to better match other perl5 module packages. This release mirrors the
Xerces-C 1.7.0 releases and implements a DOM API (though that seems to
be broken). Also, several major memory leaks were fixed.
features and new and modified APIs from version 2.4:
* Added support for Unicode 4.0
* Added support for Unicode regular expressions
* Enhanced sorting
* Added support for international domain names
* Added service registration for pluggable ICU modules
* Added layout engine API for language-specific glyphs
* Separated currencies from locales
* Added POSIX-like API for message catalogs
* Added new charset converters
This package extracts DVD subtitles from a subtitle stream and
converts it to pgm or ppm images or into VobSub format. The main
purpose is to provide the required input to OCR software to convert
the subtitle images into ASCII text. Please note that the conversion
into ASCII is not part of this package but requires an OCR program
like gocr.
GOCR is an optical character recognition program, released under the GNU
General Public License. It reads images in many formats (pnm, pbm, pgm, ppm,
some pcx and tga image files (or PNM from stdin); if PNM-tools installed and
running linux-like system you can also use pnm.gz, pnm.bz2, png, jpg, tiff,
gif, bmp and others) and outputs a text file. You do not have to train the
program or store large font bases. Simply call gocr from the command line
and get your results.
changes since 3.02:
4.0.7
Changes: This version fixes some rare infinite loops and wrong results
that occurred when the s command had a numeric option and the regular
expression could match the empty string.
4.0.6
Changes: This release improves portability a lot, and adds a
configure-time switch to choose between the included regex matcher and
the system matcher (on glibc systems only).
4.0.5
Changes: This release fixes a few portability bugs and improves the
quality of the documentation. "a", "i", "l", "L", and "=" now accept
address ranges.
4.0.3
Changes: A packaging problem with two missing translation catalogs was
fixed.
4.0.2
Changes: This release makes sed able to bootstrap itself on operating
systems that cannot run the configure script.
4.0.1
Changes: This release includes translations for over 20 languages, and
cleans up the binary and documentation files from references to
super-sed.
4.0
Changes: This release has been merged with super-sed and supports all
of its features, except for Perl regular expressions. This includes
in-place editing, new sed commands, better documentation, and support
for changing the case of the characters with the `s' command. The
performance and correctness of the regular expression matcher have
also been improved.
Changes in version 1.7:
* scanner for Ruby programming language (thanks to Noritsugu Nakamura
<nnakamur@mxq.mesh.ne.jp>).
* fixed explicit rules in Makefiles that generated errors with some
versions of non-GNU makes.
- Handle no translations in get_newest_po_age (Rodney Dawes)
- Cleaned up tests and fixed few bugs (Kenneth)
- Send error messages to stderr, and fix other similar bugs (bug #105013,
Denis Barbier)
- handle using a different output file. (James Henstridge)
- Correct the invocation of INTLTOOL_MERGE. (bug #106898, Malcolm Tredinnick)
- Extract attributes in XML when there are spaces arount the equal sign
(Alberto Manuel Brand?o Sim?es)
- use iconv -f, instead of iconv --f (bug #105181, Hidetoshi Tajima)
- Added variable substitution. (Kenneth)
- Added small section for packages not using the aufo* tools (Kenneth)
- Call msgmerge with options before non-options to fix potential non-GNU
getopt issues. (Gregory Stark, Jordi Mallach)
- Make it work when running an UTF-8 locale (Kenneth)
- Fix some iconv/charset issues (bug #113121, Morten Welinder)
libstdc++ in gcc3.
when defined, USE_GCC_SHLIB ensures that the correct rpath is passed
to the linker, and a full dependency on the compiler package is
registered.
packages which define USE_GCC_SHLIB should not include
mk/gcc.buildlink2.mk (or gcc{,3}/buildlink2.mk) as it is handled
automatically.
