fastcsv 0.0.3 → 0.0.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 1f1c56bdc3a7600dfb311eee7a98f07fe0d0e575
-  data.tar.gz: 2fae9ec519e877178a81dfb6f139c0ce94c974d2
+  metadata.gz: e5c917e36ec5bc0f784fa36281e33e23a14784c3
+  data.tar.gz: 38ac5f6dc7c494759b98c06b4879c54770e088d7
 SHA512:
-  metadata.gz: d9063634cca29ad95e961ee7bfa7269dc906c39a70a8317e35ea6ffc5991bd0caadf5b985040638d392bcbd49f1931d991261377c9a94c04d31b0147a8e9d721
-  data.tar.gz: 51314a4e948a996ad1546097ac7ad5c9d9d72eb6e8ef38a8874039920814445bee5de121225b81d5e2d461037b7fe23b1595c6be1f54dc93576aa7497681d91f
+  metadata.gz: bfcc78d6223c83a270b658921a334e70f58dc600d380935d3baf27e980c0881cfcc5018af08adbb5f039f839aa52c153f161013fa1ae544f74c5968fac1daf86
+  data.tar.gz: 0e7c462ec4f1d41174885faa9af6fd93db533cb143e1403fa7dc0b3fa18bc8e6a69a4e3db3697ee798f430d3af5191f31d9c2eee04f6e9e2e40edcccb0ca45dc

data/ext/fastcsv/fastcsv.c

@@ -24,9 +24,6 @@ if (enc2 != NULL) { \
 #define FREE \
 if (buf != NULL) { \
   free(buf); \
-} \
-if (row_sep != NULL) { \
-  free(row_sep); \
 }
 
 static VALUE cClass, cParser, eError;
@@ -38,11 +35,11 @@ typedef struct {
 } Data;
 
 
-#line 170 "ext/fastcsv/fastcsv.rl"
+#line 152 "ext/fastcsv/fastcsv.rl"
 
 
 
-#line 46 "ext/fastcsv/fastcsv.c"
+#line 43 "ext/fastcsv/fastcsv.c"
 static const int raw_parse_start = 4;
 static const int raw_parse_first_final = 4;
 static const int raw_parse_error = 0;
@@ -50,7 +47,7 @@ static const int raw_parse_error = 0;
 static const int raw_parse_en_main = 4;
 
 
-#line 173 "ext/fastcsv/fastcsv.rl"
+#line 155 "ext/fastcsv/fastcsv.rl"
 
 // 16 kB
 #define BUFSIZE 16384
@@ -83,11 +80,11 @@ static void rb_io_ext_int_to_encs(rb_encoding *ext, rb_encoding *intern, rb_enco
 
 static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
   int cs, act, have = 0, curline = 1, io = 0;
-  char *ts = 0, *te = 0, *buf = 0, *eof = 0, *mark_row_sep = 0, *row_sep = NULL;
+  char *ts = 0, *te = 0, *buf = 0, *eof = 0;
 
   VALUE port, opts, r_encoding;
   VALUE row = rb_ary_new(), field = Qnil, bufsize = Qnil;
-  int done = 0, unclosed_line = 0, len_row_sep = 0, buffer_size = 0, taint = 0;
+  int done = 0, unclosed_line = 0, buffer_size = 0, taint = 0;
   rb_encoding *enc = NULL, *enc2 = NULL, *encoding = NULL;
 
   Data *d;
@@ -225,6 +222,8 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
     encoding = rb_enc_get(r_encoding);
   }
 
+  // In case #raw_parse is called multiple times on the same parser. Note that
+  // using IO methods on a re-used parser can cause segmentation faults.
   rb_ivar_set(self, s_row, Qnil);
 
   buffer_size = BUFSIZE;
@@ -244,7 +243,7 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
   }
 
   
-#line 248 "ext/fastcsv/fastcsv.c"
+#line 247 "ext/fastcsv/fastcsv.c"
 	{
 	cs = raw_parse_start;
 	ts = 0;
@@ -252,12 +251,12 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
 	act = 0;
 	}
 
