wikitext 1.6 → 1.7

data/ext/parser.h

@@ -35,8 +35,6 @@ VALUE Wikitext_parser_sanitize_link_target(VALUE self, VALUE string);
 
 VALUE Wikitext_parser_encode_link_target(VALUE self, VALUE in);
 
-VALUE Wikitext_parser_encode_special_link_target(VALUE self, VALUE in);
-
 VALUE Wikitext_parser_parse(int argc, VALUE *argv, VALUE self);
 
 VALUE Wikitext_parser_profiling_parse(VALUE self, VALUE string);

data/ext/str.c

@@ -23,6 +23,10 @@
 
 #include "str.h"
 
+// when allocating memory, reserve a little more than was asked for,
+// which can help to avoid subsequent allocations
+#define STR_OVERALLOC 256
+
 str_t *str_new(void)
 {
     str_t *str      = ALLOC_N(str_t, 1);
@@ -32,31 +36,13 @@ str_t *str_new(void)
     return str;
 }
 
-str_t *str_new_size(long len)
-{
-    str_t *str      = ALLOC_N(str_t, 1);
-    str->ptr        = ALLOC_N(char, len);
-    str->len        = 0;
-    str->capacity   = len;
-    return str;
-}
-
-str_t *str_new_copy(char *src, long len)
+str_t *str_new_copy(const char *src, long len)
 {
     str_t *str      = ALLOC_N(str_t, 1);
-    str->ptr        = ALLOC_N(char, len);
+    str->ptr        = ALLOC_N(char, len + STR_OVERALLOC);
     memcpy(str->ptr, src, len);
     str->len        = len;
-    str->capacity   = len;
-    return str;
-}
-
-str_t *str_new_no_copy(char *src, long len)
-{
-    str_t *str      = ALLOC_N(str_t, 1);
-    str->ptr        = src;
-    str->len        = len;
-    str->capacity   = len;
+    str->capacity   = len + STR_OVERALLOC;
     return str;
 }
 
@@ -76,23 +62,23 @@ void str_grow(str_t *str, long len)
     if (str->capacity < len)
     {
         if (str->ptr)
-            REALLOC_N(str->ptr, char, len);
+            REALLOC_N(str->ptr, char, len + STR_OVERALLOC);
         else
-            str->ptr = ALLOC_N(char, len);
-        str->capacity = len;
+            str->ptr = ALLOC_N(char, len + STR_OVERALLOC);
+        str->capacity = len + STR_OVERALLOC;
     }
 }
 
-void str_append(str_t *str, char *src, long len)
+void str_append(str_t *str, const char *src, long len)
 {
     long new_len = str->len + len;
     if (str->capacity < new_len)
     {
         if (str->ptr)
-            REALLOC_N(str->ptr, char, new_len);
+            REALLOC_N(str->ptr, char, new_len + STR_OVERALLOC);
         else
-            str->ptr = ALLOC_N(char, new_len);
-        str->capacity = new_len;
+            str->ptr = ALLOC_N(char, new_len + STR_OVERALLOC);
+        str->capacity = new_len + STR_OVERALLOC;
     }
     memcpy(str->ptr + str->len, src, len);
     str->len = new_len;
@@ -103,12 +89,9 @@ void str_append_str(str_t *str, str_t *other)
     str_append(str, other->ptr, other->len);
 }
 
-void str_swap(str_t **a, str_t **b)
+void str_append_string(str_t *str, VALUE other)
 {
-    str_t *c;
-    c = *a;
-    *a = *b;
-    *b = c;
+    str_append(str, RSTRING_PTR(other), RSTRING_LEN(other));
 }
 
 void str_clear(str_t *str)

data/ext/str.h

@@ -37,15 +37,8 @@ typedef struct
 // create a new, empty string struct
 str_t *str_new(void);
 
-// create a new, empty string struct with capacity len
-str_t *str_new_size(long len);
-
 // create a new string struct and initialize it with a copy of the buffer of length len pointed to by src
-str_t *str_new_copy(char *src, long len);
-
-// create a new string struct and initialize it with the buffer of length len pointed to by src
-// no copy is made; the struct takes ownership of the buffer and will free it when the struct is disposed of
-str_t *str_new_no_copy(char *src, long len);
+str_t *str_new_copy(const char *src, long len);
 
