domain_id" ]; then
_err "Could not determine the domain id for: $root_domain"
return 1
fi
_debug "Retrieved root domain id: $domain_id"
return 0
fi
done
}
#
# Usage: _ab_dns_record_id _acme-challenge.www.domain.com "aaaabbbbcccc111122223333" "{JSON_CONTENT}"
#
# Extract the record id of the first TXT record whose name and content match the input values
#
# Variables
# _record_name: $1
# _txt_value: $2
# _dns_details: $3 - the json returned by a previous call to '_ab_dns_details() $root_domain'
# _record_ids
# _record_names
# _record_types
# _record_contents
# _record_ids_count
# _record_names_count
# _record_types_count
# _record_contents_count
# _i
# dns_record_id
#
# Notes
# TXT correspond to record type 5
# ArubaBusiness appends a terminating dot (.) to the record name
# The content field may contain the following character sequence: \"
# All record names are always converted to lowercase
#
_ab_dns_record_id() {
_record_name=$1
_txt_value=$2
_dns_details=$3
_record_name_lowercase=$(printf "%s" "$_record_name" | _lower_case)
# Extract the record ids, which are integers and contain no commas, colons or spaces
# The first id is skipped because it refers to the domain id
_record_ids=$(printf "%s" "$_dns_details" | sed 's/"Id":/\n"Id":/g' | _egrep_o '"Id":[^,]*' | _tail_n +2 | cut -d : -f 2 | tr -d ' ' | tr '\n' ' ')
# Extract the record names, which are strings but cannot contain commas, colons, spaces and quotes
# The first name is skipped because it refers to the domain name
_record_names=$(printf "%s" "$_dns_details" | sed 's/"Name":/\n"Name":/g' | _egrep_o '"Name":[^,]*' | _tail_n +2 | cut -d : -f 2 | tr -d ' "' | tr '\n' ' ')
# Extract the record types, which are integers (except for the first one) and contain no commas, colons or spaces
# The first type is skipped because it refers to the domain type
_record_types=$(printf "%s" "$_dns_details" | sed 's/"Type":/\n"Type":/g' | _egrep_o '"Type":[^,]*' | _tail_n +2 | cut -d : -f 2 | tr -d ' ' | tr '\n' ' ')
# Extract the record contents, which are strings and may contain no quotes except for TXT records, which must be delimited by two \" literals
# Note: There is no domain related entry here
# Note: A " character is appended at the end of each content to make it easier to process the list later
_record_contents=$(printf "%s" "$_dns_details" | sed 's/"Content":/\n"Content":/g' | sed 's/\\"//g' | _egrep_o '"Content": *"[^"]*"' | cut -d : -f 2- | sed -n 's/"\(.*\)"/\1/p' | tr '\n' '#')
_info "IDS: $_record_ids"
_info "NAMES: $_record_names"
_info "TYPEs: $_record_types"
_info "CONTENTS: $_record_contents"
_record_ids_count=$(printf "%s" "$_record_ids" | tr ' ' '\n' | wc -l)
_record_names_count=$(printf "%s" "$_record_names" | tr ' ' '\n' | wc -l)
_record_types_count=$(printf "%s" "$_record_types" | tr ' ' '\n' | wc -l)
_record_contents_count=$(printf "%s" "$_record_contents" | tr '#' '\n' | wc -l)
_info "Ids: $_record_ids_count, names: $_record_names_count, types: $_record_types_count, contents: $_record_contents_count"
if [ "$_record_ids_count" != "$_record_names_count" ] || [ "$_record_ids_count" != "$_record_types_count" ] || [ "$_record_ids_count" != "$_record_contents_count" ]; then
_err "Failed to parse record elements. Ids: $_record_ids_count, names: $_record_names_count, types: $_record_types_count, contents: $_record_contents_count"
return 1
fi
_info "Looking for a TXT record matching inputs - name: $_record_name_lowercase value: $_txt_value"
_i=1
while [ "$_i" -le "$_record_ids_count" ]; do
_current_name=$(printf "%s" "$_record_names" | cut -d " " -f "$_i")
_current_type=$(printf "%s" "$_record_types" | cut -d " " -f "$_i")
_current_content=$(printf "%s" "$_record_contents" | cut -d "#" -f "$_i")
if [ "$_record_name_lowercase." = "$_current_name" ] && [ "5" = "$_current_type" ] && [ "$_txt_value" = "$_current_content" ]; then
dns_record_id=$(printf "%s" "$_record_ids" | cut -d " " -f "$_i")
_info "Found matching record with id: $dns_record_id"
return 0
else
_debug2 "Record does not match - type: '$_current_type' name: '$_current_name' value: '$_current_content'; Expected '$_record_name_lowercase.' '5' '$_txt_value'"
fi
_i=$(_math "$_i" + 1)
done
_debug2 "No matching record was found in $_dns_details"
return 1
}
#
# Usage: _ab_dns_details domain.com
#
# Retrieve dns info for the given input domain
#
# Variables
# _domain: $1
# dns_details: the json returned by the call to $ARUBABUSINESS_API/api/domains/dns/$_domain/details (if return status is 0)
# response
#
_ab_dns_details() {
_domain=$1
if ! _ab_rest GET "api/domains/dns/$_domain/details" || ! _contains "$response" "DomainId"; then
return 1
fi
dns_details="$response"
return 0
}
#
# Usage: _ab_authenticate
#
# Read account conf, update domain conf and perform user authentication to acquire an access token
#
# Variables
# AB_Key
# AB_User
# AB_Pass
# AB_Token
#
_ab_authenticate() {
AB_Key="${AB_Key:-$(_readaccountconf_mutable AB_Key)}"
AB_User="${AB_User:-$(_readaccountconf_mutable AB_User)}"
AB_Pass="${AB_Pass:-$(_readaccountconf_mutable AB_Pass)}"
if [ -z "$AB_Key" ] || [ -z "$AB_User" ] || [ -z "$AB_Pass" ]; then
AB_Key=""
AB_User=""
AB_Pass=""
_err "Either the ArubaBusiness API key, the user or the password has not been defined yet."