Lots of bug fixes and improvements.
o Initial support for timestamp PI. From now you can use
<?dbtimestamp format="Y-m-d H:M:S"?> to get current datetime
in your document. Added localization support for datetime
PI
o Added level 6 to test for section depth in section.level
template so that section.title.level6.properties will be
used for sections that are 6 deep or deeper. This should
also cause a h6 to be created in html output.
o Don't use SVG graphics if use.svg=0
o Now uses number-and-title-template for sections only if
section.autolabel is not zero.
o Added missing 'english-language-name' attribute to the l10n
element, and the missing 'style' attribute to the
template element so the current gentext documents will
validate.
o Corrected several references to parameter qanda.defaultlabel
that were missing the "$".
o Now accepts admon.textlabel parameter to turn off Note,
Warning, etc. label.
o FeatReq #684561: support more XEP metadata
o Added hyphenation support. Added support for coref. Added
beginpage support. (does nothing; see TDG).
o Added support for hyphenation-character,
hyphenation-push-character-count, and
hyphenation-remain-character-count
o Added root.properties, ebnf.assignment, and ebnf.statement.terminator
o Support bgcolor PI in table cells; make sure rowsep and
colsep don't have any effect on the last row or column
o Handle othercredit on titlepage a little better
o Applied fix from Jeff Beal that fixed the bug that put
secondary page numbers on primary entries. Same with tertiary
page numbers on secondary entries.
o Added definition of missing variable collection.
o Make footnote formatting 'normal' even when it occurs in a
context that has special formatting
o Added warning when glossary.collection is not blank, but it
cannot open the specified file.
o Pick up the frame attribute on table and informaltable.
o indexdiv/title in non-autogenerated indexes are now picked up.
o Removed (unused) component.title.properties
o Move IDs from page-sequences down to titlepage blocks
o Use proportional-column-width(1) on more tables.
Use proportional-column-width() for header/footer tables;
suppress relative-align when when using FOP
o Check for glossterm.auto.link when linking firstterms; don't
output gl. prefix on glossterm links
o Generate Part ToCs
o Support glossary, bibliography, and index in component ToCs.
o Refactored chunking code so that customization of chunk
algorithm and chunk elements is more practical
o Support textobject/phrase on inlinemediaobject.
o Support 'start' PI on ordered lists
o Fixed test of $toc PI to turn on qandaset TOC.
o Added process.chunk.footnotes to sect2 through 5 to fix bug
of missing footnotes when chunk level greater than 1.
o Added paramater toc.max.depth which controls maximal depth
of ToC as requested by PHP-DOC group.
- honor PKGLOCALEDIR in PLIST (thanks pkglint(1))
- incorporate some filter processing from namazu's stable branch.
filter/pdf.pl:Adapted to xpdf 2.02
filter/mhonarc.pl: Add codeconv::toeuc() after uncommentize().
- added test to catch old versions of Storable which lack locking support
- removed new-style loop which broke on 5.005_03
- suppress more uninitialised variable warnings
- fixed strict mode requiring ForceArray on output
- fixed warnings about uninitialised values
- fixed warnings when NormaliseSpace undefined
- added support for specifying ForceArray using regular expressions
- added check to escape_value to guard against undefined argument
- added NoIndent option
- integrated a patch
+ new facility for removing extra levels of indirection (using
the new 'GroupTags' option)
+ new facility for rolling the dreaded 'content' hash up into a
scalar if there are no keys left after array folding (using the
'-' prefix mode on the ContentKey option)
+ new facility for doing variable substitution in the XML; variables
can be defined in Perl (using the new 'Variables' option) or in
the XML document (using the new 'VarAttr' option)
- added 'NormaliseSpace' option for tidying up hash keys and other
text content if required
- option names are now case-insensitive and can include underscores
- XMLin() and XMLout() are now aliased to xml_in() and xml_out() when
called as methods or imported explicitly
- option names passed to XML::Simple->new() are now validated
1.0.30: May 4 2003:
- Fixes and new APIs to handle Result Value Trees and avoid leaks
- Fixes for: EXSLT math pow() function (Charles Bozeman), global
parameter and global variables mismatch, a segfault on pattern
compilation errors, namespace copy in xsl:copy-of, python generator
problem, OpenVMS trio update, premature call to xsltFreeStackElem (Igor),
current node when templates applies to attributes
- Delay calling tput until delimiters are needed, and don't call if
user specifies his/her own delimiters first
- Don't run tput if TERM is set to empty string
- Fixed bug where XML::DOM doesn't encode some characters in
attribute values under Perl 5.8.0
- Added t/dom_encode.t test to check encoding on attribute values
- Fixed warning message and use in XML::DOM::PerlSAX
- included XML-Parser-2.31.patch, required for XML::Parser
to work with 5.8.0 unicode
- use utf8 in unicode test scripts, fixes 5.8.0 test failures
NOTE - you should use the utf8 pragma in your programs if you
are passing utf8 to XML::DOM.