 // convenience method for testing
 str_t *str_new_from_string(VALUE string);
@@ -56,14 +49,13 @@ VALUE string_from_str(str_t *str);
 // grows a string's capacity to the specified length
 void str_grow(str_t *str, long len);
 
-void str_append(str_t *str, char *src, long len);
+void str_append(str_t *str, const char *src, long len);
 
 // appends the "other" string struct onto str
 void str_append_str(str_t *str, str_t *other);
 
-// this is a temporary convenience measure
-// later on if I develop in-place variants of some functions this won't be needed
-void str_swap(str_t **a, str_t **b);
+// appends the "other" string (a Ruby String) onto str
+void str_append_string(str_t *str, VALUE other);
 
 // don't actually free the memory yet
 // this makes str structs very useful when reusing buffers because it avoids reallocation

data/ext/token.c

@@ -105,7 +105,7 @@ VALUE Wikitext_parser_token_types(VALUE self)
 }
 
 // for testing and debugging only
-VALUE _Wikitext_token(token_t *token)
+VALUE wiki_token(token_t *token)
 {
     VALUE object = rb_class_new_instance(0, NULL, cWikitextParserToken);
     (void)rb_iv_set(object, "@start",           LONG2NUM((long)token->start));

data/ext/token.h

@@ -24,8 +24,8 @@
 #include "ruby_compat.h"
 #include <stdint.h>     /* uint32_t */
 
-#define TOKEN_TEXT(token)   rb_str_new((const char *)token->start, (token->stop - token->start))
 #define TOKEN_LEN(token)    (token->stop - token->start)
+#define TOKEN_TEXT(token)   rb_str_new((const char *)token->start, TOKEN_LEN(token))
 
 typedef struct
 {
@@ -114,4 +114,4 @@ enum token_types {
 
 VALUE Wikitext_parser_token_types(VALUE self);
 
-VALUE _Wikitext_token(token_t *token);
+VALUE wiki_token(token_t *token);

data/ext/wikitext.c

@@ -44,7 +44,6 @@ void Init_wikitext()
     rb_define_method(cWikitextParser, "fulltext_tokenize", Wikitext_parser_fulltext_tokenize, -1);
     rb_define_singleton_method(cWikitextParser, "sanitize_link_target", Wikitext_parser_sanitize_link_target, 1);
     rb_define_singleton_method(cWikitextParser, "encode_link_target", Wikitext_parser_encode_link_target, 1);
-    rb_define_singleton_method(cWikitextParser, "encode_special_link_target", Wikitext_parser_encode_special_link_target, 1);
     rb_define_attr(cWikitextParser, "line_ending", Qtrue, Qtrue);
     rb_define_attr(cWikitextParser, "internal_link_prefix", Qtrue, Qtrue);
     rb_define_attr(cWikitextParser, "img_prefix", Qtrue, Qtrue);

data/lib/wikitext/version.rb

@@ -22,5 +22,5 @@
 # POSSIBILITY OF SUCH DAMAGE.
 
 module Wikitext
-  VERSION = '1.6'
+  VERSION = '1.7'
 end # module Wikitext

data/spec/external_link_spec.rb

@@ -326,4 +326,21 @@ describe Wikitext::Parser, 'external links' do
       @parser.parse("> [http://google.com/\n").should == expected # was a bug
     end
   end
+
+  describe 'regressions' do
+    # assorted examples
+    it 'should not turn failed absolute links into external hyperlinks' do
+      # was emitting: <p>[<a href="/hello" class="external">/hello</a> this</p>\n
+      expected =   %Q{<p>[<a href="/hello">/hello</a> this</p>\n}
+      @parser.parse('[/hello this').should == expected
+
+      # was emitting: <p>[<a href="/hello" class="external">/hello</a> </p>\n
+      expected =   %Q{<p>[<a href="/hello">/hello</a> </p>\n}
+      @parser.parse('[/hello ').should == expected
+
+      # was emitting: <h1>hello [<a href="/hello" class="external">/hello</a> </h1>\n
+      expected =   %Q{<h1>hello [<a href="/hello">/hello</a> </h1>\n}
+      @parser.parse('= hello [/hello =').should == expected
+    end
+  end
 end