-#line 366 "ext/fastcsv/fastcsv.rl"
+#line 350 "ext/fastcsv/fastcsv.rl"
 
   while (!done) {
     VALUE str;
     char *p, *pe;
-    int len, space = buffer_size - have, tokstart_diff, tokend_diff, start_diff, mark_row_sep_diff;
+    int len, space = buffer_size - have, tokstart_diff, tokend_diff, start_diff;
 
     if (io) {
       if (space == 0) {
@@ -265,7 +264,6 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
         tokstart_diff = ts - buf;
         tokend_diff = te - buf;
         start_diff = d->start - buf;
-        mark_row_sep_diff = mark_row_sep - buf;
 
         buffer_size += BUFSIZE;
         REALLOC_N(buf, char, buffer_size);
@@ -275,7 +273,6 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
         ts = buf + tokstart_diff;
         te = buf + tokend_diff;
         d->start = buf + start_diff;
-        mark_row_sep = buf + mark_row_sep_diff;
       }
       p = buf + have;
 
@@ -314,7 +311,7 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
 
     pe = p + len;
     
-#line 318 "ext/fastcsv/fastcsv.c"
+#line 315 "ext/fastcsv/fastcsv.c"
 	{
 	if ( p == pe )
 		goto _test_eof;
@@ -333,7 +330,7 @@ tr0:
 	}
 	goto st4;
 tr5:
-#line 49 "ext/fastcsv/fastcsv.rl"
+#line 46 "ext/fastcsv/fastcsv.rl"
 	{
     if (p == ts) {
       // Unquoted empty fields are nil, not "", in Ruby.
@@ -344,16 +341,16 @@ tr5:
       ENCODE;
     }
   }
-#line 95 "ext/fastcsv/fastcsv.rl"
+#line 92 "ext/fastcsv/fastcsv.rl"
 	{
     rb_ary_push(row, field);
     field = Qnil;
   }
-#line 166 "ext/fastcsv/fastcsv.rl"
+#line 148 "ext/fastcsv/fastcsv.rl"
 	{te = p+1;}
 	goto st4;
 tr9:
-#line 138 "ext/fastcsv/fastcsv.rl"
+#line 120 "ext/fastcsv/fastcsv.rl"
 	{
     if (d->start == 0 || p == d->start) {
       rb_ivar_set(self, s_row, rb_str_new2(""));
@@ -370,41 +367,28 @@ tr9:
       rb_yield(row);
     }
   }
-#line 168 "ext/fastcsv/fastcsv.rl"
+#line 150 "ext/fastcsv/fastcsv.rl"
 	{te = p+1;}
 	goto st4;
 tr12:
-#line 95 "ext/fastcsv/fastcsv.rl"
+#line 92 "ext/fastcsv/fastcsv.rl"
 	{
     rb_ary_push(row, field);
     field = Qnil;
   }
-#line 166 "ext/fastcsv/fastcsv.rl"
+#line 148 "ext/fastcsv/fastcsv.rl"
 	{te = p+1;}
 	goto st4;
 tr15:
-#line 168 "ext/fastcsv/fastcsv.rl"
+#line 150 "ext/fastcsv/fastcsv.rl"
 	{te = p;p--;}
 	goto st4;
 tr16:
-#line 100 "ext/fastcsv/fastcsv.rl"
+#line 97 "ext/fastcsv/fastcsv.rl"
 	{
     d->start = p;
-
-    if (len_row_sep) {
-      if (p - mark_row_sep != len_row_sep || row_sep[0] != *mark_row_sep || len_row_sep == 2 && row_sep[1] != *(mark_row_sep + 1)) {
-        FREE;
-
-        rb_raise(eError, "Unquoted fields do not allow \\r or \\n (line %d).", curline - 1);
-      }
-    }
-    else {
-      len_row_sep = p - mark_row_sep;
-      row_sep = ALLOC_N(char, p - mark_row_sep);
-      memcpy(row_sep, mark_row_sep, p - mark_row_sep);
-    }
   }
-#line 167 "ext/fastcsv/fastcsv.rl"
+#line 149 "ext/fastcsv/fastcsv.rl"
 	{te = p;p--;}
 	goto st4;
 st4:
@@ -417,7 +401,7 @@ st4:
 case 4:
 #line 1 "NONE"
 	{ts = p;}
-#line 421 "ext/fastcsv/fastcsv.c"
+#line 405 "ext/fastcsv/fastcsv.c"
 	switch( (*p) ) {
 		case 0: goto tr13;
 		case 10: goto tr3;
@@ -441,7 +425,7 @@ case 1:
 tr2:
 #line 1 "NONE"
 	{te = p+1;}
-#line 49 "ext/fastcsv/fastcsv.rl"
+#line 46 "ext/fastcsv/fastcsv.rl"
 	{
     if (p == ts) {
       // Unquoted empty fields are nil, not "", in Ruby.
@@ -452,7 +436,7 @@ tr2:
       ENCODE;
     }
   }
-#line 138 "ext/fastcsv/fastcsv.rl"
+#line 120 "ext/fastcsv/fastcsv.rl"
 	{
     if (d->start == 0 || p == d->start) {
       rb_ivar_set(self, s_row, rb_str_new2(""));
@@ -469,14 +453,14 @@ tr2:
       rb_yield(row);
     }
   }
-#line 168 "ext/fastcsv/fastcsv.rl"
+#line 150 "ext/fastcsv/fastcsv.rl"
 	{act = 3;}
 	goto st5;
 st5:
 	if ( ++p == pe )
 		goto _test_eof5;
 case 5:
-#line 480 "ext/fastcsv/fastcsv.c"
+#line 464 "ext/fastcsv/fastcsv.c"
 	switch( (*p) ) {
 		case 0: goto tr2;
 		case 10: goto tr3;
@@ -486,7 +470,7 @@ case 5:
 	}
 	goto st1;
 tr3:
-#line 49 "ext/fastcsv/fastcsv.rl"
+#line 46 "ext/fastcsv/fastcsv.rl"
 	{
     if (p == ts) {
       // Unquoted empty fields are nil, not "", in Ruby.
@@ -497,10 +481,8 @@ tr3:
       ENCODE;
     }
   }
-#line 117 "ext/fastcsv/fastcsv.rl"
+#line 101 "ext/fastcsv/fastcsv.rl"
 	{
-    mark_row_sep = p;
-
     curline++;
 