- only use encodeText for Perl < 5.6.0
- replace match w/ substitution in AttDef::new, workaround for 5.8.0 unicode
- replace match w/ substitution in Default handler for non-paramter
entity reference, workaround for 5.8.0 unicode
- Fixed problem when defining user LWP object
- Autodetect whether to 'use bytes'
- Added dispose method to XML::DOM::Parser Synopsis
- Fixed warning message in Attr.getValue method
From the readme:
This program is a simple filter that reads the output of the 'diff' program,
and produces a histogram of the total number of lines that were changed. It is
useful for scanning a patch file to see which files were changed.
- Relax-NG: Compiling to regexp and streaming validation on top of the
xmlReader interface, added to xmllint --stream
- xmlReader: Expand(), Next() and DOM access glue, bug fixes
- Support for large files: RGN validated a 4.5GB instance
- Thread support is now configured in by default
- Fixes: update of the Trio code (Bjorn), WXS Date and Duration fixes
(Charles Bozeman), DTD and namespaces (Brent Hendricks), HTML push parser
and zero bytes handling, some missing Windows file path conversions,
behaviour of the parser and validator in the presence of "out of memory"
error conditions
- extended the API to be able to plug a garbage collecting memory
allocator, added xmlMallocAtomic() and modified the allocations
accordingly.
- Performances: removed excessive malloc() calls, speedup of the push and
xmlReader interfaces, removed excessive thread locking
- Documentation: man page (John Fleck), xmlReader documentation
- Python: adding binding for xmlCatalogAddLocal (Brent M Hendricks)
Troff
-----
o Input encoding files for latin-9 (a.k.a. latin-0 and ISO 8859-15) and
latin-2 (ISO 8859-2) have been added. Example use:
groff -Tdvi -mlatin9 my_file > my_file.dvi
You still need proper fonts with the necessary glyphs. Out of the box,
the groff package supports latin-9 only for -Tps, -Tdvi, and -Tutf8,
and latin-2 only for -Tdvi and -Tutf8.
o Composite glyphs are now supported. To do this, a subset of the Adobe
Glyph List (AGL) Algorithm as described in
http://partners.adobe.com/asn/developer/typeforum/unicodegn.html
is used to construct glyph names based on Unicode character codes. The
existing groff glyph names are frozen; no glyph names which can't be
constructed algorithmically will be added in the future.
The \[...] escape sequence has been extended to specify multiple glyph
components. Example:
\[A ho]
this accesses a glyph with the name `u0041_0328'.
Some groff glyphs which are useful as composites map to `wrong' Unicode
code points. For example, `ho' maps to U+02DB which is a spacing
ogonek, whereas a non-spacing ogonek U+0328 is needed for composite
glyphs. The new request
.composite from to
will change the mapping while a composite glyph name is constructed.
To make \[A ho] yield the expected result,
.composite ho u0328
is needed. [The new file `composite.tmac' loaded at start-up already
contains proper calls to `.composite'.]
Please refer to the info pages of groff and to the groff_char man page
for more details.
o A new request `fschar' has been added to define font-specific fallback
characters. They are searched after the list of fonts declared with the
`fspecial' request but before the list of fonts declared with `special'.
o Fallback characters defined with `fschar' can be removed with the
new `rfschar' request.
o A new request `schar' has been added to define global fallback
characters. They are searched after the list of fonts declared with the
`special' request but before the already mounted special fonts.
o In groff versions 1.18 and 1.18.1, \D'f ...' didn't move the current
point horizontally. Despite of being silly, this change has been
reverted for backwards compatibility. Consequently, the intermediate
output command `Df' also moves the position horizontally again.
\D'f ...' is deprecated since it depends on the horizontal resolution of
the output device (given with the `hor' parameter in the DESC file).
Use the new \D'Fg ...' escape instead.
o For orthogonality, new \D subcommands to change the fill color are
available:
\D'Fr ...' (rgb)
\D'Fc ...' (cmy)
\D'Fg ...' (gray)
\D'Fk ...' (cmyk)
\D'Fd' (default color)
The arguments are the same as with the `defcolor' request. The current
position is *not* changed.
o The values set with \H and \S are now available in number registers
\n[.height] and \n[.slant], respectively.
o The `.pe' number register isn't new but hasn't been documented before.
