<feed xmlns='http://www.w3.org/2005/Atom'>
<title>lua-cjson/g_fmt.c, branch or-1.31.0</title>
<subtitle>A mirror of https://github.com/openresty/lua-cjson
</subtitle>
<id>https://git.lua4.win/lua-cjson/atom?h=or-1.31.0</id>
<link rel='self' href='https://git.lua4.win/lua-cjson/atom?h=or-1.31.0'/>
<link rel='alternate' type='text/html' href='https://git.lua4.win/lua-cjson/'/>
<updated>2012-03-04T08:24:35+00:00</updated>
<entry>
<title>Fix string length returned by g_fmt.c for |x|&lt;1</title>
<updated>2012-03-04T08:24:35+00:00</updated>
<author>
<name>Mark Pulford</name>
<email>mark@kyne.com.au</email>
</author>
<published>2012-01-19T10:22:07+00:00</published>
<link rel='alternate' type='text/html' href='https://git.lua4.win/lua-cjson/commit/?id=bb296abe813f7ded82e74909fdb0f7e4f1f42416'/>
<id>urn:sha1:bb296abe813f7ded82e74909fdb0f7e4f1f42416</id>
<content type='text'>
Fix string length returned by built in fpconv_g_fmt() so it doesn't
include the trailing null when converting numbers |x|&lt;1.
</content>
</entry>
<entry>
<title>Use internal dtoa/strtod for double conversion</title>
<updated>2012-03-04T08:24:34+00:00</updated>
<author>
<name>Mark Pulford</name>
<email>mark@kyne.com.au</email>
</author>
<published>2011-12-29T12:33:58+00:00</published>
<link rel='alternate' type='text/html' href='https://git.lua4.win/lua-cjson/commit/?id=ca42b9a996f9046ba3876ad8a81cda1d935b39cf'/>
<id>urn:sha1:ca42b9a996f9046ba3876ad8a81cda1d935b39cf</id>
<content type='text'>
The internal Lua CJSON dtoa/strtod routines have locale support
disabled. This avoids problems under locales with comma decimal
separators.

Build changes:
- CMake: Check for big endian architectures
- Makefile: Provide option to build with dtoa.c

Modifications to dtoa.c:
- Include locale dtoa_config.h configuration
- Rename Infinity/NaN to inf/nan to match common C libraries
- Rename strtod() -&gt; internal_strtod() to prevent conflict with libc
  function

Modifications to g_fmt.c:
- Return output string length (instead of original buffer pointer)
- Provide precision as an argument to g_fmt()
- Silence compilations warnings from vendor source
  - while(a = b)
  - Unused label "done:"
- Only swap to scientific notation when once the number of decimal
  digits required exceeds the precision available. This matches
  standard printf format %g.
- Display a "0" in front of numbers &lt; 1.
</content>
</entry>
<entry>
<title>Added Netlib dtoa.c/g_fmt.c routines (20110428)</title>
<updated>2011-12-28T21:40:57+00:00</updated>
<author>
<name>Mark Pulford</name>
<email>mark@kyne.com.au</email>
</author>
<published>2011-12-28T21:40:57+00:00</published>
<link rel='alternate' type='text/html' href='https://git.lua4.win/lua-cjson/commit/?id=23d19ba2551e6f432ada6f7cd5ac3e874b4d35d1'/>
<id>urn:sha1:23d19ba2551e6f432ada6f7cd5ac3e874b4d35d1</id>
<content type='text'>
See "www.netlib.org/fp/changes" for details.
</content>
</entry>
</feed>