data/spec/internal_link_spec.rb

@@ -823,9 +823,15 @@ describe Wikitext::Parser, 'internal links (space to underscore on)' do
     end
 
     describe 'missing link text' do
-      it 'should use link target' do
+      it 'should use link target (zero-width link text)' do
         @parser.parse('[[foo|]]').should == %Q{<p><a href="/wiki/foo">foo</a></p>\n}
       end
+
+      # was a bug in version <= 1.6
+      # emitted: <p><a href="/wiki/foo"></a></p>\n
+      it 'should use link target (blank link text)' do
+        @parser.parse('[[foo| ]]').should == %Q{<p><a href="/wiki/foo">foo</a></p>\n}
+      end
     end
 
     describe 'link cut off at separator (end-of-file)' do

data/spec/link_encoding_spec.rb

@@ -94,51 +94,4 @@ describe Wikitext, 'encoding a link target' do
       Wikitext::Parser.encode_link_target(string).should == URI.escape(string, reserved).downcase
     end
   end
-
-  # "special" links don't get transformed in any way
-  describe 'special links' do
-
-    # as of version 1.4.0 the encode_link_target function no longer handles special links
-    it 'should (no longer) recognize links which match /\A[a-z]+\/\d+\z/ as being special' do
-      string = 'foo/10'
-      Wikitext::Parser.encode_special_link_target(string).should == 'foo%2f10'
-      Wikitext::Parser.encode_link_target(string).should == 'foo%2f10'
-    end
-
-    it "should not recognize links which don't match at /\A/ as being special" do
-      string = '  foo/10'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-      string = '..foo/10'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-      string = '12foo/10'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-    end
-
-    it "should not recognize links which don't match at /\z/ as being special" do
-      string = 'foo/10  '
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-      string = 'foo/10__'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-      string = 'foo/10##'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-      string = 'foo/10ab'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-    end
-
-    it "should not recognize links which don't match at /[a-z]/ (case differences) as being special" do
-      string = 'FOO/10'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-    end
-
-    it "should not recognize links which don't match at /[0-9]/ (case differences) as being special" do
-      string = 'foo/xx'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-    end
-
-    it "should not recognize links which don't match at /\// as being special" do
-      string = 'foo 10'
-      Wikitext::Parser.encode_special_link_target(string).should_not == string
-    end
-  end
 end
-

data/spec/link_sanitizing_spec.rb

@@ -103,7 +103,7 @@ describe Wikitext, 'sanitizing a link target' do
     Wikitext::Parser.sanitize_link_target('foo, "bar" & baz €').should == 'foo, "bar" & baz €'
   end
 
-  # here we're exercising the _Wikitext_utf8_to_utf32 function
+  # here we're exercising the wiki_utf8_to_utf32 function
   describe 'with invalidly encoded input' do
     it 'should raise an exception for missing second byte' do
       lambda {

data/spec/regressions_spec.rb

@@ -130,7 +130,7 @@ describe Wikitext::Parser, 'regressions' do
 
   # this is the general case of the bug covered in the previous spec
   # any token that appears as the first token after a PRE token can manifest this bug
-  # PRINTABLE didn't only because it called _Wikitext_start_para_if_necessary(), which handled the pending CRLF
+  # PRINTABLE didn't only because it called wiki_start_para_if_necessary(), which handled the pending CRLF
   it 'should emit pending newlines for all token types found inside PRE and PRE_START blocks' do
     # PRE_START
     input = dedent <<-END
@@ -441,7 +441,7 @@ describe Wikitext::Parser, 'regressions' do
     END
     @parser.parse(input).should == expected
 
-    # these tokens weren't affected by the bug, seeing as they either call _Wikitext_start_para_if_necessary()
+    # these tokens weren't affected by the bug, seeing as they either call wiki_start_para_if_necessary()
     # or they can only appear in PRE_START (not PRE) thanks to the tokenizer
     # but we add specs for them to make sure that the issue never crops up for them in the future
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: wikitext
 version: !ruby/object:Gem::Version 
-  version: "1.6"
+  version: "1.7"
 platform: ruby
 authors: 
 - Wincent Colaiuta
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-05-07 00:00:00 +02:00
+date: 2009-05-13 00:00:00 +02:00
 default_executable: 
 dependencies: []