| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
A lot of header clean-up.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- export of .odg files was completely broken on Windows as
"documentUri = Inkscape::URI(filename);" always failed
(Windows paths are never valid URIs because of '\')
- Linked images did not seem to be working on any OS
(probably regression in 6f844ef457690c841b0be91d70b1e54b61c04812)
I assume support for linked images is still sketchy but seems to
work in the most common cases; support for embedded images seems
to be non-existent, though.
Fixed bug:
- https://bugs.launchpad.net/inkscape/+bug/1654034
|
| |
|
|
|
|
|
|
|
| |
Replace "\stackrel" with "\tabular" to achieve constant line spacing.
The former varies line spacing depending on the content of the line
(i.e. an line containing "-" will have a very small height)
Add basic support for line-height.
(currently only the top level text element is considered)
|
| |
|
| |
Found via `codespell`
|
| |
|
|
|
| |
Fixed bug:
- https://bugs.launchpad.net/inkscape/+bug/1739497
|
| |
|
|
|
|
| |
As the parser seems to wrap newlines into their own tspan which has
the same style as the preceding tspan the latex-text renderer
produced output like "\texbf{\\}" which broke compilation in LaTeX
|
| |
|
|
|
|
|
|
|
|
|
| |
Wrap text in "\shortstack" to allow for newlines using "\\".
The output extension already produced those but they are ignored
in a "\smash" environment.
Patch by Michael Jäntsch
Fixed bug:
- https://bugs.launchpad.net/inkscape/+bug/771959
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This also fixes an issue with multipage PDF output where an empty
string would have been assumed to require emission of the previous
page when in fact the CairoRenderer does not produce a new page in
this case resulting in another facet of
https://bugs.launchpad.net/ubuntu/+source/inkscape/+bug/1417470
Invisible characters might still cause this issue but it seems to be
a very unlikely scenario and there's no way to know this without
actually rendering the glyphs (which seems to be overkill here)
|
| |
|
|
|
|
| |
It happen often that the rendering was interrupted early but the
final "cairo_surface_status" claims success which results in
incomplete output but no user-visible warning.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The previous implementation [1] allowed to sandwich text between
graphical objects by outputting multiple (partial) PDF pages which
are then stacked in the final document to reveal the full image.
However this code failed for clipped/masked/transparent objects
as those are treated specially by the renderer resulting in missing
pages in the output causing [2].
The attempt to workaround this issue which was committed in
e4dea66a338824037b6c35b262aa8db4004b6581 (now reverted)
fixed document creation in LaTeX by inserting blank pages but did
not actually fix the issue with clipped/masked/transparent objects
typically resulting in a single page with the full image and all
text put on top.
This commit resolves the underlying issue, making the former
workaround unnecessary and allowing for proper overlaying of text
and arbitrarily clipped/masked/transparent objects
Fixed bugs:
- https://bugs.launchpad.net/inkscape/+bug/771957 [1]
- https://bugs.launchpad.net/inkscape/+bug/1417470 [2]
|
| |
|
|
| |
This reverts commit e4dea66a338824037b6c35b262aa8db4004b6581.
|
| |
|
|
|
|
|
|
|
|
| |
- do not call cairo_show_page as the GtkPrintOperation seems to do
that internally (calling it manually will emit the current page,
clear it, and the cleared page seems to be emitted in error later)
- we can (and should) still call cairo_surface_destroy as it's doing
reference counting internally (we increase the reference count in
CairoRenderContext::setSurfaceTarget) and won't actually destroy
anything until the count reaches zero
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Previously only the first page was printed properly,
all following pages were either blank or were not printed at all.
This was caused by finishing/destroying the cairo surface of the
cairo context associated with the GtkPrintContext used for printing
which made it impossible to draw to it for any consecutive page
Fixed bugs:
- https://bugs.launchpad.net/inkscape/+bug/1733424
- https://bugs.launchpad.net/inkscape/+bug/490866
|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Found using `codespell -q 3 -w --skip="*.svg,*.po,*.ts,./share/tutorials,./src/libavoid,./packaging/win32/languages,./man,./src/2geom" -I ../inkscape-whitelist.txt`
whereby whitelist file contained:
```
dum
iff
glight
substract
te
upto
```
|
| |/ |
|
| |\ |
|
| | |
| |
| |
| |
| | |
This replaces all usage of boost::scoped_ptr with std::unique_ptr.