     if (d->start == 0 || p == d->start) {
@@ -520,10 +502,8 @@ tr3:
   }
 	goto st6;
 tr10:
-#line 117 "ext/fastcsv/fastcsv.rl"
+#line 101 "ext/fastcsv/fastcsv.rl"
 	{
-    mark_row_sep = p;
-
     curline++;
 
     if (d->start == 0 || p == d->start) {
@@ -546,10 +526,10 @@ st6:
 	if ( ++p == pe )
 		goto _test_eof6;
 case 6:
-#line 550 "ext/fastcsv/fastcsv.c"
+#line 530 "ext/fastcsv/fastcsv.c"
 	goto tr16;
 tr4:
-#line 49 "ext/fastcsv/fastcsv.rl"
+#line 46 "ext/fastcsv/fastcsv.rl"
 	{
     if (p == ts) {
       // Unquoted empty fields are nil, not "", in Ruby.
@@ -560,10 +540,8 @@ tr4:
       ENCODE;
     }
   }
-#line 117 "ext/fastcsv/fastcsv.rl"
+#line 101 "ext/fastcsv/fastcsv.rl"
 	{
-    mark_row_sep = p;
-
     curline++;
 
     if (d->start == 0 || p == d->start) {
@@ -583,10 +561,8 @@ tr4:
   }
 	goto st7;
 tr11:
-#line 117 "ext/fastcsv/fastcsv.rl"
+#line 101 "ext/fastcsv/fastcsv.rl"
 	{
-    mark_row_sep = p;
-
     curline++;
 
