7 # Biopieces are command line scripts and uses OptionParser to parse command line
8 # options according to a list of casts. Each cast prescribes the long and short
9 # name of the option, the type, if it is mandatory, the default value, and allowed
10 # and disallowed values. An optional list of extra casts can be supplied, and the
11 # integrity of the casts are checked. Following the command line parsing, the
12 # options are checked according to the casts. Methods are also included for handling
13 # the parsing and emitting of Biopiece records, which are ASCII text records consisting
14 # of lines with a key/value pair seperated by a colon and a white space ': '.
15 # Each record is separated by a line with three dashes '---'.
17 # Initialize a Biopiece and write the status to file.
18 # Options are for testing purposes only.
19 def initialize(test=nil, input=STDIN, output=STDOUT)
24 @status.set unless @test
28 # Check the integrity of a list of casts, followed by parsion options from argv
29 # and finally checking the options according to the casts. Returns nil if argv
30 # is empty, otherwise an options hash.
31 def parse(argv, cast_list=[], script_path=$0)
32 casts = Casts.new(cast_list)
33 option_handler = OptionHandler.new(argv, casts, script_path, @test)
34 @options = option_handler.options_parse
37 # Open Biopiece input stream if not open and iterate over all Biopiece
38 # records in the stream.
40 @in = Stream::open(@options, mode="r", @input) unless @in.is_a? IO
44 @in.each_line do |line|
46 when /^([^:]+): (.*)$/
49 yield record unless record.empty?
52 raise "Bad record format: #{line}"
56 yield record unless record.empty?
61 alias :each :each_record
63 # Open Biopiece output stream if not open and puts record to the stream.
65 @out = Stream::open(@options, mode="w", @output) unless @out.is_a? IO
67 record.each do |key,value|
68 @out.print "#{key}: #{value}\n"
74 # Create a temporary directory inside the ENV["BP_TMP"] dir.
79 path = ENV["BP_TMP"] + "/" + [user, time + pid, pid, "bp_tmp"].join("_")
81 @status.set_tmpdir(path)
88 # Install signal handlers
89 %w( INT TERM QUIT ).each do |signal|
90 Signal.trap(signal) do
98 @in.close if @in.respond_to? :close
99 @out.close if @out.respond_to? :close
101 rescue Exception => exception
103 when 'INT' then exit_status = 'INTERRUPTED'
104 when 'TERM' then exit_status = 'TERMINATED'
106 exit_status = "DIED" unless exit_status =~ /INT|TERM|QUIT/
107 $stderr.puts exit_status
108 $stderr.puts exception.backtrace
112 FileUtils.remove_entry_secure(status.get_tmpdir)
113 status.log(exit_status)
121 # Error class for all exceptions to do with option casts.
122 class CastError < StandardError; end
125 # Class to handle casts of command line options. Each cast prescribes the long and
126 # short name of the option, the type, if it is mandatory, the default value, and
127 # allowed and disallowed values. An optional list of extra casts can be supplied,
128 # and the integrity of the casts are checked.
130 TYPES = %w[flag string list int uint float file file! files files! dir dir! genome]
131 MANDATORY = %w[long short type mandatory default allowed disallowed]
133 # Initialize cast object with an optional options cast list to which
134 # ubiquitous casts are added after which all casts are checked.
135 def initialize(cast_list=[])
136 @cast_list = cast_list
139 self.push *@cast_list
144 # Add ubiquitous options casts.
146 @cast_list << {:long => 'help', :short => '?', :type => 'flag', :mandatory => false, :default => nil, :allowed => nil, :disallowed => nil}
147 @cast_list << {:long => 'stream_in', :short => 'I', :type => 'files!', :mandatory => false, :default => nil, :allowed => nil, :disallowed => nil}
148 @cast_list << {:long => 'stream_out', :short => 'O', :type => 'file', :mandatory => false, :default => nil, :allowed => nil, :disallowed => nil}
149 @cast_list << {:long => 'verbose', :short => 'v', :type => 'flag', :mandatory => false, :default => nil, :allowed => nil, :disallowed => nil}
152 # Check integrity of the casts.
159 # Check if all mandatory keys are present in casts and raise if not.
161 @cast_list.each do |cast|
162 MANDATORY.each do |mandatory|
163 raise CastError, "Missing symbol in cast: '#{mandatory.to_sym}'" unless cast.has_key? mandatory.to_sym
168 # Check if all values in casts are valid.
