| Age | Commit message (Collapse) | Author |
|---|
|
On 2.3.x, opening files from the Android email app (not Gmail), was causing
a crash due to a SecurityException. On 4.x, it was failing with an error
message.
I think we should have just been calling
getContentResolver().openInputStream(uri) to get the file data, rather than
reading the row from the content resolver and looking at the _data field.
On 2.3.x querying this row caused a security error. On 4.x, we got back
a _data field containing a path, but this path was internal to the email
app, and not for our consumption.
Previously we were calling openInputStream() only as a fallback; now we try
it first, and fall back to the other method if that fails. I suspect we can
delete the other code, but I can't test on the 3.x version of the Transformer
Prime, so I'm leaving it be for now.
|
|
Use the UTI 'public.cbz-archive' instead of the more generic
'public.zip-archive'. It's likely another app (or the OS) has an
Exported UTI for 'public.zip-archive' which would take precedence over
our Imported UTI and not have a 'cbz' extension.
It's not entirely clear what the correct UTI to use for CBZs is (there
are a few), but public.cbz-archive seems to be most likely candidate.
Remove the 'zip' extension, as we probably don't want to load arbitrary
zip files.
Add mimetypes for all completeness, and separate OXPS from XPS as it has
a different mimetype.
|
|
The GUI layout tool instantiates custom controls classes to display a
preview in the IDE. It relies on the 2-argument constructor being
implemented.
Use a different means to get the window manager that works for non-activity
contexts, and avoid creating gesture recognizers in this situation.
Based on a patch supplied by Masaki Muranaka
|
|
|
|
|
|
|
|
|
|
|
|
(Revert the fix for potential memory leak - there wasn't a potential memory
leak as the device only takes ownership of it's display list if
dev->free_user is set).
This reverts commit ada560eb015e7b2208c492d47f28da8093733c4a.
|
|
Use intptr_t when casting between a jlong and a pointer to suppress errors
about different size words.
Add a 'u' suffix to unsigned values output by the cmap dump utility.
|
|
|
|
Correct an off-by-one error in the file deletion code.
|
|
Fixes bug #694711. As far as I could see, opening an XPS worked fine, the
first time. However, if the same file was opened a second time, then it
should have been saved with a number, e.g. foo(1).xps. However, the period
was ommitted, so the file wasn't recognised as an XPS.
This presumably wasn't an issue for PDF files, as PDF format is assumed
by default (I am guessing).
|
|
Most were pretty harmless, and were addressed by renaming functions to match the Core
Foundation naming conventions, but there was one actual memory leak, and some potential
uses of uninitialised data.
|
|
Fixes bug #695191 - Mupdf Build49/armv7a & Android 3.1: cycles
through subset of pages & page scrubber
The problem here was that in Honeycomb, various bitmap operations
(including drawing via JNI) do not update the bitmap generation count.
When hardware acceleration is enabled, this means that the underlying GL
layer is not aware that the bitmap has changed, and ends up reusing old
textures.
To workaround this, we erase the bitmap before drawing the page. Erase
appears to be the only operation I could find (after pouring through the
source), which actually increments the generation count. The other option
would have been to disable hardware acceleration, but that was far less ideal.
|
|
env/clazz pointers in globals.
The env/clazz pointers are stashed in the globals structure so that they can
be accessed by callback functions in mupdf.c (such as bufferStreamSeek()).
The intention is that only one thread (i.e. the AsyncTask background thread)
reads/writes these stashed pointers. Because cookies are created/destroyed
in the main thread, we add a new version of get_globals() which doesn't trample
these pointers.
|
|
to file and open from file works
Patch was supplied by gong_su@hotmail.com. Thanks!
|
|
then pressing back button again.
I've also added an onCancel() handler, so that the back button only needs
to be pressed once to return to the file picker view.
Spotted while looking at bug #693719 - Attached PDF file does not display (edit)
|
|
|
|
crash when rotating the device.
When cancelling a render async task, we now wait for it to actually finish
before continuing. The benefit of this is that we should be able to guarantee
that its Bitmap becomes eligible for GC before we continue to create any
new bitmaps.
This should hopefully help with the OOM errors seen when rotating
the device and trying to create the new bitmaps.
