Commit Graph

58 Commits

Author SHA1 Message Date
kpdecker b695aaec7c Use depths array rather than passing array args
Approximately doubles the throughput performance of depthed templates and clears the way for doing recursive path lookups in pathed mode.
2014-08-13 14:11:53 -05:00
kpdecker f6dc5ad716 Create contextName helper for context lookup 2014-08-13 11:41:27 -05:00
kpdecker e93d516672 Render indent for standalone partials 2014-08-12 14:41:58 -05:00
kpdecker 63c5eb7a82 Remove unused JSCompiler.register method 2014-08-11 20:10:12 -05:00
kpdecker 271106d43f Do not lookup pathed helpers on the helper stack
Fixes #764
2014-07-12 12:50:54 -05:00
kpdecker 1fb7b51ee6 Fix rendering of paths that resolve to zero
Fixes #820
2014-07-12 11:43:45 -05:00
kpdecker c90cfe247c Use ID lookup for ambiguous helper evaluation
Final change for #731
2014-07-07 00:20:02 -05:00
kpdecker b5a5c76ceb Rework lookup null protector logic
- Move the lookup null protection out of `nameLookup` and into that contexts that are aware of the needs for falsy vs. not displayed values.
- Optimize lookup for nested path operations

Fixes #731
2014-07-06 23:41:15 -05:00
kpdecker 4aad72d223 Move lambda resolution to runtime
This has a very positive impact on precompiled output size, particularly for known-helpers cases, and little or no impact on the throughput numbers.
2014-07-06 12:53:01 -05:00
kpdecker b79e31e5a5 Remove disassemble log statement
Fixes #772
2014-06-28 18:44:11 -07:00
kpdecker 55782f08c5 Fix evaluation of paths and subexprs
Fixes #743
2014-03-05 17:49:39 -06:00
Kevin Decker 7dd64e5e92 Merge pull request #749 from jenseng/option_subexpressions_fix
properly handle multiple subexpressions in the same hash, fixes #748
2014-03-05 14:38:26 -06:00
Jon Jensen 5999baf746 fix bug in IE<=8 (no Array::map), closes #751
looks like this broke in 2812fe27

also s/this.i/var i/, since the functions that needed it were removed in
a1edab6ef
2014-03-04 22:18:50 -07:00
Jon Jensen 8493822b62 properly handle multiple subexpressions in the same hash, fixes #748
push all hash params before popping any so as to avoid the last stackN var
stomping previous ones
2014-02-20 17:37:25 -06:00
kpdecker 16f135835e Add support for depthed resolution of data fields 2014-02-09 18:35:22 -06:00
kpdecker 7a7ad74ff1 Optimize buffer generate first and all edge cases 2014-02-09 14:24:18 -06:00
Kevin Decker bc8f6792e8 Merge pull request #729 from wycats/precompile-literal
Convert template spec to object literal
2014-02-08 20:01:52 -06:00
kpdecker 2812fe2775 Convert template spec to object literal
This allows for metadata to be associated with the template and a simplification of the template init logic.
2014-02-06 23:05:45 -08:00
kpdecker b0e7298523 Remove unnecessary conditional 2014-02-06 23:01:43 -08:00
kpdecker ace2896ec8 Add trackIds compiler flag
Allows helpers that care about where a particular field came from derive this data while maintaining backward compatibility with existing helpers.
2014-01-17 23:15:17 -06:00
kpdecker 33921beaa0 Fix missing parameters for pathed mustaches
Fixes #658
2014-01-17 23:09:59 -06:00
kpdecker 45ae86a248 Implement partial hash evaluation 2014-01-17 17:51:52 -06:00
Blake Embrey 13633e7896 Improve usefulness of extend util, properly use namespace property, update setup options to use a hash helper. 2014-01-16 21:24:26 +10:00
Kevin Decker 5659db4877 Merge pull request #634 from wycats/name-option
It would be great to have the helper name passed to `blockHelperMissing`
2014-01-15 08:42:32 -08:00
kpdecker f650b0d69e Expose the initial context via @root 2014-01-15 09:41:53 -06:00
kpdecker d385e2cab6 Cleanup now unused code in jscompiler 2014-01-06 16:07:56 -06:00
kpdecker 0f328f375e Include name option for all helper calls
All helper calls will have access to `options.name` which is the first id value of the mustache operation.

As part of this the helperMissing call has been simplified to remove the indexed name in order to optimize the call. This is a breaking change.