170 @cast_list.each do |cast|
172 check_val_short(cast)
174 check_val_mandatory(cast)
175 check_val_default(cast)
176 check_val_allowed(cast)
177 check_val_disallowed(cast)
181 # Check if the values to long are legal and raise if not.
182 def check_val_long(cast)
183 unless cast[:long].is_a? String and cast[:long].length > 1
184 raise CastError, "Illegal cast of long: '#{cast[:long]}'"
188 # Check if the values to short are legal and raise if not.
189 def check_val_short(cast)
190 unless cast[:short].is_a? String and cast[:short].length == 1
191 raise CastError, "Illegal cast of short: '#{cast[:short]}'"
195 # Check if values to type are legal and raise if not.
196 def check_val_type(cast)
199 type_hash[type] = true
202 unless type_hash.has_key? cast[:type]
203 raise CastError, "Illegal cast of type: '#{cast[:type]}'"
207 # Check if values to mandatory are legal and raise if not.
208 def check_val_mandatory(cast)
209 unless cast[:mandatory] == true or cast[:mandatory] == false
210 raise CastError, "Illegal cast of mandatory: '#{cast[:mandatory]}'"
214 # Check if values to default are legal and raise if not.
215 def check_val_default(cast)
216 unless cast[:default].nil? or
217 cast[:default].is_a? String or
218 cast[:default].is_a? Integer or
219 cast[:default].is_a? Float
220 raise CastError, "Illegal cast of default: '#{cast[:default]}'"
224 # Check if values to allowed are legal and raise if not.
225 def check_val_allowed(cast)
226 unless cast[:allowed].is_a? String or cast[:allowed].nil?
227 raise CastError, "Illegal cast of allowed: '#{cast[:allowed]}'"
231 # Check if values to disallowed are legal and raise if not.
232 def check_val_disallowed(cast)
233 unless cast[:disallowed].is_a? String or cast[:disallowed].nil?
234 raise CastError, "Illegal cast of disallowed: '#{cast[:disallowed]}'"
238 # Check cast for duplicate long or short options names.
241 @cast_list.each do |cast|
242 raise CastError, "Duplicate argument: '--#{cast[:long]}'" if check_hash.has_key? cast[:long]
243 raise CastError, "Duplicate argument: '-#{cast[:short]}'" if check_hash.has_key? cast[:short]
244 check_hash[cast[:long]] = true
245 check_hash[cast[:short]] = true
251 # Class for parsing argv using OptionParser accordind to given casts.
252 # Default options are set, file glob expressions expanded, and options are
253 # checked according to the casts. Usage information is printed and exit called
256 REGEX_LIST = /^(list|files|files!)$/
257 REGEX_INT = /^(int|uint)$/
258 REGEX_STRING = /^(string|file|file!|dir|dir!|genome)$/
260 def initialize(argv, casts, script_path, test=nil)
263 @script_path = script_path
267 # Parse options from argv using OptionParser and casts denoting long and
268 # short option names. Usage information is printed and exit called.
269 # A hash with options is returned.
273 option_parser = OptionParser.new do |option|
274 @casts.each do |cast|
275 if cast[:type] == 'flag'
276 option.on("-#{cast[:short]}", "--#{cast[:long]}") do |o|
277 @options[cast[:long]] = o
279 elsif cast[:type] =~ REGEX_LIST
280 option.on( "-#{cast[:short]}", "--#{cast[:long]} A", Array) do |a|
281 @options[cast[:long]] = a
283 elsif cast[:type] =~ REGEX_INT
284 option.on("-#{cast[:short]}", "--#{cast[:long]} I", Integer) do |i|
285 @options[cast[:long]] = i
287 elsif cast[:type] =~ REGEX_STRING
288 option.on("-#{cast[:short]}", "--#{cast[:long]} S", String) do |s|
289 @options[cast[:long]] = s
291 elsif cast[:type] == 'float'
292 option.on("-#{cast[:short]}", "--#{cast[:long]} F", Float) do |f|
293 @options[cast[:long]] = f
296 raise ArgumentError, "Unknown option type: '#{cast[:type]}'"
301 option_parser.parse!(@argv)
304 print_usage_and_exit(true)
305 elsif print_usage_short?
316 # Given the script name determine the path of the wiki file with the usage info.
318 path = ENV["BP_DIR"] + "/bp_usage/" + File.basename(@script_path, ".rb") + ".wiki"
319 raise "No such wiki file: #{path}" unless File.file? path
323 # Check if full "usage info" should be printed.