_err "Please configure them and try again."
return 1
fi
_saveaccountconf_mutable AB_Key "$AB_Key"
_saveaccountconf_mutable AB_User "$AB_User"
_saveaccountconf_mutable AB_Pass "$AB_Pass"
if ! _ab_get_token || [ -z "$AB_Token" ]; then
_err "Failed to acquire an access token"
return 1
fi
return 0
}
#
# Usage: _ab_get_token
#
# Try acquiring a temporary access token. The token should have a 24h lifespan
#
# Variables
# _ab_user_enc
# _ab_pass_enc
# _ab_authdata
# AB_User
# AB_Pass
# AB_Token
# response
# _H2
#
_ab_get_token() {
_ab_user_enc=$(printf "%s" "$AB_User" | _url_encode)
_ab_pass_enc=$(printf "%s" "$AB_Pass" | _url_encode)
_ab_authdata="grant_type=password&username=$_ab_user_enc&password=$_ab_pass_enc"
_H2="Content-Type: application/x-www-form-urlencoded"
if ! _ab_rest POST "auth/token" "$_ab_authdata" || ! _contains "$response" "access_token"; then
_err "Authentication failure"
return 1
fi
AB_Token="$(printf "%s" "$response" | _egrep_o '"access_token":"[^\"]*"' | cut -d : -f 2 | tr -d '"')"
if [ -z "$AB_Token" ]; then
_err "Could not extract access token"
return 1
fi
_debug "Acquired access token"
return 0
}
#
# Usage: _ab_rest POST "example/endpoint" "password=123"
#
# Perform a REST request using the given method, endpoint and data
#
# Variables
# _method: $1 - The http method
# _endpoint: $2 - The api path (relative to $ARUBABUSINESS_API)
# _data: $3 - The body of the request (optional)
# _key_trimmed
# _token_trimmed
# _ret_code
# AB_Key
# AB_Token
# ARUBABUSINESS_API
# _H1
# _H2
# _H3
# _H4
#
_ab_rest() {
_method=$1
_endpoint="$2"
_data="$3"
_key_trimmed=$(printf "%s" "$AB_Key" | tr -d '"')
_token_trimmed=$(printf "%s" "$AB_Token" | tr -d '"')
_H1="Accept: application/json"
if [ -z "$_H2" ]; then
# Default to application/json
_H2="Content-Type: application/json"
fi
if [ "$_key_trimmed" ]; then
_H3="Authorization-Key: $_key_trimmed"
else
_err "Missing Api Key"
_ab_cleanup_headers
return 1
fi
if [ "$_token_trimmed" ]; then
_H4="Authorization: Bearer $_token_trimmed"
else
_debug "No access token set"
fi
if [ "$_method" != "GET" ]; then
response="$(_post "$_data" "$ARUBABUSINESS_API/$_endpoint" "" "$_method")"
else
response="$(_get "$ARUBABUSINESS_API/$_endpoint")"
fi
_ret_code=$?
if [ "$_ret_code" = "0" ] && _ab_call_is_success; then
# Normalize the json response
response="$(printf "%s" "$response" | _normalizeJson)"
_ret_code=0
else
_err "Failed to call endpoint: $_endpoint"
_ret_code=1
fi
_ab_cleanup_headers
return $_ret_code
}
#
# Usage: _ab_cleanup_headers
#
# Unset header variables to avoid interfering with other calls
#
# Variables
# _H1
# _H2
# _H3
# _H4
#
_ab_cleanup_headers() {
# Cleanup request headers
unset _H1 _H2 _H3 _H4 _H5
# Cleanup response headers
if [ -f "$HTTP_HEADER" ]; then
: >"$HTTP_HEADER"
fi
}
#
# Usage: _ab_call_is_success
#
# Check whether a call's response http status is one of 200, 201, 202 or 204 (other 2xx are not handled)
#
# Variables
# _status
# _http_status
# _success_http_codes
# HTTP_HEADER
#
_ab_call_is_success() {
_success_http_codes="200 201 202 204"
if [ -f "$HTTP_HEADER" ]; then
_http_status=$(_egrep_o "^HTTP[\/0-9. ]*" <"$HTTP_HEADER" | _head_n 1 | cut -d " " -f 2)
for _status in $_success_http_codes; do
if [ "$_status" = "$_http_status" ]; then
return 0
fi
done
fi
return 1
}