Also removes the corresponding includes.
|
| |/ |
|
| |
|
|
| |
Just use Gio::File when access to non-local files is needed.
|
| | |
|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The crash only affected embedded images.
(we were already catching the exception when loading a linked file)
While at it add some error reporting via debug messages,
fix a memory leak and do some minor cleanup.
Fixed bug:
- https://bugs.launchpad.net/inkscape/+bug/1720330
|
| |/ |
|
| | |
|
| |
|
|
|
|
| |
The idea behind inkscape-version.h / inkscape-version.cpp is to
avoid rebuilding most of the code base when the version number
changes. Defining a version number in config.h defeats this purpose.
|
| | |
|
| | |
|
| |
|
|
| |
get_filename methods
|
| |
|
|
|
| |
Fixes Bug:
- https://launchpad.net/bugs/1417470
|
| |
|
| |
(bzr r15724)
|
| |
|
|
|
| |
longer tied to Adobe).
(bzr r15723)
|
| |
|
| |
(bzr r15715)
|
| |
|
|
|
| |
placeholder instead
(bzr r15653)
|
| |
|
| |
(bzr r15633.1.2)
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This unconditionally sets "LC_NUMERIC" to "C" to use dots (.) in all scanf/printf-like functions to prevent parsing issues with locales that would use a comma (,) instead.
Fixed bugs:
- https://launchpad.net/bugs/1549015
- https://launchpad.net/bugs/1675755
- https://launchpad.net/bugs/1335660
- https://launchpad.net/bugs/1089857
- https://launchpad.net/bugs/1599763
(bzr r15617)
|
| |
|
|
|
|
| |
Remove rotation via viewbox.
Still some work to do...
(bzr r15603)
|
| |
|
| |
(bzr r15586)
|
| |
|
| |
(bzr r15585)
|
| |
|
|
|
|
|
|
|
|
| |
polygon series"
The problem where calls of "sprintf()" to format SVG path data. Unfortunately depending on the locale set at this point via "setlocale()" the output could contain commas (,) instead of dots (.) as decimal spearator resulting in invalid path data.
Fixed bugs:
- https://launchpad.net/bugs/1549015
(bzr r15568)
|
| |
|
|
|
|
| |
Fixed bugs:
- https://launchpad.net/bugs/1664372
(bzr r15532)
|
| |
|
| |
(bzr r15526)
|
| |
|
| |
(bzr r15525)
|
| |
|
| |
(bzr r15505)
|
| |
|
|
|
|
|
| |
Fixed bugs:
- https://launchpad.net/bugs/1662686
- https://launchpad.net/bugs/1662683
(bzr r15493)
|
| |
|
|
|
|
| |
Fixed bugs:
- https://launchpad.net/bugs/1537497
(bzr r15446)
|
| |\
| |
| | |
(bzr r15142.1.38)
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Windows if the filename contains non-ASCII characters
The problem is that "RVNGFileStream" uses "fopen()" internally which only supports ANSI filenames on Windows (i.e. a complete fix would require upstream code changes).
By using "g_win32_locale_filename_from_utf8()" the problem can be worked around in most cases, though:
* the filename is converted to the current codepage (i.e. all 255 characters that are available in the current character encoding are allowed in the filename)
* even if the filename contains a character that's not available in the current character encoding it's attempted to use the alternative short (8.3) file name instead
Therefore the input operation will only fail in the unlikely case that the filename contains a character not available in the current ANSI code page while at the same time short file names are disabled on the file system (which is not the case in standard configurations).
Fixed bugs:
- https://launchpad.net/bugs/1656763
- https://launchpad.net/bugs/1656763
(bzr r15421)
|
| | |
| |
| | |
(bzr r15364)
|
| | |
| |
| | |
(bzr r15363)
|