To prevent the UI thread from being blocked for too long while we're waiting
for the async task to finish, we use a fz_cookie and set the 'abort' flag to
request the render be stopped as soon as possible.
|
|
|
|
|
|
Android sometimes calls the 'getSelectedView()' method of an AdapterView.
This can be made to happen more predicatably by enabling the Talkback
accessibility feature.
Remove the UnsupportedOperationException and just return null, as we the
ReaderView does not have the concept of a selected page.
|
|
If an iOS app uses too much memory, the OS asks it to free up some space.
If it doesn't do so in a timely manner, it will get a second warning before
being killed by the OS.
In other platforms, where malloc() return NULL in OOM, the store scavenger
releases memory when mallocs fail. In iOS, mallocs usually never return NULL
because the app is killed before this can happen. Therefore, we need to
initiate a scavenge from the low memory notification instead.
We evict the store to 50% of its current size when a memory warning occurs
when it is in the foreground, and 0% when a memory warning occurs whilst
it is in the background. Having said this, I didn't manage to get a background
warning to occur, presumably because we don't request background execution
Therefore, I think in practice the OS just kills the process. However, this
will be useful if we ever add background execution.
|
|
renders a specific PDF
Some async tasks were still trying to access the globals pointer, which had
been nulled during shutdown, leading to a crash.
Check for NULL and return cleanly in this case.
Possibly we should also be calling AsyncTask.get() to wait for the tasks to
finish, before nulling the global pointer?
|
|
|
|
Android NDK
(security issue because a variable is used as a format string with no parameters).
|
|
|
|
|
|
When throwing an error during fz_alpha_from_gray, the stack depth
can get confused. Fix this by moving some more code into the
appropriate fz_try().
In the course of fixing this bug, I added some new optional debug
code to display the stack level as it runs. This is committed here
disabled; just change the appropriate #define in draw-device.c to
enable it.
Also, add some code to run_xobject, to avoid throwing in an fz_always()
clause.
|
|
Return the null object rather than throwing an exception when parsing
indirect object references with negative object numbers.
Do range check for object numbers (1 .. length) when object numbers
are used instead.
Object number 0 is not a valid object number. It must always be 'free'.
|
|
Replace the DroidSansFallback TTF files with a TTC that has two fonts:
The original and a copy where the OpenType 'vert' substitution
lookup has been pre-applied by copying the uniXXXX.vert glyph data
to uniXXXX.
|
|
pdf_create_document leaks the trailer and in pdf-device.c many objects
are inserted into dictionaries using pdf_dict_puts and leaked instead
of using pdf_dict_puts_drop.
|
|
...like the one Microsoft Word generates.
|
|
If we hit an out of memory error in fz_draw_end_mask, then pop the
stack, and rethrow. Ensure that the generic device code catches this
error and sets the error_depth to 1 so that the final pop is ignored.
|
|
|
|
|
|
|
|
|
|
|
|
Add a new function pdfapp_autozoom to fit the page to the window by comparing
the aspect ratios of the page and the window to choose whether to fit
horizontally or vertically.
|
|
Various functions (such as fz_begin_group) handle errors internally
by use of the error_depth parameter. This means that if we call
them, we MUST ensure that we call the appropriate closing function.
Similarly, if we don't call them, we should NOT call the closing
function.
In order to ensure we do this correctly, we introduce a cleanup_state
variable that says which ones we tried to call.
This cures the original bug.
|
|
We used to extract the outline using the combined TM*CTM matrix and
use the identity transform for stroking, thus ending up with the wrong
line width.
If we instead extract using the TM and then stroke with the CTM we get
the correct results.
|
|
|
|
Allows compilation with -DFT_DEBUG_LEVEL_TRACE to track down buggy fonts.
|
|
When I changed the stream implementations to use implementation
specific buffers, rather than a generic public one in every fz_stream,
I changed fz_read_byte to only get a single byte at a time.
I noted at the time that the underlying stream was free to decode
larger blocks if it wanted too, but I forgot to actually do this for
the flate decoder. Fixing this here should solve the speed issues.
|
|
Without this, comparefiles/Bug695086 renders the barcode test upside
down.
|
|
Fix broken solution file and add project entries for new files.
|
|
|
|
Grow the edge list using an exponential realloc pattern.
Use qsort for huge paths and only fall back to the simple
shell sort for small paths.
|
