parse_fasta 1.9.1 → 1.9.2

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    MDAyZjJjNDJlMGZkM2Y2YmZmOTNmZTIwOTE2ZmVlMTRlYjhiNThkNA==
+    NmM5ZWYwOGM5YWIxMzU2YjBmZTk4Y2I5YzI0NjY0MzUwM2YwMjgyOA==
   data.tar.gz: !binary |-
-    ZmQxMDI5MDI4MjEyN2UyMjQ3ZGQwNDU2YzZhYmI0ZTNjMGQ2ZWFiOQ==
+    NzI2NDY1MWZmYmUwNDUxMTk2MmI4YjgwYWVlYjcyZDI4MDUzMzk4NA==
 SHA512:
   metadata.gz: !binary |-
-    NDZjNTk2OTUzYmM1MzE3MmJmOWJlMGM0NjQxY2RlODRiY2RiOTNlNWRiNmM1
-    NTA3OWNmMzdiZTRlNmVlZWZiOTU1ZjhkZjE4YTA3ZjQ5OTFhOWQ3NjViOTM3
-    N2IwY2I5YjI1MDIwMDczY2M2MzNmZTE4OTdiMzI0NWQ4ODcwMGE=
+    ODY1ZTQ1MzU4MTc2MDhhMjA0OThiYzM4Yzk4YjJiZjU4ZGY4MGM5NTRjYTE5
+    OWZkODk0M2ZmODE5ODY1MjE3NTQ5MzgyNTFjMTk2NzU2NGVjN2NkNGUzYzA3
+    ODliNjRlOGJjOGJhNjhlMWZmMmU1NjkyMjgwNzAyODQ1MDExOTI=
   data.tar.gz: !binary |-
-    NmNkZTI0NzE3NjU5M2VmNzgzYTk0MDM1NGM2NTdmNDA2YWY0ZmY2OWRhNTE4
-    ODE1MzIxYTcyNzhhMTJmMTc5NmE5MjI3OWEyYjUwMzM5NzJhYzAxZjJkYTE1
-    YTNmMTcyNWE1ZDg1NTNhMTg0ZDU3N2I0Y2E0YzlmOWZhOWRmNmE=
+    YWY4NWU3NDFiYTVmMmE1Y2MxMDI3ZjE3NTIyY2Q1N2Q2ZDQxM2ZlZjI4NjUy
+    MWM5OTZhNzEzZWNmMGVlYTQ1MDc1MzViMDBkOTQ0YzQyY2IxYjlmOGQwNzRh
+    YmIyOTg2Yjk0OTFlNWVhOGU3MTMzM2I1ZGY0ZjlkMzExZGNkZDk=

data/README.md

@@ -66,6 +66,10 @@ Read fasta file into a hash.
 
 ## Versions ##
 
+### 1.9.2 ###
+
+Speed up fastA `each_record` and `each_record_fast`.
+
 ### 1.9.1 ###
 
 Speed up fastQ `each_record` and `each_record_fast`. Courtesy of
@@ -221,60 +225,19 @@ Last version with File monkey patch.
 
 ## Benchmark ##
 
-**NOTE**: These benchmarks are against an older version of
-  `parse_fasta`.
-
 Some quick and dirty benchmarks against `BioRuby`.
 
 ### FastaFile#each_record ###
 
-Calculating sequence length length for each fasta record with both the
-`each_record` method from this gem and using the `FastaFormat` class
-from BioRuby. You can see the test script in `benchmark.rb`.
-
-The test file contained 2,009,897 illumina reads and the file size
-was 1.1 gigabytes. Here are the results from Ruby's `Benchmark` class:
+You can see the test script in `benchmark.rb`.
 
-                      user     system      total        real
-    parse_fasta  64.530000   1.740000  66.270000 ( 67.081502)
-    bioruby     116.250000   2.260000 118.510000 (120.223710)
+                           user     system      total        real
+    parse_fasta        1.920000   0.160000   2.080000 (  2.145932)
+    parse_fasta fast   1.210000   0.160000   1.370000 (  1.377770)
+    bioruby            4.330000   0.290000   4.620000 (  4.655567)
 
 Hot dog! It's faster :)
 
-### FastqFile#each_record ###
-
-The same sequence length test as above, but this time with a fastq
-file containing 4,000,000 illumina reads.
-
-                        user     system      total        real
-    this_fastq     62.610000   1.660000  64.270000 ( 64.389408)
-    bioruby_fastq 165.500000   2.100000 167.600000 (167.969636)
-
-### Sequence#gc ###
-
-The test is done on random strings matcing `/[AaCcTtGgUu]/`. `this_gc`
-is `Sequence.new(str).gc`, and `bioruby_gc` is
-`Bio::Sequence::NA.new(str).gc_content`.
-
-To see how the methods scales, the test 1 string was 2,000,000 bases,
-test 2 was 4,000,000 and test 3 was 8,000,000 bases.
-
-                       user     system      total        real
-    this_gc 1      0.030000   0.000000   0.030000 (  0.029145)
-    bioruby_gc 1   2.030000   0.010000   2.040000 (  2.157512)
-
-	this_gc 2      0.060000   0.000000   0.060000 (  0.059408)
-    bioruby_gc 2   4.060000   0.020000   4.080000 (  4.334159)
-
-	this_gc 3      0.120000   0.000000   0.120000 (  0.185434)
-    bioruby_gc 3   8.060000   0.020000   8.080000 (  8.659071)
-
-Nice!
-
-Troll: "When will you find the GC of an 8,000,000 base sequence?"
-
-Me: "Step off, troll!"
-
 ## Notes ##
 
