Sha256: c4d4572ef1b9b605c2c3de947f08d33435aa6d75a22fb55b1c76aff8c26f7846
Contents?: true
Size: 1.72 KB
Versions: 7
Compression:
Stored size: 1.72 KB
Contents
# encoding: utf-8
# This file is distributed under New Relic's license terms.
# See https://github.com/newrelic/newrelic-ruby-agent/blob/main/LICENSE for complete details.
module NewRelic
module Agent
module Database
module PostgresExplainObfuscator
extend self
# Note that this regex can't be shared with the ones in the
# Database::Obfuscator class because here we don't look for
# backslash-escaped strings.
QUOTED_STRINGS_REGEX = /'(?:[^']|'')*'|"(?:[^"]|"")*"/
LABEL_LINE_REGEX = /^([^:\n]*:\s+).*$/.freeze
def obfuscate(explain)
# First, we replace all single-quoted strings.
# This is necessary in order to deal with multi-line string constants
# embedded in the explain output.
#
# Note that we look for both single or double quotes but do not
# replace double quotes in order to avoid accidentally latching onto a
# single quote character embedded within a quoted identifier (such as
# a table name).
#
# Note also that we make no special provisions for backslash-escaped
# single quotes (\') because these are canonicalized to two single
# quotes ('') in the explain output.
explain.gsub!(QUOTED_STRINGS_REGEX) do |match|
match.start_with?('"') ? match : '?'
end
# Now, mask anything after the first colon (:).
# All parts of the query that can appear in the explain output are
# prefixed with "<label>: ", so we want to preserve the label, but
# remove the rest of the line.
explain.gsub!(LABEL_LINE_REGEX, '\1?')
explain
end
end
end
end
end
Version data entries
7 entries across 7 versions & 1 rubygems