bamfcsv 0.2.0 → 0.3.1

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    bamfcsv (0.1.4)
+    bamfcsv (0.3.1)
 
 GEM
   remote: http://rubygems.org/

data/ext/bamfcsv/bamfcsv_ext.c

@@ -1,11 +1,13 @@
 #include <ruby/ruby.h>
+#include <ruby/encoding.h>
 #include <stdlib.h>
 #include <stdio.h>
+#include <stdbool.h>
 
 VALUE BAMFCSV_module;
 VALUE BAMFCSV_MalformedCSVError_class;
 
-VALUE bamfcsv_finalize_cell(char *cell_start, char *cell_end, int quote_count) {
+VALUE bamfcsv_finalize_cell(char *cell_start, char *cell_end, int quote_count, rb_encoding *enc) {
   if (*cell_end == '\r')
     cell_end--;
 
@@ -17,12 +19,20 @@ VALUE bamfcsv_finalize_cell(char *cell_start, char *cell_end, int quote_count) {
     cell_end--;
   }
 
-  VALUE cell_str = rb_str_new(cell_start, cell_end-cell_start+1);
+  VALUE cell_str = rb_enc_str_new(cell_start, cell_end-cell_start+1, enc);
 
   return cell_str;
 }
 
-VALUE bamfcsv_build_matrix(char *buf, unsigned long bufsize) {
+bool quotes_end_line(char* cur) {
+  return *(cur-1) == '"' || (*(cur-1) == '\r' && *(cur-2) == '"');
+}
+
+VALUE bamfcsv_parse_string(VALUE self, VALUE string) {
+  char *buf = RSTRING_PTR(string);
+  long bufsize = RSTRING_LEN(string);
+  rb_encoding *enc = rb_enc_from_index(ENCODING_GET(string));
+
   unsigned long num_rows = 1, cell_count = 1;
   int quote_count = 0, quotes_matched = 1;
 
@@ -57,7 +67,7 @@ VALUE bamfcsv_build_matrix(char *buf, unsigned long bufsize) {
         if (quote_count && *(cur-1) != '"')
           rb_raise(BAMFCSV_MalformedCSVError_class, "Unclosed quoted field on line %lu, cell %lu.", num_rows, cell_count);
 
-        VALUE cell_str = bamfcsv_finalize_cell(cell_start, cur-1, quote_count);
+        VALUE cell_str = bamfcsv_finalize_cell(cell_start, cur-1, quote_count, enc);
         if (quote_count)
           rb_funcall(cell_str, gsub_bang, 2, dbl_dquote, dquote);
 
@@ -69,10 +79,10 @@ VALUE bamfcsv_build_matrix(char *buf, unsigned long bufsize) {
 
       } else if (*cur == '\n') {
         
-        if (quote_count && !(*(cur-1) == '"' || *(cur-1) == '\r' && *(cur-2) == '"'))
+        if (quote_count && !quotes_end_line(cur))
             rb_raise(BAMFCSV_MalformedCSVError_class, "Unclosed quoted field on line %lu, cell %lu: EOL", num_rows, cell_count);
 
-        VALUE cell_str = bamfcsv_finalize_cell(cell_start, cur-1, quote_count);
+        VALUE cell_str = bamfcsv_finalize_cell(cell_start, cur-1, quote_count, enc);
         if (quote_count)
           rb_funcall(cell_str, gsub_bang, 2, dbl_dquote, dquote);
         /* Completely blank lines don't even get a nil. This matches CSV's behavior. */
@@ -95,10 +105,10 @@ VALUE bamfcsv_build_matrix(char *buf, unsigned long bufsize) {
 
   if (!quotes_matched) /* Reached EOF without matching quotes */
     rb_raise(BAMFCSV_MalformedCSVError_class, "Illegal quoting on line %lu, cell %lu: File ends without closing '\"'", num_rows, cell_count);
-  else if (quote_count && *(cur-1) != '"') /* Quotes closed before end of final cell */
+  else if (quote_count && !quotes_end_line(cur)) /* Quotes closed before end of final cell */
     rb_raise(BAMFCSV_MalformedCSVError_class, "Unclosed quoted field on line %lu, cell %lu: EOF", num_rows, cell_count);
 
