-
Notifications
You must be signed in to change notification settings - Fork 1
/
changelog.txt
217 lines (217 loc) · 12.8 KB
/
changelog.txt
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
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
* 1.44 - October 28, 2010
o Added new feature for collating PDF page scans: shuffle. Please see
the man page for usage details.
o Introduced update_info_utf8, dump_data_utf8 and
dump_data_fields_utf8 to provide UTF-8 companions to update_info,
dump_data and dump_data_fields. These latter operations use XML
numerical entities to encode non-ASCII characters. In version 1.43,
we changed the encoding for update_info to UTF-8, but that made it
incompatible with dump_data and also broke some downstream
applications. By introducing these UTF-8 operations, we can revert
update_info to its original behavior.
o Burst feature now copies the metadata (including XMP) from the
input file to the output pages.
o Updated Bouncy Castle library to 1.45.
o Removed or replaced third-party code that wasn't compatible with
pdftk's GPL license.
o Updated third-party license information.
* 1.43 - September 30, 2010
o Improved input handle detection to reduce false hits.
o Improved keyword detection logic to eliminate false hits when input
filenames happen to include pdftk keywords even, odd and end.
o Added option of prompting the user for the output when bursting a
PDF. Also reviewed other filename prompting code.
o Changed the PDF parser to accept name tokens longer than 127
characters -- the PDF Specification says that 127 is the limit.
This isn't related to file names. The issue arose with PDFs created
by Acrobat Web Capture 9.0.
o Fixed a problem with filling form choice fields in some PDFs where
the old form value was 'sticking.'
o Changed pdftk behavior when handling subset fonts so it doesn't
alter font name "tags." This was causing printing problems with
Acrobat 3.01 on Windows.
o Fixed a stream parsing bug that was causing page content to
disappear after merge of PDFs generated by Microsoft Reporting
Services PDF Rendering Extension 10.0.0.0.
o Added multistamp and multibackground features provided by a Debian
patch -- thanks!
o Clear the signal mask as workaround to environments that turn off
signals before calling pdftk. This problem is known to cause pdftk
to hang_in_some_Python_web_setups as well as in PHP.
o Set locale to C as workaround to an unusual exception. This is a
Debian_patch. Please let me know if it causes any troubles.
o Improved reporting of output errors via Debian patch -- thanks!
o Added support for UTF-8 data in update_info via Debian patch -
- thanks!
o Added support for UTF-8 filenames via Debian patch -- thanks!
o Updated build procedure to work better with newer versions of GCC.
Maintained compatibility with older versions of GCC.
o Added license information to the source tree for the third-party
libraries that pdftk uses.
* 1.41 - November 28, 2006
o Fixed a bug that corrupted output PDF xref tables. This corruption
was mild but universal. Most PDF tools can cope with the corrupted
PDFs, but I recommend upgrading from 1.40 to 1.41 as soon as
possible. This bug was introduced in version 1.40 -- version 1.12
does not have this bug.
o Fixed a bug that prevented XFDF form data from being passed to
pdftk via stdin.
o Commented out some unused code from pdftk.cc.
* 1.40 - September 19, 2006
o Added the stamp operation, a natural complement to the existing
background operation.
o Added the page rotating patch provided by David Fabel -- thanks!
Tweaked the patch so it handles a greater variety of input syntax
(e.g., 1-20evenE).
o Added the generate_fdf patch provided Bernhard R. Link -- thanks! I
actually rewrote the patch so it uses FDF features built into the
iText library. Please let me know my changes break anything
downstream.
o The fill_form operation can now take XFDF data as well as FDF data.
This feature was sponsored by Vesaria -- thanks!
o Added the drop_xfa option so pdftk could fill forms created with
newer versions of Acrobat or Adobe Designer. Read more about this
above.
o Added the keep_first_id and keep_final_id options for more PDF fun.
o Upgraded the iText library we use to itext-paulo rev. 155. This
makes pdftk harder to compile on older versions of gcc.
o Added the -O2 optimizing switch to Makefile GCJFLAGS. This should
make pdftk leaner and meaner, but could be dropped if your build
acts funny (like segfaulting).
o Fixed a bug that caused pdftk to create bloated PDFs when input PDF
pages had links on their pages.
o Added License-Adobe.txt to the fonts folder, as required for
distribution of Adobe's AFM files.
* 1.12 - November 9, 2004
o Fixed a bug where the presense of page annotations in some PDFs
would cause pdftk to crash. This bug first emerged when processing
a PDF created by FPDF (version 1.52) that contained web links.
Turns out that pdftk erroneously expected all page annotations to
be indirect objects. This assumption has been removed from the
code.
* 1.11 - November 3, 2004
o Fixed a couple bugs in the dump_data_fields form field reporting.
Also improved this feature so it now reports all possible settings
for check box, radio button, list box and combo box form fields.
