Commit graph

30 commits

Author SHA1 Message Date
Ferdinand Majerech 601f10bf2f New loader ctor in errors unittests. 2014-07-31 03:22:13 +02:00
Ferdinand Majerech 9fe6950df0 Style. 2014-07-31 03:21:58 +02:00
Ferdinand Majerech ee2df932a0 New Loader ctor used in Emitter tests. 2014-07-31 03:16:22 +02:00
Ferdinand Majerech eb29c2626d Style. 2014-07-31 03:15:56 +02:00
Ferdinand Majerech 337938f090 Using new Loader ctor in unicode input tests. 2014-07-31 03:12:57 +02:00
Ferdinand Majerech ee0a449977 Style. 2014-07-31 03:12:25 +02:00
Ferdinand Majerech c5b7987e68 Using non-Stream Loader ctor in Representer test. 2014-07-31 03:12:02 +02:00
Ferdinand Majerech 21700fdb5d Style. 2014-07-31 03:11:37 +02:00
Ferdinand Majerech 505a26ff4e Reader unittest now constructs Reader from buffer. 2014-07-31 02:32:23 +02:00
Ferdinand Majerech 3c36d4ad63 Replaced some old 'in' params with const 2014-07-28 23:20:35 +02:00
Ferdinand Majerech 405a61891f Reader unittest fixes. 2014-07-22 01:13:01 +02:00
Ferdinand Majerech c856a35eb0 More readable tests. 2014-07-19 04:09:09 +02:00
Ferdinand Majerech 4faaf77950 Removed deprecation errors, improved readability. 2014-07-19 03:51:52 +02:00
Ferdinand Majerech a341340d6c Using the dyaml.all module instead of yaml (which will eventually be removed). 2013-12-17 15:09:19 +01:00
Kiith-Sa a409cceeff DMD 2.053 compatibility. 2013-05-23 14:33:34 +02:00
kiith-sa edf3e2a799 Using std.typecons.Flag where possible, and more in->const. 2012-09-14 01:21:01 +02:00
Ferdinand Majerech 9596806644 Red-Black Trees are now used for duplicate detection, and planned
to be used for unordered map storage. This is because AAs still
don't work correctly and even if they did, require the user to
define both toHash and opCmp/opEquals for every YAML
struct/class. Now only opCmp needs to be defined.
Documentation/tutorials/examples have been updated accordingly.
2012-01-23 15:57:26 +01:00
Ferdinand Majerech fbc962623d Simplified the Constructor API. 2011-11-17 23:53:24 +01:00
Ferdinand Majerech 2c9d464389 Reader was reimplemented.
After experiments with loading the whole file at once, and
with decoding and parsing in separate thread, lazy reader
turned to be the fastest/least memory intensive solution.
Characters are now decoded in small batches.
This improved parsing speed by ~20%.

No global state anymore. Anchors are now zero terminated strings
and TagDirectives are a simple array. Event structure was changed
to prevent size increase.
Minor fixes and improvements.
2011-11-16 03:10:29 +01:00
Ferdinand Majerech 8208e817de Updated Node for much more constness (unfortunately, quite some
code duplication as well). No longer need a custom Variant.
2011-10-30 18:12:02 +01:00
Ferdinand Majerech 9d68b6fa9a Optimized GC performance when loading,
greatly decreasing memory usage and
improving speed by 30% or more.
2011-10-23 20:17:37 +02:00
Ferdinand Majerech 13ea5f0c24 Added a shortcut alias called "as" for Node.get(), and replaced
get() with as() all over the code, tutorials, examples and docs.
Fixed a bug in YAML benchmark makefile.
Fixed a bug in autoddoc configuration.
2011-10-22 17:06:32 +02:00
Ferdinand Majerech 33a376b038 String representation of a Node's tag can now be accessed by the
user.
2011-10-20 15:48:21 +02:00
Ferdinand Majerech 548480b06b Changed the Constructor API (for loading of custom types) to
make it easier to load custom classes/structs. Updated API docs,
tutorials and examples accordingly.
2011-10-17 12:53:38 +02:00
Ferdinand Majerech 34b11405d4 Implemented the resolver unittest.
Changed Loader API to be in line with Dumper,
and updated examples, tutorials and docs.
2011-10-12 23:49:42 +02:00
Ferdinand Majerech 934df763ad Most of emitter work done. There are some more changes left until 0.2,
and the turorials and documentation still need some updating.
2011-10-11 15:58:23 +02:00
Ferdinand Majerech b0f2ce201a Added Node constructors needed to create new nodes to emit. 2011-08-21 03:56:21 +02:00
Ferdinand Majerech dde7d2f64f Node equality test now takes tag into account, as the
specification requires.
2011-08-20 22:38:32 +02:00
Ferdinand Majerech 7192503fe6 Implemented a Tag struct taking as little memory as possible.
Removed endMark from Node to keep it in 32 bytes on 64bit.
This will result in slightly worse debugging messages, but we
still have the start position of a node.

Tag is needed for better compliance with the spec and emitting
support for multiple tags with the same D data type.
2011-08-20 22:15:20 +02:00
Ferdinand Majerech 283c42bf8f Initial commit. 2011-08-16 14:53:13 +02:00