-  VALUE cell_str = bamfcsv_finalize_cell(cell_start, cur-1, quote_count);
+  VALUE cell_str = bamfcsv_finalize_cell(cell_start, cur-1, quote_count, enc);
   if (quote_count)
     rb_funcall(cell_str, gsub_bang, 2, dbl_dquote, dquote);
   /* Completely blank lines don't even get a nil. This matches CSV's behavior. */
@@ -110,12 +120,6 @@ VALUE bamfcsv_build_matrix(char *buf, unsigned long bufsize) {
 
 }
 
-VALUE bamfcsv_parse_string(VALUE self, VALUE string) {
-
-  return bamfcsv_build_matrix(RSTRING_PTR(string), NUM2ULONG(rb_str_length(string)));
-
-}
-
 void Init_bamfcsv() {
 
   BAMFCSV_module = rb_define_module("BAMFCSV");

data/lib/bamfcsv/table.rb

@@ -26,8 +26,12 @@ module BAMFCSV
       @row_cache[idx] ||= Row.new(@header_map, @matrix[idx])
     end
 
+    def empty?
+      @matrix.empty?
+    end
+
     def inspect
-      "#<BAMFCSV::Table>"
+      "[#{self.map{|r| r.inspect}.join(", ")}]"
     end
 
     private
@@ -50,6 +54,15 @@ module BAMFCSV
       def [](key)
         @fields[@header_map[key]]
       end
+
+      def inspect
+        pairs = []
+        headers.each do |h|
+          pairs << "#{h.inspect} => #{self[h].inspect}"
+        end
+        "{#{pairs.join(", ")}}"
+      end
+
     end
   end
 end

data/lib/bamfcsv/version.rb

@@ -1,3 +1,3 @@
 module BAMFCSV
-  VERSION = "0.2.0"
+  VERSION = "0.3.1"
 end

data/lib/bamfcsv.rb

@@ -9,7 +9,12 @@ module BAMFCSV
 
   def self.parse(csv_str, opts={})
     return [] if csv_str.empty?
-    matrix = __parse_string(csv_str)
+    # We need to do this because the C extension currently overwrites
+    # the input, and all of String#clone, String#dup, and String.new
+    # copy the pointer, not the contents. So we make a copy, parse
+    # that, and throw away the copy.
+    copy = "" + csv_str
+    matrix = __parse_string(copy)
     if opts[:headers]
       Table.new(matrix)
     else

data/spec/lib/bamfcsv_spec.rb

@@ -56,7 +56,15 @@ describe BAMFCSV do
       BAMFCSV.parse("1,2").should == [["1","2"]]
     end
 
-    it 'correctly escaptes ""' do
+    it "correctly parses the last cell when followed by CRLF" do
+      BAMFCSV.parse("1,2\r\n").should == [["1","2"]]
+    end
+
+    it "correctly parses the last cell when quotes and followed by CRLF" do
+      BAMFCSV.parse("1,\"2\"\r\n").should == [["1","2"]]
+    end
+
+    it 'correctly escapes ""' do
       BAMFCSV.parse(%Q|1,"""2"""\n|).should == [["1", '"2"']]
     end
 
@@ -74,6 +82,19 @@ describe BAMFCSV do
       BAMFCSV.parse("1\r\n2").should == [["1"],["2"]]
     end
 