* 1.10 - October 27, 2004
o Fixed the background feature so it handles rotated pages (input PDF
or background PDF) better. Pdftk will transform the background PDF
page so that its orientation is preserved on every page of the
output PDF, even on input PDF pages that are rotated. I chose this
logic so as to give the user greater control over the results;
rotate pages before processing to achieve the desired output. Let
me know if this logic is too inconveniet for you.
o Fixed form field handling when combining PDF pages. Pdftk used to
permit duplicate form field names, which is illegal PDF. Now, pdftk
detects duplicates and adds name prefixes as needed. If no
duplicates occur, then no changes are made. If an input PDF has a
field represented by multiple annotations, then that is respected
and preserved in the output.
An especially nice upshot to this new handling is that you can now
assemble duplicate PDF forms and not end up with all of their
fields echoing each other (as you get with Acrobat). Run pdftk
A=form.pdf cat A A A output formX3.pdf and you'll get a form that
behaves as you would expect.
o Added stdin support for input PDF, FDF, or Info files (thanks to
Bart Orbons for this patch).
o Added a means for users to control the output PDF filenames when
using the burst feature: pass in a printf-styled format string via
output (documented above).
o Changed background command-line syntax, so it is an operation
instead of an output option. The old syntax also works, for
backward compatibility.
o Now shuffling subset font name prefixes for input PDFs, to prevent
duplicates.
o Updated Makefile.Mandrake according to feedback from Larry
Gilliland.
o Reduced the Windows EXE filesize using UPX, as suggested by Ralf
Koenig.
* 1.00 - August 14, 2004
o Upgraded the iText library we use to itext-paulo rev. 132, which
resolved a bug involving bookmark page references in dump_data
output.
o Fixed the problem of form fields getting corrupted by splitting or
merging PDF form pages.
o Building the Windows binary using libgcj 3.4 seems to have fixed
the problem of using accented characters in filenames and paths.
o Added these new operations: fill_form, update_info, attach_file,
and unpack_file.
o Added the background and flatten output options.
o Added the do_ask interactive mode (the default on Windows) that
asks before overwriting files and asks for passwords to input PDFs,
if necessary. Also added the dont_ask mode (the default on Linux),
for hands-free operation.
o Many input fields can be substituted with PROMPT, which cues pdftk
to ask the user for a filename or password upon execution.
o Added output to stdout via output -.
o Using the uncompress option now also adds page numbers to page
dictionaries, for easy lookup. Find page N (1-based) by searching
for /pdftk_PageNum N. Using the compress option removes these
markers.
o Added Mac OS X Makefile, and removed the optimization flag from the
GCJ flags (which would cause trouble on older versions of gcc, such
as 3.2.2) in all Makefiles.
o Now catching PDF output open exceptions.
o Builds now pack iText font afm files into pdftk, which are required
for the new form filling feature.
* 0.941 - March 28, 2004
o Fixed the 'Input_UnicodeBig not found' error encountered by Windows
users when using the dump_data or the burst operations on some
PDFs.
o Added an optimization flag to the gcj arguments. This can be
adjusted or omitted by editing your platform-specific Makefile.
o Renamed the CC_OPTS Makefile macro to CXXFLAGS, for uniformity.
* 0.94 - March 24, 2004
o Fixed a string copy bug in pdftk.cc.
o Fixed unicode string output so it drops initial, signature
character.
o Fixed nagging gnu.java.locale.Calendar static linking problem
(Windows).
o Made more improvements towards gcc/gcj 3.2 compatibility (e.g.,
RedHat 8, 9).
o Added macros to Makefiles, for easier porting.
o Added simple return codes: 0 --> Success, 1 --> Error, 2 --
> Exception. Some exceptions will return an "Error" return code.
o Removed warning issued when an input PDF has no ID string.
o Empty Info fields no longer reported on dump_data.
o Added newline to end of --help output.
* 0.93 - March 7, 2004
o Removed restriction on the number of input documents. For example,
you can now run:
pdftk *.pdf cat output combined.pdf
to assemble any number of PDFs into a single document.
o Made pdftk run silently by default, and added the verbose output
option for when you want detailed feedback.
o Changed the encryption strength default from 40-bit to 128-bit.
o Improved file open error handling and reporting. If pdftk can't
open a PDF, it tells you why.
o Added RedHat9 and Mandrake makefiles (Thanks to Andre Gompel and
Pablo Rodri'guez). Support for these platforms is still
experimental.
o Copied the MD5 code from libgcj into our tree, to improve support
for older compilers/libraries. This should improve RedHat9 and
Mandrake support.
o Removed pointless warning sometimes issued by the libgcj security
class.
o Added debian directory and Aure'lien's man page, updated man page.
o Reorganized Makefiles (thanks Andre).
* 0.92
o Added logical page numbering (a/k/a page labels) to dump_data
operation.
o Appended .omit extension to a few iText files we don't use, to
speed compiling.
* 0.91
o Removed restriction on adding the same page from the same PDF more
than once.
o Added Solaris Makefile.
o AllFeatures permission now implies 'top quality' printing, not
'degraded' printing.
o Fixed handling of 'empty owner password' case during output
encryption.
o Added test to make sure the user password is not the same as the
owner pw.
o CopyContents also allows ScreenReaders.
o ModifyAnnotation also sets the FillIn bit.
o ModifyContents also sets the Assembly bit.
o Updated docs.
o Added --version switch.