/
Changes
66 lines (55 loc) · 2.72 KB
/
Changes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Release 0.1.3
- Changed AST for PostScript 'if' and 'ifelse'. Now returns a 'cond' node.
- Added :strict option to PDF::Grammar::Content::Actions. Will cause parser
to abort whith an error when unrecognised content is encountered.
Release 0.1.2
- Optimized hex string parsing. Hex codes in names must have 2-digits
- Avoid shape errors in xref sections, when actual entries and
object count don't match (let PDF::Reader detect and handle this).
- Regen META6.json via mi6
Release 0.1.1
- PDF::Grammar::Doc::Actions now produces compact uint64 shaped indexes
- Change PDF::Grammar :AST-Types from a Set to an Enum.
- Allow whitespace between trailer dict and 'startxref'
Release 0.1.0
- Changed xref entries AST from a hash to a uint32 @array[3]
Release 0.0.8
- Renamed tokens char_delimiter -> char-delimiter, literal_delimiter -> literal-delimiter
- remove custom PDF::Grammar.parsefile method. Not needed in latest rakudo
Release 0.0.7
- Optimized stream and image tokens
- Added PDF::Grammar::Content::Fast. A faster, more forgiving
alternative to PDF::Grammar::Content
Release 0.0.6
- Make META.info version consistent with PDF::Grammar:ver<...>
- Declare "perl6" : "6" in META.info
- new <xref-first> rule. A variation of <xref> that assumes we've
hopped over the first few bytes. Needed by PDF::Reader to (gulp)
handle the /T offset entry in linearization dictionary.
- minor optimisations to <ws> and <name> rules
- replaced a couple of regex with rules/tokens
- <xref-section> - allow zero <xref-entry>'s
As observed in some real-world PDF's
Release 0.0.5
- Replaced JSON::Tiny with JSON::Fast
- Removed PDF::Grammar::Stream
- Slightly relax trailer parse. Don't require a \n before trailer dictionary
- Fixed whitespace handling in <PDF::Grammar::Doc::ind-obj-nibble> rule
Release 0.0.4
- Added `AST Reference` section to README.
- Renamed a couple of token tags for consistency:
-- object-first-num => obj-first-num
-- object-count => obj-count
Release 0.0.3
- Reworked 'stream' AST to return :encoded stream content. Rather
than :start and :end positions of the stream.
This was originally intended as a optimization. In practice the reader
only uses this rule in :repair mode, in which case it always extracts
the stream data anyway.
- tweaked ind-obj-nibble rule to use .codes rather than .chars to compute
offset of stream start - safer in a post-NFG world. Latin encoding now
has "\n\r" as a combining newline character.
Release 0.0.2
- General fixes for the 2015 Rakudo GLR
Release 0.0.1
- Initial development