     if (d->start == 0 || p == d->start) {
@@ -609,14 +585,14 @@ st7:
 	if ( ++p == pe )
 		goto _test_eof7;
 case 7:
-#line 613 "ext/fastcsv/fastcsv.c"
+#line 589 "ext/fastcsv/fastcsv.c"
 	if ( (*p) == 10 )
 		goto st6;
 	goto tr16;
 tr13:
 #line 1 "NONE"
 	{te = p+1;}
-#line 49 "ext/fastcsv/fastcsv.rl"
+#line 46 "ext/fastcsv/fastcsv.rl"
 	{
     if (p == ts) {
       // Unquoted empty fields are nil, not "", in Ruby.
@@ -627,7 +603,7 @@ tr13:
       ENCODE;
     }
   }
-#line 138 "ext/fastcsv/fastcsv.rl"
+#line 120 "ext/fastcsv/fastcsv.rl"
 	{
     if (d->start == 0 || p == d->start) {
       rb_ivar_set(self, s_row, rb_str_new2(""));
@@ -644,14 +620,14 @@ tr13:
       rb_yield(row);
     }
   }
-#line 168 "ext/fastcsv/fastcsv.rl"
+#line 150 "ext/fastcsv/fastcsv.rl"
 	{act = 3;}
 	goto st8;
 st8:
 	if ( ++p == pe )
 		goto _test_eof8;
 case 8:
-#line 655 "ext/fastcsv/fastcsv.c"
+#line 631 "ext/fastcsv/fastcsv.c"
 	switch( (*p) ) {
 		case 10: goto tr15;
 		case 13: goto tr15;
@@ -660,7 +636,7 @@ case 8:
 	}
 	goto st1;
 tr14:
-#line 41 "ext/fastcsv/fastcsv.rl"
+#line 38 "ext/fastcsv/fastcsv.rl"
 	{
     unclosed_line = curline;
   }
@@ -669,7 +645,7 @@ st2:
 	if ( ++p == pe )
 		goto _test_eof2;
 case 2:
-#line 673 "ext/fastcsv/fastcsv.c"
+#line 649 "ext/fastcsv/fastcsv.c"
 	switch( (*p) ) {
 		case 0: goto st0;
 		case 34: goto tr8;
@@ -679,7 +655,7 @@ st0:
 cs = 0;
 	goto _out;
 tr8:
-#line 60 "ext/fastcsv/fastcsv.rl"
+#line 57 "ext/fastcsv/fastcsv.rl"
 	{
     if (p == ts) {
       field = rb_enc_str_new("", 0, encoding);
@@ -714,7 +690,7 @@ tr8:
       }
     }
   }
-#line 45 "ext/fastcsv/fastcsv.rl"
+#line 42 "ext/fastcsv/fastcsv.rl"
 	{
     unclosed_line = 0;
   }
@@ -723,7 +699,7 @@ st3:
 	if ( ++p == pe )
 		goto _test_eof3;
 case 3:
-#line 727 "ext/fastcsv/fastcsv.c"
+#line 703 "ext/fastcsv/fastcsv.c"
 	switch( (*p) ) {
 		case 0: goto tr9;
 		case 10: goto tr10;
@@ -757,9 +733,16 @@ case 3:
 	_out: {}
 	}
 
-#line 427 "ext/fastcsv/fastcsv.rl"
+#line 409 "ext/fastcsv/fastcsv.rl"
 
     if (done && cs < raw_parse_first_final) {
+      if (d->start == 0 || p == d->start) {
+        rb_ivar_set(self, s_row, rb_str_new2(""));
+      }
+      else if (p > d->start) {
+        rb_ivar_set(self, s_row, rb_str_new(d->start, p - d->start));
+      }
+
       FREE;
 
       if (unclosed_line) {

data/ext/fastcsv/fastcsv.rl

@@ -22,9 +22,6 @@ if (enc2 != NULL) { \
 #define FREE \
 if (buf != NULL) { \
   free(buf); \
-} \
-if (row_sep != NULL) { \
-  free(row_sep); \
 }
 
 static VALUE cClass, cParser, eError;
@@ -99,24 +96,9 @@ typedef struct {
 
   action mark_row {
     d->start = p;
-
-    if (len_row_sep) {
-      if (p - mark_row_sep != len_row_sep || row_sep[0] != *mark_row_sep || len_row_sep == 2 && row_sep[1] != *(mark_row_sep + 1)) {
-        FREE;
-
-        rb_raise(eError, "Unquoted fields do not allow \\r or \\n (line %d).", curline - 1);
-      }
-    }
-    else {
-      len_row_sep = p - mark_row_sep;
-      row_sep = ALLOC_N(char, p - mark_row_sep);
-      memcpy(row_sep, mark_row_sep, p - mark_row_sep);
-    }
   }
 
   action new_row {
-    mark_row_sep = p;
-
     curline++;
 
     if (d->start == 0 || p == d->start) {
@@ -202,11 +184,11 @@ static void rb_io_ext_int_to_encs(rb_encoding *ext, rb_encoding *intern, rb_enco
 
 static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
   int cs, act, have = 0, curline = 1, io = 0;
-  char *ts = 0, *te = 0, *buf = 0, *eof = 0, *mark_row_sep = 0, *row_sep = NULL;
+  char *ts = 0, *te = 0, *buf = 0, *eof = 0;
 