Fixes #634
2014-01-06 16:07:56 -06:00
kpdecker daeecf825c Track stringParams mode in local state vars 2014-01-06 02:55:13 -06:00
Blake Embrey 3d9685c6a2 Remove redundant conditions. 2014-01-05 12:14:24 -05:00
kpdecker 6d996ef270 Simplify ambiguous code
Remove if conditional in favor of boolean failover.
2014-01-04 09:19:58 -06:00
kpdecker e0d0ad5027 Merge branch 'compiler-options' of github.com:blakeembrey/handlebars.js into blakeembrey-compiler-options
Conflicts:
	lib/handlebars/compiler/javascript-compiler.js
2014-01-01 19:38:31 -06:00
kpdecker 3c85720137 Remove duplication from generated subexpressions 2014-01-01 18:59:21 -06:00
kpdecker e7e94dc697 Whitespace cleanup 2013-12-31 21:20:59 -06:00
kpdecker c1a93d33e7 Use literal for data lookup 2013-12-31 21:19:00 -06:00
kpdecker cd885bf855 Add stack handling sanity checks 2013-12-31 21:18:48 -06:00
kpdecker ddfe457abf Fix stack id "leak" on replaceStack 2013-12-31 21:17:50 -06:00
kpdecker f4d337d252 Fix incorrect stack pop when replacing literals 2013-12-31 21:17:16 -06:00
kpdecker af3358d195 Fix multiple hash handling in subexpressions 2013-12-31 16:10:51 -06:00
machty b09333db79 Added support for subexpressions
Handlebars now supports subexpressions.

    {{foo (bar 3)}}

Subexpressions are always evaluated as helpers; if
`3` were omitted from the above example, `bar`
would be invoked as a param-less helper, even
though a top-levell `{{bar}}` would be considered
ambiguous.

The return value of a subexpression helper is
passed in as a parameter of a parent subexpression
helper, even in string params mode. Their type,
as listed in `options.types` or `options.hashTypes`
in string params mode, is "sexpr".

The main conceptual change in the Handlebars code
is that there is a new AST.SexprNode that manages
the params/hash passed into a mustache, as well
as the logic that governs whether that mustache
is a helper invocation, property lookup, etc.
MustacheNode, which used to manage this stuff,
still exists, but only manages things like
white-space stripping and whether the mustache
is escaped or not. So, a MustacheNode _has_
a SexprNode.

The introduction of subexpressions is fully
backwards compatible, but a few things needed
to change about the compiled output of a template
in order to support subexpressions. The main one
is that the options hash is no longer stashed in
a local `options` var before being passed to
either the helper being invoked or the
`helperMissing` fallback. Rather, the options
object is inlined in these cases. This does
mean compiled template sizes will be a little
bit larger, even those that don't make use of
subexpressions, but shouldn't have any noticeable
impact on performance when actually rendering
templates as only one of these inlined objects
will actually get evaluated.
2013-12-30 19:07:04 -05:00
Blake Embrey 150e55aa00 Pull options out from param setup to allow easier extension. 2013-12-28 22:31:45 -05:00
Kevin Decker 2dae244185 Merge pull request #684 from blakeembrey/javascript-variable
Allow any number of trailing characters for valid JavaScript variable
2013-12-25 10:12:11 -08:00
Blake Embrey f3a90812a6 Allow any number of trailing characters. 2013-12-24 16:14:57 -05:00
kpdecker bbc7c7d080 Expose compilerInfo and checkRevision as APIs
Fixes #656
2013-12-23 21:00:59 -06:00
kpdecker ded0a1617f Protect context-lookups from undefined values
Fixes #166
Fixes #587
2013-11-03 10:58:37 -06:00
kpdecker c776f171ab Implement strip opcode 2013-10-14 22:50:50 -05:00
kpdecker 6b0b3fa2ca Defer content output
Allows for stripping of the content after the fact.
2013-10-14 22:50:50 -05:00
kpdecker a1edab6ef5 Kill unused functions 2013-10-14 22:50:50 -05:00
kpdecker 4cf49732a2 Push Source helper method 2013-10-14 22:50:49 -05:00
kpdecker 06d94fed56 Fix argument not found error 2013-10-14 22:49:34 -05:00
kpdecker 4fda71364e Pass options hash to context function exec
Fixes #599
2013-10-14 21:30:58 -05:00