324 def print_usage_full?
328 # Check if short "usage info" should be printed.
329 def print_usage_short?
332 elsif @options["stream_in"]
334 elsif @options["data_in"]
336 elsif wiki_path =~ /^(list_biopieces|list_genomes|list_mysql_databases|biostat)$/ # TODO get rid of this!
343 # Print usage info by Calling an external script 'print_wiki'
344 # using a system() call and exit. An optional 'full' flag
345 # outputs the full usage info.
346 def print_usage_and_exit(full=nil)
351 system("print_wiki --data_in #{wiki_path} --help")
353 system("print_wiki --data_in #{wiki_path}")
356 raise "Failed printing wiki: #{wiki_path}" unless $?.success?
362 # Set default options value from cast unless a value is set.
364 @casts.each do |cast|
366 @options[cast[:long]] = cast[:default] unless @options.has_key? cast[:long]
371 # Expands glob expressions to a full list of paths.
372 # Examples: "*.fna" or "foo.fna,*.fna" or "foo.fna,/bar/*.fna"
374 @casts.each do |cast|
375 if cast[:type] == 'files' or cast[:type] == 'files!'
376 if @options.has_key? cast[:long]
379 @options[cast[:long]].each do |path|
381 Dir.glob(path).each do |file|
382 files << file if File.file? file
389 @options[cast[:long]] = files
395 # Check all options according to casts.
397 @casts.each do |cast|
398 options_check_mandatory(cast)
399 options_check_int(cast)
400 options_check_uint(cast)
401 options_check_file(cast)
402 options_check_files(cast)
403 options_check_dir(cast)
404 options_check_allowed(cast)
405 options_check_disallowed(cast)
409 # Check if a mandatory option is set and raise if it isn't.
410 def options_check_mandatory(cast)
412 raise ArgumentError, "Mandatory argument: --#{cast[:long]}" unless @options.has_key? cast[:long]
416 # Check int type option and raise if not an integer.
417 def options_check_int(cast)
418 if cast[:type] == 'int' and @options.has_key? cast[:long]
419 unless @options[cast[:long]].is_a? Integer
420 raise ArgumentError, "Argument to --#{cast[:long]} must be an integer, not '#{@options[cast[:long]]}'"
425 # Check uint type option and raise if not an unsinged integer.
426 def options_check_uint(cast)
427 if cast[:type] == 'uint' and @options.has_key? cast[:long]
428 unless @options[cast[:long]].is_a? Integer and @options[cast[:long]] >= 0
429 raise ArgumentError, "Argument to --#{cast[:long]} must be an unsigned integer, not '#{@options[cast[:long]]}'"
434 # Check file! type argument and raise if file don't exists.
435 def options_check_file(cast)
436 if cast[:type] == 'file!' and @options.has_key? cast[:long]
437 raise ArgumentError, "No such file: '#{@options[cast[:long]]}'" unless File.file? @options[cast[:long]]
441 # Check files! type argument and raise if files don't exists.
442 def options_check_files(cast)
443 if cast[:type] == 'files!' and @options.has_key? cast[:long]
444 @options[cast[:long]].each do |path|
445 raise ArgumentError, "No such file: '#{path}'" unless File.file? path
450 # Check dir! type argument and raise if directory don't exist.
451 def options_check_dir(cast)
452 if cast[:type] == 'dir!' and @options.has_key? cast[:long]
453 raise ArgumentError, "No such directory: '#{@options[cast[:long]]}'" unless File.directory? @options[cast[:long]]
457 # Check options and raise unless allowed.
458 def options_check_allowed(cast)
459 if cast[:allowed] and @options.has_key? cast[:long]
461 cast[:allowed].split(',').each { |a| allowed_hash[a] = 1 }
463 raise ArgumentError, "Argument '#{@options[cast[:long]]}' to --#{cast[:long]} not allowed" unless allowed_hash.has_key? @options[cast[:long]]
467 # Check disallowed argument values and raise if disallowed.
468 def options_check_disallowed(cast)
469 if cast[:disallowed] and @options.has_key? cast[:long]
470 cast[:disallowed].split(',').each do |val|
471 raise ArgumentError, "Argument '#{@options[cast[:long]]}' to --#{cast[:long]} is disallowed" if val == @options[cast[:long]]
477 # Class for manipulating the execution status of Biopieces by setting a
478 # status file with a time stamp, process id, and command arguments. The
479 # status file is used for creating log entries and for displaying the
480 # runtime status of Biopieces.
