bamfcsv 0.0.1

data/.gitignore

@@ -0,0 +1,11 @@
+*.o
+*.bundle
+*.dSYM
+a.out
+*.so
+*~
+\#*
+*.csv*
+!spec/fixtures/*.csv*
+Makefile
+pkg

data/.rspec

@@ -0,0 +1,2 @@
+--format Fuubar
+--colour

data/Gemfile

@@ -0,0 +1,4 @@
+source "http://rubygems.org"
+
+# Specify your gem's dependencies in bamfcsv.gemspec
+gemspec

data/Gemfile.lock

@@ -0,0 +1,35 @@
+PATH
+  remote: .
+  specs:
+    bamfcsv (0.0.1)
+
+GEM
+  remote: http://rubygems.org/
+  specs:
+    chalofa_ruby-progressbar (0.0.9.1)
+    diff-lcs (1.1.2)
+    fuubar (0.0.4)
+      chalofa_ruby-progressbar (~> 0.0.9)
+      rspec (~> 2.0)
+      rspec-instafail (~> 0.1.4)
+    rake (0.8.7)
+    rake-compiler (0.7.6)
+      rake
+    rspec (2.5.0)
+      rspec-core (~> 2.5.0)
+      rspec-expectations (~> 2.5.0)
+      rspec-mocks (~> 2.5.0)
+    rspec-core (2.5.1)
+    rspec-expectations (2.5.0)
+      diff-lcs (~> 1.1.2)
+    rspec-instafail (0.1.7)
+    rspec-mocks (2.5.0)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  bamfcsv!
+  fuubar (~> 0.0.2)
+  rake-compiler (~> 0.7.1)
+  rspec (~> 2.5.0)

data/README

@@ -0,0 +1,3 @@
+BAMF,CSV,parser
+OHHH,YEAH,there,are,so
+MANy,cahrs,in,here.

data/Rakefile

@@ -0,0 +1,4 @@
+require 'bundler'
+Bundler::GemHelper.install_tasks
+
+Dir['tasks/*.rake'].sort.each { |f| load f }

data/bamfcsv.gemspec

@@ -0,0 +1,26 @@
+# -*- encoding: utf-8 -*-
+$:.push File.expand_path("../lib", __FILE__)
+require "bamfcsv/version"
+
+Gem::Specification.new do |s|
+  s.name        = "bamfcsv"
+  s.version     = BAMFCSV::VERSION
+  s.platform    = Gem::Platform::RUBY
+  s.authors     = ["Jon Distad", "Alex Redington"]
+  s.email       = ["jon@thinkrelevance.com", "lovemachine@thinkrelevance.com"]
+  s.homepage    = "https://github.com/jondistad/bamfcsv"
+  s.summary     = %q{BAMF!!! Your csv is parsed.}
+  s.description = %q{BAMFCSV parses csv like a BAMF. BAMF!!}
+
+  s.rubyforge_project = "bamfcsv"
+
+  s.files         = `git ls-files`.split("\n")
+  s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
+  s.executables   = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
+  s.require_paths = ["lib", "ext"]
+  s.extensions = %w{ext/bamfcsv/extconf.rb}
+
+  s.add_development_dependency "rspec", "~> 2.5.0"
+  s.add_development_dependency "fuubar", "~> 0.0.2"
+  s.add_development_dependency 'rake-compiler', "~> 0.7.1"
+end

