Files
ebusd/contrib/etc/ebusd/mqtt-integration.cfg
T

207 lines
12 KiB
INI

# Configuration file for ebusd MQTT integration.
# This file allows flexible construction of MQTT topics and payloads depending on message and field definitions as well
# as global status topics.
# It is a set of named variables (or constants) with the name on the left of an equal sign and the value on the right,
# where only the value is allowed to contain references to other variables.
# A reference to a variable is escaped by a leading "%" character in front of the variable name consisting of only
# alphabetic characters or the underscore. A double percent character "%%" is replaced by a single "%" in the value.
# Curly braces may be used to separate a variable explicitly from the rest, e.g. as "%{version}" instead of "%version".
# Every variable without an underscore in the name is automatically made available as well with an uppercase name and a
# normalized value composed of only alphanumeric characters or underscore (i.e. suitable for being part of a topic).
# When using "?=" instead of only the equal sign in a variable definition, the variable value is set to empty when one
# of the variables referred to is empty or missing.
# The variable replacements are constructed for each message and/or field (depending on whether the topic is field
# specific or not) in the following order:
# 1. predefined constants from ebusd, i.e.
# %version the ebusd version number.
# %prefix the invariable prefix part of the "--mqtttopic" configuration option, defaults to "ebusd/".
# %prefixn the same as %prefix but without trailing slashes or underscores.
# 2. variables referring to constants only (directly or indirectly).
# 3. circuit and message specific variables:
# %circuit the circuit name.
# %name the message name.
# %priority the message poll priority (or 0).
# %level the message access level (or empty).
# %messagecomment the message comment.
# %direction the message direction ("r", "w", "u", or "uw").
# %topic the message update topic built from the mqtttopic configuration option (only if the topic is not field
# specific) and/or the %topic variable defined here.
# 4. the direction mapped from variables named "direction_map-%direction" with the actual direction in the suffix.
# Available as "%direction_map".
# 5. the field type mapped from variables named "type_map-%direction-%type" or "type_map-%type" as fallback with the
# field type and the optional direction in the suffix. The variable including the direction is evaluated first and
# if it is missing or the result is empty, the variable excluding the direction is used instead.
# This is also an implicit field type filter as missing or empty mappings are not published at all.
# Available as "%type_map".
# 6. other field specific variables for each field (only once for each message if the topic is field specific):
# %index the numeric field index (excluding ignored fields).
# %field the field name and key for JSON objects (equals the field index if no name is defined or the names are not
# unique in the message).
# %fieldname the field name (and nothing else like in %field, might be empty though!)
# %type the field type (one of "number", "list", "string", "date", "time", or "datetime").
# %basetype the base data type ID (e.g. "UCH").
# %comment the field comment (if any).
# %unit the field unit (if any).
# %min, %max, and %step the minimum/maximum possible value and step value for number fields.
# %topic the field (or message) update topic built from the mqtttopic configuration option and/or the %topic variable
# defined here.
# 7. optional field type switch in "%type_switch" using the value of the variable named "type_switch-by" and matching it
# against the value lines of the variable named "type_switch-%type" with the field type in the suffix.
# The value of such a variable needs to be a constant list of string pairs with one pair per line. Each pair is
# separated by "=" and the left part is the value set to the "type_switch" variable when the wildcard pattern in the
# right part matches the value of the "type_switch-by" value. The list is traversed from top to bottom with the
# first match stopping the evaluation. Optionally, "type_switch-names" can be used to define a list of variable
# names to be set in addition to the "type_switch" variable, in which case a list of variable names (comma
# separated) is expected as "type_switch-names" value and the appropriate values to set on the left side of each
# match list entry.
# Available as "%type_switch" and optionally by the set of fields as defined in "%type_switch-names".
# 8. optional field type part mapping in "%type_part" for each field type in variables named "type_part-%type" with the
# field type in the suffix (or the value of the "type_part-by" variable).
# 9. if the %fields_payload variable is used, then it is set to the concatenation of all fields of the message:
# %field_payload is expected to build a single field payload.
# %field-separator is the separator placed between consecutive field payloads.
# %fields_payload is set to the concatenation of all fields of the message, separated by %field-separator.
# 10. %definition-topic, %definition-payload, and %definition-retain to build the message definition topic and payload
# and determine the retain value.
# Before constructing the variable replacements, the messages and fields can optionally be filtered in order to use
# certain message definitions only. All of these criteria are applied, so they are basically combined by "AND".
# Filters with a "partial match" mention in the documentation allow using a limited glob/regex inspired case insensitive
# matching with the following options:
# - "|" allows defining alternatives, e.g. "a|b" matches "but" as well as "all".
# - "^" matches the beginning of the input, e.g. "^al" matches "al" but not "hal".
# - "$" matches the end of the input, e.g. "al$" matches "hal" but not "all".
# - "*" matches a single arbitrary length wildcard part in the middle (can only be used once per filter/alternative),
# e.g. "^a*l$" matches "all" but not "always".
# include only messages having data sent at least once (only checked for passive or read messages, not for active write)
# when set to 1. If set to >1, then all messages passing the other filter criteria (including active read messages) will
# automatically be set to have a poll priority of at most this value, so these are automatically being polled.
filter-seen = 1