482 # Write the status to a status file.
484 time0 = Time.new.strftime("%Y-%m-%d %X")
486 File.open(path, mode="w") { |file| file.puts [time0, ARGV.join(" ")].join(";") }
489 # Append the a temporary directory path to the status file.
490 def set_tmpdir(tmp_path)
491 status = File.open(path, mode="r").read.chomp
493 File.open(path, mode="w") do |fh|
494 fh << [status, tmp_path].join(";") + "\n"
498 # Extract the temporary directory path from the status file,
499 # and return this or nil if not found.
501 tmp_path = File.open(path, mode="r").read.chomp.split(";").last
503 File.directory?(tmp_path) ? tmp_path : nil
506 # Write the Biopiece status to the log file.
508 time1 = Time.new.strftime("%Y-%m-%d %X")
510 script = File.basename($0, ".rb")
512 stream = File.open(path)
513 time0, args, tmp_dir = stream.first.split(";")
516 elap = time_diff(time0, time1)
517 command = [script, args].join(" ")
518 log_file = ENV["BP_LOG"] + "/biopieces.log"
520 File.open(log_file, mode="a") { |file| file.puts [time0, time1, elap, user, exit_status, command].join("\t") }
523 # Delete status file.
530 # Path to status file
533 script = File.basename($0, ".rb")
535 path = ENV["BP_TMP"] + "/" + [user, script, pid, "status"].join(".")
538 # Get the elapsed time from the difference between two time stamps.
539 def time_diff(t0, t1)
540 Time.at((DateTime.parse(t1).to_time - DateTime.parse(t0).to_time).to_i).gmtime.strftime('%X')
546 # Open Biopieces output data stream for reading from stdin or a file
547 # specified in options["stream_in"] OR writing to stdout or a file
548 # specified in options["stream_out"].
549 def self.open(options, mode, stdio)
551 $stdin.tty? ? read(options["stream_in"]) : stdio
553 options["stream_out"] ? self.write(options["stream_out"], options["compress"]) : stdio
555 raise "Bad mode #{mode}"
561 # Opens a reads stream to a list of files.
563 self.zipped?(files) ? self.zread(files) : self.nread(files)
566 # Opens a write stream to a file and returns a _io_ object.
567 def self.write(file, zip=nil)
568 zip ? self.zwrite(file) : self.nwrite(file)
571 # Opens a list of gzipped files for reading and return an _io_ object.
572 def self.zread(files)
573 stdin, stdout, stderr = Open3.popen3("zcat " + files.join(' '));
579 # Opens a file for gzipped writing and return an _io_ object.
580 def self.zwrite(file)
581 stdin, stdout, stderr = Open3.popen3("gzip -f > #{file}")
587 # Opens a list of files for reading and return an _io_ object.
588 def self.nread(files)
589 stdin, stdout, stderr = Open3.popen3("cat " + files.join(' '));
595 # Opens a file for writing and return an _io_ object.
596 def self.nwrite(file)
597 File.open(file, mode="w")
600 # Test if a list of files are gzipped or not.
601 # Raises if files are mixed zipped and unzipped.
602 def self.zipped?(files)
606 type = `file #{file}`
608 if type =~ /gzip compressed/
609 type_hash[:gzip] = true
611 type_hash[:ascii] = true
615 raise "Mixture of zipped and unzipped files" if type_hash.size == 2
622 # Stuff to be done at begin such as setting the exit status
623 # and signal handlers
629 # Install signal handlers
630 %w( INT TERM QUIT ).each do |signal|
631 Signal.trap(signal) do
640 # Make sure that file streams are closed, tmpdir removed, and
641 # exit status is written to log file.
642 def do_at_exit(exit_status)
645 @in.close if @in.respond_to? :close
646 @out.close if @out.respond_to? :close
647 rescue Exception => exception
648 $stderr.puts "Exception caught!" # DEBUG
650 when 'INT' then exit_status = 'INTERRUPTED'
651 when 'TERM' then exit_status = 'TERMINATED'
653 exit_status = "DIED" unless exit_status =~ /INT|TERM|QUIT/
654 $stderr.puts exit_status
655 $stderr.puts exception.backtrace
659 FileUtils.remove_entry_secure(status.get_tmpdir)
660 status.log(exit_status)
662 puts "EXIT STATUS: #{exit_status}"
667 exit_status = do_at_begin
668 do_at_exit(exit_status)