data/ext/bamfcsv/bamfcsv_ext.c

@@ -0,0 +1,191 @@
+#include <stdlib.h>
+#include <ruby/ruby.h>
+#include <fcntl.h>
+#include <sys/mman.h>
+#include "bamfcsv_ext.h"
+
+struct s_Row *alloc_row() {
+
+  struct s_Row *new_row = malloc(sizeof(struct s_Row));
+
+  new_row -> first_cell = 0;
+  new_row -> next_row = 0;
+  new_row -> cell_count = 0;
+
+  return new_row;
+
+}
+
+struct s_Cell *alloc_cell() {
+
+  struct s_Cell *new_cell = malloc(sizeof(struct s_Cell));
+
+  new_cell -> start = 0;
+  new_cell -> len = 0;
+  new_cell -> next_cell = 0;
+  new_cell -> has_quotes = 0;
+
+  return new_cell;
+
+}
+
+void free_cell(struct s_Cell *cell) {
+
+  if (cell != 0) {
+    free_cell(cell->next_cell);
+    free(cell);
+  }
+
+}
+
+void free_row(struct s_Row *row) {
+
+  if (row != 0) {
+
+    free_row(row->next_row);
+    free_cell(row->first_cell);
+    free(row);
+
+  }
+
+}
+
+VALUE build_matrix_from_pointer_tree(struct s_Row *first_row, int num_rows) {
+  VALUE matrix;
+  VALUE row;
+  VALUE new_string;
+  int i,j;
+  struct s_Row *cur_row;
+  struct s_Cell *cur_cell;
+
+  cur_row = first_row;
+  matrix = rb_ary_new2(num_rows);
+
+  ID gsub = rb_intern("gsub!");
+  VALUE dquote = rb_str_new2("\"\""), quote = rb_str_new2("\"");
+
+  for (i = 0; i < num_rows; i++) {
+
+    cur_cell = cur_row->first_cell;
+    row = rb_ary_new2(cur_row->cell_count);
+    rb_ary_store(matrix,i,row);
+    for (j = 0; j < cur_row->cell_count; j++) {
+      if (*(cur_cell->start) == '"' 
+          && *((cur_cell->start)+((cur_cell->len-1)*sizeof(char))) == '"')
+        new_string = rb_str_new(cur_cell->start+sizeof(char), cur_cell->len-(sizeof(char)*2));
+      else
+        new_string = rb_str_new(cur_cell->start, cur_cell->len);
+      if (cur_cell->has_quotes) {
+        rb_funcall(new_string, gsub, 2, dquote, quote);
+      }
+      rb_ary_store(row, j, new_string);
+      cur_cell = cur_cell->next_cell;
+    }
+    cur_row = cur_row->next_row;
+  }
+
+  return matrix;
+}
+
+void finalize_cell(struct s_Cell *cell, char *cur) {
+  if (*(cur-sizeof(char)) == '\r') 
+    cell->len = cur-(cell->start)-sizeof(char);
+  else
+    cell->len = cur-(cell->start);
+}
+
+VALUE build_matrix(char *buf, int bufsize) {
+  int str_start = 0;
+  int num_rows = 1;
+  int in_quote = 0;
+
+  struct s_Row *first_row = alloc_row();
+  struct s_Row *cur_row = first_row;
+  struct s_Cell *cur_cell = alloc_cell();
+  first_row->first_cell = cur_cell;
+  cur_cell->start = buf;
+
+  VALUE matrix;
+
+  char *cur;
+  
+  for (cur = buf; cur < buf+bufsize; cur++) {
+
+    if (*cur == '"') {
+      if (in_quote)
+        if (*(cur+1) != ',')
+          cur_cell->has_quotes = 1;
+      in_quote = !in_quote;
+    }
+
+    if (!in_quote) {
+
+      if (*cur == ',') {
+        
+        finalize_cell(cur_cell,cur);
+        cur_cell->next_cell = alloc_cell();
+        cur_cell = cur_cell->next_cell;
+        cur_cell->start = cur+sizeof(char);
+        cur_row->cell_count += 1;
+
+      }
+      
+      if (*cur == '\n') {
+        
+        finalize_cell(cur_cell,cur);
+        cur_row->cell_count += 1;
+        cur_row->next_row = alloc_row();
+        cur_row = cur_row -> next_row;
+        cur_row->first_cell = alloc_cell();
+        cur_cell = cur_row->first_cell;
+        cur_cell->start = cur+sizeof(char);
+        
+        num_rows++;
+
+      }
+      
+    }
+
+  }
+
+  if (cur_row->cell_count == 0) {
+    num_rows--;
+  }
+
+  matrix = build_matrix_from_pointer_tree(first_row, num_rows);
+
+  free_row(first_row);
+
+  return matrix;
+
+}
+
+VALUE mm_parse(const char *file) {
+
+  char *mmapped_csv;
+  int filesize, csv;
+
+  csv = open(file, O_RDONLY);
+  filesize = lseek(csv, 0, SEEK_END);
+  mmapped_csv = (char*) mmap(0, filesize, PROT_READ, MAP_SHARED, csv, 0);
+
+  VALUE matrix = build_matrix(mmapped_csv,filesize);
+
+  munmap(mmapped_csv, filesize);
+  close(csv);
+
+  return matrix;
+}
+
+VALUE read_path(VALUE self, VALUE file) {
+
+  return mm_parse(RSTRING_PTR(file));
+
+}
+
+void Init_bamfcsv() {
+
+  VALUE module = rb_define_module("BAMFCSV");
+  rb_define_module_function(module, "read_path", read_path, 1);
+
+}

