Sha256: d1bbd74a566a201ec1b51421512f86f6820de022499ece845128ea52f4518612
Contents?: true
Size: 1.9 KB
Versions: 9
Compression:
Stored size: 1.9 KB
Contents
require "yaml"
MetricFu.metrics_require do
[
"hotspots/hotspot",
"hotspots/analysis/analyzer_tables",
"hotspots/analysis/analyzed_problems",
"hotspots/analysis/rankings"
]
end
module MetricFu
class HotspotAnalyzer
COMMON_COLUMNS = %w{metric}
GRANULARITIES = %w{file_path class_name method_name}
def tool_analyzers
MetricFu::Hotspot.analyzers
end
def initialize(result_hash)
# we can't depend on the result
# returning a parsed yaml file as a hash?
result_hash = YAML::load(result_hash) if result_hash.is_a?(String)
setup(result_hash)
end
def hotspots
analyzed_problems.worst_items
end
def analyzed_problems
@analyzed_problems = MetricFu::HotspotAnalyzedProblems.new(@rankings, @analyzer_tables)
end
private
# TODO clarify each of these steps in setup
# extract into its own method
# remove unnecessary constants,
# turn into methods
def setup(result_hash)
# TODO There is likely a clash that will happen between
# column names eventually. We should probably auto-prefix
# them (e.g. "roodi_problem")
analyzer_columns = COMMON_COLUMNS + GRANULARITIES + tool_analyzers.map(&:columns).flatten
# though the tool_analyzers aren't returned, they are processed in
# various places here, then by the analyzer tables
# then by the rankings
# to ultimately generate the hotspots
@analyzer_tables = MetricFu::AnalyzerTables.new(analyzer_columns)
tool_analyzers.each do |analyzer|
analyzer.generate_records(result_hash[analyzer.name], @analyzer_tables.table)
end
@analyzer_tables.generate_records
@rankings = MetricFu::HotspotRankings.new(@analyzer_tables.tool_tables)
@rankings.calculate_scores(tool_analyzers, GRANULARITIES)
# TODO does it not need to return something here?
analyzed_problems
end
end
end
Version data entries
9 entries across 9 versions & 3 rubygems