+    it "parses data outside the 7-bit range" do
+      BAMFCSV.parse("age \u226540 years").should == [["age \u226540 years"]]
+    end
+
+    it "doesn't alter the input" do
+      original = %Q{this,that,"the ""other"" thing"\r\n1,2,3\n}
+      # String#dup, String#clone, and String.new copy the pointer but
+      # share the same underlying buffer, d'oh!
+      input = "" + original
+      BAMFCSV.parse(input)
+      input.should == original
+    end
+
     describe "default CSV module compatibility" do
       it "adds a nil cell after a trailing comma with no newline" do
         BAMFCSV.parse("1,2,").should == [["1","2",nil]]
@@ -179,5 +200,55 @@ describe BAMFCSV do
         expect { table.each }.should_not raise_error(LocalJumpError)
       end
     end
+
+    describe "Table#inspect" do
+      it "is an Array of Hashes" do
+        csv = <<CSV
+foo,bar
+1,2
+baz,quux
+CSV
+
+        inspected = '[{"foo" => "1", "bar" => "2"}, {"foo" => "baz", "bar" => "quux"}]'
+
+        BAMFCSV.parse(csv, :headers => true).inspect.should == inspected
+      end
+    end
+
+    describe "Table#empty?" do
+      it "returns true for an empty table with headers" do
+        table = BAMFCSV.parse("column1,column2\n", :headers => true)
+        table.empty?.should be_true
+      end
+
+      it "returns false for a non-empty table with headers" do
+        table = BAMFCSV.parse("column1,column2\nfoo,bar", :headers => true)
+        table.empty?.should be_false
+      end
+
+      it "returns true for an empty table without headers" do
+        table = BAMFCSV.parse("", :headers => false)
+        table.empty?.should be_true
+      end
+
+      it "returns false for a non-empty table without headers" do
+        table = BAMFCSV.parse("foo,bar", :headers => false)
+        table.empty?.should be_false
+      end
+    end
+
+    describe "Table::Row#inspect" do
+      it "is a Hash" do
+        csv = <<CSV
+foo,bar
+1,2
+baz,quux
+CSV
+
+        inspected = '{"foo" => "1", "bar" => "2"}'
+
+        BAMFCSV.parse(csv, :headers => true).first.inspect.should == inspected
+      end
+    end
   end
 end

metadata

@@ -1,63 +1,59 @@
---- !ruby/object:Gem::Specification 
+--- !ruby/object:Gem::Specification
 name: bamfcsv
-version: !ruby/object:Gem::Version 
+version: !ruby/object:Gem::Version
+  version: 0.3.1
   prerelease: 
-  version: 0.2.0
 platform: ruby
-authors: 
+authors:
 - Jon Distad
 - Alex Redington
 autorequire: 
 bindir: bin
 cert_chain: []
-
-date: 2011-04-29 00:00:00 -04:00
-default_executable: 
-dependencies: 
-- !ruby/object:Gem::Dependency 
+date: 2011-08-10 00:00:00.000000000Z
+dependencies:
+- !ruby/object:Gem::Dependency
   name: rspec
-  prerelease: false
-  requirement: &id001 !ruby/object:Gem::Requirement 
+  requirement: &2157181100 !ruby/object:Gem::Requirement
     none: false
-    requirements: 
+    requirements:
     - - ~>
-      - !ruby/object:Gem::Version 
+      - !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 
+  version_requirements: *2157181100
+- !ruby/object:Gem::Dependency
+  name: fuubar
+  requirement: &2157180600 !ruby/object:Gem::Requirement
     none: false
-    requirements: 
+    requirements:
     - - ~>
-      - !ruby/object:Gem::Version 
+      - !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 
+  version_requirements: *2157180600
+- !ruby/object:Gem::Dependency
+  name: rake-compiler
+  requirement: &2157180140 !ruby/object:Gem::Requirement
     none: false
-    requirements: 
+    requirements:
     - - ~>
-      - !ruby/object:Gem::Version 
+      - !ruby/object:Gem::Version
         version: 0.7.1
   type: :development
-  version_requirements: *id003
+  prerelease: false
+  version_requirements: *2157180140
 description: BAMFCSV parses csv like a BAMF. BAMF!!
-email: 
+email:
 - jon@thinkrelevance.com
 - lovemachine@thinkrelevance.com
 executables: []
-
-extensions: 
+extensions:
 - ext/bamfcsv/extconf.rb
 extra_rdoc_files: []
-
-files: 
+files:
 - .gitignore
 - .rspec
 - Gemfile
@@ -81,36 +77,32 @@ files:
 - 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: 
+require_paths:
 - lib
 - ext
-required_ruby_version: !ruby/object:Gem::Requirement 
+required_ruby_version: !ruby/object:Gem::Requirement
   none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      version: "0"
-required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      version: "0"
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
 requirements: []
-
 rubyforge_project: bamfcsv
-rubygems_version: 1.5.3
+rubygems_version: 1.8.6
 signing_key: 
 specification_version: 3
 summary: BAMF!!! Your csv is parsed.
-test_files: 
+test_files:
 - spec/fixtures/bamf-comma-comma.csv
 - spec/fixtures/double-quotes.csv
 - spec/fixtures/empty.csv