data/ext/bamfcsv/bamfcsv_ext.h

@@ -0,0 +1,19 @@
+#ifndef _BAMFCSV_EXT_H
+#define _BAMFCSV_EXT_H
+
+struct s_Row {
+  struct s_Cell *first_cell;
+  struct s_Row *next_row;
+  int cell_count;
+};
+
+struct s_Cell {
+  char *start;
+  int len;
+  int has_quotes;
+  struct s_Cell *next_cell;
+};
+
+void Init_bamfcsv();
+
+#endif

data/ext/bamfcsv/extconf.rb

@@ -0,0 +1,2 @@
+require 'mkmf'
+create_makefile('bamfcsv')

data/lib/bamfcsv.rb

@@ -0,0 +1,13 @@
+require 'bamfcsv/bamfcsv'
+
+module BAMFCSV
+
+  def self.read(thing_to_read)
+    if  String === thing_to_read
+      raise Errno::ENOENT.new("#{thing_to_read} does not exist") unless File.exist? thing_to_read
+      raise Errno::EISDIR.new("#{thing_to_read} is a directory") if File.directory? thing_to_read
+      read_path(thing_to_read)
+    end
+  end
+
+end

data/lib/bamfcsv/version.rb

@@ -0,0 +1,3 @@
+module BAMFCSV
+  VERSION = "0.0.1"
+end

data/spec/fixtures/bamf-comma-comma.csv

@@ -0,0 +1 @@
+BAMF,,CSV

data/spec/fixtures/double-quotes.csv

@@ -0,0 +1,3 @@
+this is a semicolon:, ;
+this is a comma:," ,"
+this is a quote:," """

data/spec/fixtures/escapes.csv

@@ -0,0 +1,2 @@
+this is a semicolon:, ;
+this is a comma:," ,"

data/spec/fixtures/one-column.csv

@@ -0,0 +1,2 @@
+BAMF
+CSV

data/spec/fixtures/terminated-with-cr.csv

@@ -0,0 +1,2 @@
+a
+b

data/spec/fixtures/test.csv

@@ -0,0 +1,3 @@
+BAMF,CSV,parser
+OHHH,YEAH,there,are,so
+MANy,cahrs,in,here.

data/spec/lib/bamfcsv_spec.rb

@@ -0,0 +1,49 @@
+require 'spec_helper'
+
+describe BAMFCSV do
+  it "has a parse method" do
+    BAMFCSV.should respond_to(:read)
+  end
+
+  describe "#read" do
+    it "is a matrix given a filename" do
+      BAMFCSV.read("spec/fixtures/test.csv").should be_instance_of Array
+    end
+
+    it "is an empty array passed an empty file" do
+      BAMFCSV.read("spec/fixtures/empty.csv").should == []
+    end
+
+    it "is a 1xn matrix with a one column csv file" do
+      BAMFCSV.read("spec/fixtures/one-column.csv").should == [["BAMF"],["CSV"]]
+    end
+
+    it "interprets empty cells correctly" do
+      BAMFCSV.read("spec/fixtures/bamf-comma-comma.csv").should == [["BAMF","","CSV"]]
+    end
+
+    it "escapes cells that are quoted" do
+      BAMFCSV.read("spec/fixtures/escapes.csv").should == [["this is a semicolon:", " ;"],["this is a comma:", " ,"]]
+    end
+
+    it "escapes cells that are quoted" do
+      BAMFCSV.read("spec/fixtures/double-quotes.csv").should == [["this is a semicolon:", " ;"], ["this is a comma:", " ,"], ["this is a quote:", " \""]]
+    end
+
+    it "doesn't create a row when the file terminates with [CR]LF" do
+      BAMFCSV.read("spec/fixtures/terminated-with-cr.csv").should == [["a"],["b"]]
+    end
+
+    it "raises Errno::ENOENT when the file does not exist" do
+      expect do
+        BAMFCSV.read("spec/fixtures/this-file-does-not-not-exist.csv")
+      end.should raise_error Errno::ENOENT
+    end
+
+    it "raises Errno::EISDIR when the file is not a flat file" do
+      expect do
+        BAMFCSV.read("spec/fixtures/")
+      end.should raise_error Errno::EISDIR
+    end
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1,12 @@
+root_path = File.expand_path(File.join(File.dirname(__FILE__), '..'))
+$:.unshift(File.join(root_path, "lib"))
+$:.unshift(File.join(root_path, "ext"))
+require 'bundler'
+require 'bamfcsv'
+Bundler.require(:development)
+
+RSpec.configure do |config|
+  config.filter_run :focused => true
+  config.alias_example_to :fit, :focused => true
+  config.run_all_when_everything_filtered = true
+end