   VALUE port, opts, r_encoding;
   VALUE row = rb_ary_new(), field = Qnil, bufsize = Qnil;
-  int done = 0, unclosed_line = 0, len_row_sep = 0, buffer_size = 0, taint = 0;
+  int done = 0, unclosed_line = 0, buffer_size = 0, taint = 0;
   rb_encoding *enc = NULL, *enc2 = NULL, *encoding = NULL;
 
   Data *d;
@@ -369,7 +351,7 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
   while (!done) {
     VALUE str;
     char *p, *pe;
-    int len, space = buffer_size - have, tokstart_diff, tokend_diff, start_diff, mark_row_sep_diff;
+    int len, space = buffer_size - have, tokstart_diff, tokend_diff, start_diff;
 
     if (io) {
       if (space == 0) {
@@ -377,7 +359,6 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
         tokstart_diff = ts - buf;
         tokend_diff = te - buf;
         start_diff = d->start - buf;
-        mark_row_sep_diff = mark_row_sep - buf;
 
         buffer_size += BUFSIZE;
         REALLOC_N(buf, char, buffer_size);
@@ -387,7 +368,6 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
         ts = buf + tokstart_diff;
         te = buf + tokend_diff;
         d->start = buf + start_diff;
-        mark_row_sep = buf + mark_row_sep_diff;
       }
       p = buf + have;
 
@@ -428,6 +408,13 @@ static VALUE raw_parse(int argc, VALUE *argv, VALUE self) {
     %% write exec;
 
     if (done && cs < raw_parse_first_final) {
+      if (d->start == 0 || p == d->start) {
+        rb_ivar_set(self, s_row, rb_str_new2(""));
+      }
+      else if (p > d->start) {
+        rb_ivar_set(self, s_row, rb_str_new(d->start, p - d->start));
+      }
+
       FREE;
 
       if (unclosed_line) {

data/fastcsv.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |s|
   s.name        = "fastcsv"
-  s.version     = '0.0.3'
+  s.version     = '0.0.4'
   s.platform    = Gem::Platform::RUBY
   s.authors     = ["Open North"]
   s.email       = ["info@opennorth.ca"]

data/lib/fastcsv.rb

@@ -8,6 +8,10 @@ class FastCSV < CSV
     Parser.new.raw_parse(*args, &block)
   end
 
+  def row
+    parser && parser.row
+  end
+
   def shift
     # COPY
     # handle headers not based on document content

data/spec/fastcsv_spec.rb

@@ -147,6 +147,7 @@ RSpec.shared_examples 'a CSV parser' do
   it 'should raise an error on mixed row separators' do
     csv = "foo\rbar\nbaz\r\n"
     expect{CSV.parse(csv)}.to raise_error(CSV::MalformedCSVError, 'Unquoted fields do not allow \r or \n (line 2).')
+    skip
     expect{FastCSV.parse(csv)}.to raise_error(FastCSV::MalformedCSVError, 'Unquoted fields do not allow \r or \n (line 2).')
   end
 

data/test/csv/test_csv_parsing.rb

@@ -160,16 +160,16 @@ class TestCSV::Parsing < TestCSV
     assert_equal(6, lines.size)
     assert_match(/\Aline,4/, lines.find { |l| l =~ /some\rjunk/ })
 
-    csv = FastCSV.new(bad_data)
-    begin
-      loop do
-        assert_not_nil(csv.shift)
-        assert_send([csv.lineno, :<, 5]) # FIXME 4
-      end
-    rescue FastCSV::MalformedCSVError
-      assert_equal( "Unquoted fields do not allow \\r or \\n (line 4).",
-                    $!.message )
-    end
+    # csv = FastCSV.new(bad_data)
+    # begin
+    #   loop do
+    #     assert_not_nil(csv.shift)
+    #     assert_send([csv.lineno, :<, 5]) # FIXME 4
+    #   end
+    # rescue FastCSV::MalformedCSVError
+    #   assert_equal( "Unquoted fields do not allow \\r or \\n (line 4).",
+    #                 $!.message )
+    # end
 
     assert_raise(FastCSV::MalformedCSVError) { FastCSV.parse_line('1,2,"3...') }
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: fastcsv
 version: !ruby/object:Gem::Version
-  version: 0.0.3
+  version: 0.0.4
 platform: ruby
 authors:
 - Open North