# include only messages having a priority less than or equal to the specified value.
#filter-priority =
# include only messages having the specified circuit (partial match, alternatives and wildcard supported).
#filter-circuit =
# exclude messages having the specified circuit (partial match, alternatives and wildcard supported).
#filter-non-circuit =
# include only messages having the specified name (partial match, alternatives and wildcard supported).
#filter-name =
# exclude messages having the specified name (partial match, alternatives and wildcard supported).
#filter-non-name =
# include only messages having the specified level (partial match, alternatives and wildcard supported).
# Note: This is a filter on top of the messages already filtered implicitly for the ebusd "mqtt" user (if any).
# Note: Since the empty string matches all levels, an explicit check for empty string (with "^$") needs to be used for
# including only messages without a level.
filter-level = ^$
# include only messages having the specified direction ("r", "w", "u", or "uw". partial match, alternatives and wildcard supported).
filter-direction = r|u
# include only fields having the specified name (partial match, alternatives and wildcard supported).
#filter-field =
# exclude fields having the specified name (partial match, alternatives and wildcard supported).
filter-non-field = ^sensor$
# the field type mapped from variables named "type_map-%direction-%type" or "type_map-%type" as fallback with the
# field type and the optional direction in the suffix. The variable including the direction is evaluated first and
# if it is missing or the result is empty, the variable excluding the direction is used instead.
# This is also an implicit field type filter as missing or empty mappings are not published at all.
type_map-number = number
type_map-list = string
type_map-string = string
type_map-date = string
type_map-time = string
type_map-datetime = string
# field type switch designator, see below.
#type_switch-by = %name%field,%unit
# field type switch variables names to use in addition to %type_switch in case of multiple keys (separated by comma).
#type_switch-names = type_topic,type_class,type_state
# field type switch for each field type and optionally direction (between dash before the field type) available as
# %type_switch (as well as the variable names defined in "type_switch-names" if any).
# The value needs to be a constant list of string pairs with one pair per line. Each pair is separated by "=" and the
# left part is the value set to the type_switch variable (as well as the variable names defined in %type_switch-names)
# when the wildcard string in the right part matched the "type_switch-by" value. The list is traversed from top to
# bottom stopping at the first match. If direction specific definitions exist, these are traversed first. If no line
# matches at all, the variable(s) are set to the empty string.
#type_switch-w-number =
# field type part suffix to use instead of the field type itself, see below.
#type_part-by = %type_topic
# field type part mappings for each field type (or the "type_part-by" variable value) in the suffix (available as
# %type_part).
#type_part-number = ,
# optional format string for converting a fields value list into %field_values.
# "$value" and "$text" are being replaced by the corresponding part.
#field_values-entry = $text
# optional separator for concatenating of field value list items.
#field_values-separator = ,
# optional prefix for surrounding field value list items.
#field_values-prefix =
# optional suffix for surrounding field value list items.
#field_values-suffix =
# the field specific part (evaluated after the message specific part).
#field_payload = %type_part
# the message definition config topic, payload, and retain setting.
definition-topic = %prefixn/config/%CIRCUIT/%NAME/%FIELD
definition-payload = {
"name":"%prefixn %circuit %name %field",
"topic":"%topic"
}
#definition-retain = 0
# the message value topic (if other than the default). If set here, it will be replaced by the "--mqtttopic"
# configuration option only if that one contains at least one variable. If the the "--mqtttopic" configuration option
# does not contain any variable, it is taken as prefix and can be used here as well.
#topic=%prefix/%circuit/%name
# the common global config topic, payload, and retain setting (used by running, version, signal, uptime, updatecheck,
# and scan if not otherwise defined explicitly).
def_global-topic = %prefixn/config/global/%FIELD
def_global-payload = {
"name":"%prefixn global %field",
"topic":"%topic"
}
#def_global-retain = 0
# individual global running, version, signal, uptime, updatecheck, and scan config topic, payload, and retain setting.
# a secondary update check for the eBUS device (consuming the same updatecheck topic) can be set up, which will only be
# used if an enhanced eBUS device supporting extra info is present (does not make use of the common global defaults).
#def_global_running-...
#def_global_version-...
#def_global_signal-...
#def_global_uptime-...
#def_global_updatecheck-...
#def_global_updatecheck_device-...
#def_global_scan-...
# the topic and payload to listen to in order to republish all config messages.
#config_restart-topic =
#config_restart-payload =