data/tasks/compile.rake

@@ -0,0 +1,11 @@
+require "rake/extensiontask"
+
+def gemspec
+  @clean_gemspec ||= eval(File.read(File.expand_path('../../bamfcsv.gemspec', __FILE__)))
+end
+
+Rake::ExtensionTask.new("bamfcsv", gemspec) do |ext|
+  ext.lib_dir = File.join 'lib', 'bamfcsv'
+  CLEAN.include "#{ext.lib_dir}/*.#{RbConfig::CONFIG['DLEXT']}"
+end
+Rake::Task[:spec].prerequisites << :compile

data/tasks/rspec.rake

@@ -0,0 +1,12 @@
+begin
+  require 'rspec'
+  require 'rspec/core/rake_task'
+
+  RSpec::Core::RakeTask.new('spec') do |t|
+    t.verbose = true
+  end
+
+  task :default => :spec
+rescue LoadError
+  puts "rspec, or one of its dependencies, is not available. Install it with: sudo gem install rspec"
+end

metadata

@@ -0,0 +1,122 @@
+--- !ruby/object:Gem::Specification 
+name: bamfcsv
+version: !ruby/object:Gem::Version 
+  prerelease: 
+  version: 0.0.1
+platform: ruby
+authors: 
+- Jon Distad
+- Alex Redington
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2011-04-01 00:00:00 -04:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        version: 2.5.0
+  type: :development
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  name: fuubar
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        version: 0.0.2
+  type: :development
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: rake-compiler
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        version: 0.7.1
+  type: :development
+  version_requirements: *id003
+description: BAMFCSV parses csv like a BAMF. BAMF!!
+email: 
+- jon@thinkrelevance.com
+- lovemachine@thinkrelevance.com
+executables: []
+
+extensions: 
+- ext/bamfcsv/extconf.rb
+extra_rdoc_files: []
+
+files: 
+- .gitignore
+- .rspec
+- Gemfile
+- Gemfile.lock
+- README
+- Rakefile
+- bamfcsv.gemspec
+- ext/bamfcsv/bamfcsv_ext.c
+- ext/bamfcsv/bamfcsv_ext.h
+- ext/bamfcsv/extconf.rb
+- lib/bamfcsv.rb
+- lib/bamfcsv/version.rb
+- spec/fixtures/bamf-comma-comma.csv
+- spec/fixtures/double-quotes.csv
+- spec/fixtures/empty.csv
+- spec/fixtures/escapes.csv
+- spec/fixtures/one-column.csv
+- spec/fixtures/terminated-with-cr.csv
+- spec/fixtures/test.csv
+- spec/lib/bamfcsv_spec.rb
+- spec/spec_helper.rb
+- tasks/compile.rake
+- tasks/rspec.rake
+has_rdoc: true
+homepage: https://github.com/jondistad/bamfcsv
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+- ext
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+requirements: []
+
+rubyforge_project: bamfcsv
+rubygems_version: 1.6.2
+signing_key: 
+specification_version: 3
+summary: BAMF!!! Your csv is parsed.
+test_files: 
+- spec/fixtures/bamf-comma-comma.csv
+- spec/fixtures/double-quotes.csv
+- spec/fixtures/empty.csv
+- spec/fixtures/escapes.csv
+- spec/fixtures/one-column.csv
+- spec/fixtures/terminated-with-cr.csv
+- spec/fixtures/test.csv
+- spec/lib/bamfcsv_spec.rb
+- spec/spec_helper.rb