It is set to 1 during a page ejection caused by the `bp' request.
o The new glyph symbol `tno' is a textual variant of `no'.
o The new glyph symbol `+e' represents U+03F5, GREEK LUNATE EPSILON SYMBOL.
(Well, it is not really new since it has been previously supported by
grolj4.) The mapping for both the dvi and lj4 symbol font has been
changed accordingly so that Greek small letter epsilon, `*e', has the
same glyph shape as with other devices.
Grops
-----
o The font `freeeuro.pfa' has been added to provide various default glyph
shapes for `eu' and `Eu'.
o It is now possible to access all glyphs in a Type 1 font, not only 256
(provided the font file created by afmtodit has proper entries). grops
constructs additional encoding vectors on the fly if necessary.
o The paper size is now emitted via the %%DocumentMedia and PageSize
mechanisms so that it is no longer required to tell `gv' or `ps2pdf'
about the paper size. The `broken' flag value 16 omits this feature
(the used PostScript command `setpagedevice' is a LanguageLevel 2
extension). Patch by Egil Kvaleberg <egil@kvaleberg.no>.
o Non-slanted PostScript metrics have been changed again; they no longer
contain negative left italic correction values. This assures correct
spacing with eqn.
Grodvi
------
o The font cmtex10 has been added as the special font `SC' to the DVI fonts.
It is used as a font-specific special font for CW and CWI.
o New options -l and -p to set landscape orientation and the paper size.
grodvi now emits a `papersize' special which is understood by DVI drivers
like dvips.
Consequently, the DESC file should contain a `papersize' keyword.
o The glyph shapes for \[*f] and \[*e] have been exchanged with \[+f] and
\[+e], respectively, to be in sync with all other devices.
o Glyphs \[HE] and \[DI] have been replaced with \[u2662] and \[u2661],
respectively, since the former two glyphs have a black (filled) shape
which grodvi doesn't provide by default (it never has actually).
Grolj4
------
o The glyphs \[*e] and \[+e] have been exchanged to be in sync with
all other devices.
o The glyph \[~=] is now called \[|=]. Similar to other devices, \[~=]
is now another name for glyph \[~~].
Grotty
------
o New option `-r'. It is similar to the -i option except it tells grotty to
use the `reverse video' attribute to render italic fonts.
Pic
---
o New command `figname' to set the name of a picture's output box in TeX
mode.
Refer
-----
o The environment variable `REFER' to override the name of the default
database isn't new but hasn't been documented before.
Soelim
------
o New option `-r' to avoid emission of `.lf' lines.
o New option `-t' to emit TeX comment lines (giving current file and the
line number) instead of `.lf' lines.
Afmtodit
--------
o Unencoded glyphs in an AFM file are output also (since grops can now emit
multiple encoding vectors for a single font).
o New option `-m' to prevent negative left italic correction values.
o The mapping and encoding file together with file `DESC' are now searched
in the default font directory also. Please refer to the man page of
afmtodit for more details.
Macro Packages
--------------
o Larry Kollar <kollar@alltel.net> and others made the man macros more
customizable.
. New command line options -rFT, -rIN, and -rSN to set the vertical
location of the footer line, the body text indentation, and the
sub-subheading indentation.
. New command line option -rHY (similar to the ms macros) to control
hyphenation.
. New macros `.PT' and `.BT' to print the header and footer strings.
They can be replaced with a customized version in `man.local'.
. The string `HF' now holds the typeface to print headings and
subheadings.
. Similar to the ms macros, the LT register now defaults to LL if not
explicitly specified on the command line.
o troff's start-up file `troffrc' now includes `papersize.tmac' to set
the paper size with the command line option `-dpaper=<size>'.
Possible values for `<size>' are the same as the predefined `papersize'
values in the DESC file (only lowercase; see the groff_font man page)
except a7-d7. An appended `l' (ell) character denotes landscape
orientation. Examples: `a4', `c3l', `letterl'.
Most output drivers need additional command line switches `-p' and
`-l' to override the default paper length and orientation as set in
the driver specific DESC file.
For example, use the following for PS output on A4 paper in landscape
orientation:
groff -Tps -dpaper=a4l -P-pa4 -P-l -ms foo.ms > foo.ps
