lib/google/cloud/bigtable/v2/bigtable_client.rb in google-cloud-bigtable-0.1.2 vs lib/google/cloud/bigtable/v2/bigtable_client.rb in google-cloud-bigtable-0.1.3
- old
+ new
@@ -243,11 +243,11 @@
# ReadRowsResponse documentation for details.
#
# @param table_name [String]
# The unique name of the table from which to read.
# Values are of the form
- # +projects/<project>/instances/<instance>/tables/<table>+.
+ # `projects/<project>/instances/<instance>/tables/<table>`.
# @param app_profile_id [String]
# This value specifies routing for replication. If not specified, the
# "default" application profile will be used.
# @param rows [Google::Bigtable::V2::RowSet | Hash]
# The row keys and/or ranges to read. If not specified, reads from all rows.
@@ -301,11 +301,11 @@
# mapreduces.
#
# @param table_name [String]
# The unique name of the table from which to sample row keys.
# Values are of the form
- # +projects/<project>/instances/<instance>/tables/<table>+.
+ # `projects/<project>/instances/<instance>/tables/<table>`.
# @param app_profile_id [String]
# This value specifies routing for replication. If not specified, the
# "default" application profile will be used.
# @param options [Google::Gax::CallOptions]
# Overrides the default settings for this call, e.g, timeout,
@@ -334,16 +334,16 @@
req = Google::Gax::to_proto(req, Google::Bigtable::V2::SampleRowKeysRequest)
@sample_row_keys.call(req, options)
end
# Mutates a row atomically. Cells already present in the row are left
- # unchanged unless explicitly changed by +mutation+.
+ # unchanged unless explicitly changed by `mutation`.
#
# @param table_name [String]
# The unique name of the table to which the mutation should be applied.
# Values are of the form
- # +projects/<project>/instances/<instance>/tables/<table>+.
+ # `projects/<project>/instances/<instance>/tables/<table>`.
# @param row_key [String]
# The key of the row to which the mutation should be applied.
# @param mutations [Array<Google::Bigtable::V2::Mutation | Hash>]
# Changes to be atomically applied to the specified row. Entries are applied
# in order, meaning that earlier mutations can be masked by later ones.
@@ -365,14 +365,14 @@
# require "google/cloud/bigtable/v2"
#
# bigtable_client = Google::Cloud::Bigtable::V2.new
# formatted_table_name = Google::Cloud::Bigtable::V2::BigtableClient.table_path("[PROJECT]", "[INSTANCE]", "[TABLE]")
#
- # # TODO: Initialize +row_key+:
+ # # TODO: Initialize `row_key`:
# row_key = ''
#
- # # TODO: Initialize +mutations+:
+ # # TODO: Initialize `mutations`:
# mutations = []
# response = bigtable_client.mutate_row(formatted_table_name, row_key, mutations)
def mutate_row \
table_name,
@@ -419,11 +419,11 @@
# require "google/cloud/bigtable/v2"
#
# bigtable_client = Google::Cloud::Bigtable::V2.new
# formatted_table_name = Google::Cloud::Bigtable::V2::BigtableClient.table_path("[PROJECT]", "[INSTANCE]", "[TABLE]")
#
- # # TODO: Initialize +entries+:
+ # # TODO: Initialize `entries`:
# entries = []
# bigtable_client.mutate_rows(formatted_table_name, entries).each do |element|
# # Process element.
# end
@@ -445,36 +445,36 @@
#
# @param table_name [String]
# The unique name of the table to which the conditional mutation should be
# applied.
# Values are of the form
- # +projects/<project>/instances/<instance>/tables/<table>+.
+ # `projects/<project>/instances/<instance>/tables/<table>`.
# @param row_key [String]
# The key of the row to which the conditional mutation should be applied.
# @param app_profile_id [String]
# This value specifies routing for replication. If not specified, the
# "default" application profile will be used.
# @param predicate_filter [Google::Bigtable::V2::RowFilter | Hash]
# The filter to be applied to the contents of the specified row. Depending
- # on whether or not any results are yielded, either +true_mutations+ or
- # +false_mutations+ will be executed. If unset, checks that the row contains
+ # on whether or not any results are yielded, either `true_mutations` or
+ # `false_mutations` will be executed. If unset, checks that the row contains
# any values at all.
# A hash of the same form as `Google::Bigtable::V2::RowFilter`
# can also be provided.
# @param true_mutations [Array<Google::Bigtable::V2::Mutation | Hash>]
- # Changes to be atomically applied to the specified row if +predicate_filter+
- # yields at least one cell when applied to +row_key+. Entries are applied in
+ # Changes to be atomically applied to the specified row if `predicate_filter`
+ # yields at least one cell when applied to `row_key`. Entries are applied in
# order, meaning that earlier mutations can be masked by later ones.
- # Must contain at least one entry if +false_mutations+ is empty, and at most
+ # Must contain at least one entry if `false_mutations` is empty, and at most
# 100000.
# A hash of the same form as `Google::Bigtable::V2::Mutation`
# can also be provided.
# @param false_mutations [Array<Google::Bigtable::V2::Mutation | Hash>]
- # Changes to be atomically applied to the specified row if +predicate_filter+
- # does not yield any cells when applied to +row_key+. Entries are applied in
+ # Changes to be atomically applied to the specified row if `predicate_filter`
+ # does not yield any cells when applied to `row_key`. Entries are applied in
# order, meaning that earlier mutations can be masked by later ones.
- # Must contain at least one entry if +true_mutations+ is empty, and at most
+ # Must contain at least one entry if `true_mutations` is empty, and at most
# 100000.
# A hash of the same form as `Google::Bigtable::V2::Mutation`
# can also be provided.
# @param options [Google::Gax::CallOptions]
# Overrides the default settings for this call, e.g, timeout,
@@ -488,11 +488,11 @@
# require "google/cloud/bigtable/v2"
#
# bigtable_client = Google::Cloud::Bigtable::V2.new
# formatted_table_name = Google::Cloud::Bigtable::V2::BigtableClient.table_path("[PROJECT]", "[INSTANCE]", "[TABLE]")
#
- # # TODO: Initialize +row_key+:
+ # # TODO: Initialize `row_key`:
# row_key = ''
# response = bigtable_client.check_and_mutate_row(formatted_table_name, row_key)
def check_and_mutate_row \
table_name,
@@ -523,11 +523,11 @@
#
# @param table_name [String]
# The unique name of the table to which the read/modify/write rules should be
# applied.
# Values are of the form
- # +projects/<project>/instances/<instance>/tables/<table>+.
+ # `projects/<project>/instances/<instance>/tables/<table>`.
# @param row_key [String]
# The key of the row to which the read/modify/write rules should be applied.
# @param rules [Array<Google::Bigtable::V2::ReadModifyWriteRule | Hash>]
# Rules specifying how the specified row's contents are to be transformed
# into writes. Entries are applied in order, meaning that earlier rules will
@@ -549,13 +549,13 @@
# require "google/cloud/bigtable/v2"
#
# bigtable_client = Google::Cloud::Bigtable::V2.new
# formatted_table_name = Google::Cloud::Bigtable::V2::BigtableClient.table_path("[PROJECT]", "[INSTANCE]", "[TABLE]")
#
- # # TODO: Initialize +row_key+:
+ # # TODO: Initialize `row_key`:
# row_key = ''
#
- # # TODO: Initialize +rules+:
+ # # TODO: Initialize `rules`:
# rules = []
# response = bigtable_client.read_modify_write_row(formatted_table_name, row_key, rules)
def read_modify_write_row \
table_name,