 Only the `SeqFile` class actually checks to make sure that you passed

data/lib/parse_fasta/fasta_file.rb

@@ -116,10 +116,28 @@ class FastaFile < File
       #   end
       # end
     else
-      f.each("\n>") do |line|
-        header, sequence = parse_line(line)
-        yield(header.strip, Sequence.new(sequence || ""))
+      header = ""
+      sequence = ""
+      f.each_line do |line|
+        line.chomp!
+        len = line.length
+        if header.empty? && line.start_with?(">")
+          header = line[1, len]
+        elsif line.start_with?(">")
+          yield(header.strip, Sequence.new(sequence || ""))
+          header = line[1, len]
+          sequence = ""
+        else
+          raise ParseFasta::SequenceFormatError if sequence.include? ">"
+          sequence << line
+        end
       end
+      yield(header, Sequence.new(sequence || ""))
+
+      # f.each("\n>") do |line|
+      #     header, sequence = parse_line(line)
+      #     yield(header.strip, Sequence.new(sequence || ""))
+      #   end
 
       # f.each_with_index(sep=/^>/) do |line, idx|
       #   if idx.zero?
@@ -161,13 +179,31 @@ class FastaFile < File
       f = self
     end
 
-    f.each("\n>") do |line|
-      header, sequence = parse_line(line)
+    header = ""
+    sequence = ""
+    f.each_line do |line|
+      line.chomp!
+      len = line.length
+      if header.empty? && line.start_with?(">")
+        header = line[1, len]
+      elsif line.start_with?(">")
+        yield(header.strip, sequence)
+        header = line[1, len]
+        sequence = ""
+      else
+        raise ParseFasta::SequenceFormatError if sequence.include? ">"
+        sequence << line
+      end
+    end
+    yield(header, sequence)
 
-      raise ParseFasta::SequenceFormatError if sequence.include? ">"
+    # f.each("\n>") do |line|
+    #   header, sequence = parse_line(line)
 
-      yield(header.strip, sequence)
-    end
+    #   raise ParseFasta::SequenceFormatError if sequence.include? ">"
+
+    #   yield(header.strip, sequence)
+    # end
 
     f.close if f.instance_of?(Zlib::GzipReader)
     return f

data/lib/parse_fasta/version.rb

@@ -17,5 +17,5 @@
 # along with parse_fasta.  If not, see <http://www.gnu.org/licenses/>.
 
 module ParseFasta
-  VERSION = "1.9.1"
+  VERSION = "1.9.2"
 end

data/test_files/benchmark.rb

@@ -28,16 +28,23 @@ def this_parse_fasta fname
   end
 end
 
+def this_parse_fasta_fast fname
+  FastaFile.open(fname, 'r').each_record_fast do |header, sequence|
+    [header, sequence.length].join("\t")
+  end
+end
+
 def bioruby_parse_fasta fname
   Bio::FastaFormat.open(fname).each do |entry|
     [entry.definition, entry.seq.length].join("\t")
   end
 end
 
-# Benchmark.bmbm do |x|
-#   x.report('parse_fasta') { this_parse_fasta(ARGV.first) }
-#   x.report('bioruby')     { bioruby_parse_fasta(ARGV.first) }
-# end
+Benchmark.bmbm do |x|
+  x.report('parse_fasta') { this_parse_fasta(ARGV.first) }
+  x.report('parse_fasta fast') { this_parse_fasta_fast(ARGV.first) }
+  x.report('bioruby')     { bioruby_parse_fasta(ARGV.first) }
+end
 
 ####
 
@@ -72,8 +79,8 @@ end
 # fastq = ARGV.first
 
 def bioruby_fastq(fastq)
-  Bio::FlatFile.open(Bio::Fastq, fastq) do |fq| 
-    fq.each do |entry| 
+  Bio::FlatFile.open(Bio::Fastq, fastq) do |fq|
+    fq.each do |entry|
       [entry.definition, entry.seq.length].join("\t")
     end
   end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: parse_fasta
 version: !ruby/object:Gem::Version
-  version: 1.9.1
+  version: 1.9.2
 platform: ruby
 authors:
 - Ryan Moore
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-04-20 00:00:00.000000000 Z
+date: 2016-05-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler