qpdfview team mailing list archive
-
qpdfview team
-
Mailing list archive
-
Message #00010
Re: Request for testing after extensive refactoring
-
To:
qpdfview@xxxxxxxxxxxxxxxxxxx
-
From:
bdjnk <bdjnks@xxxxxxxxx>
-
Date:
Thu, 22 Mar 2012 15:10:56 -0700
-
In-reply-to:
<op.wbk498t0v1wjws@localhost.localdomain>
-
User-agent:
Mozilla/5.0 (X11; Linux i686; rv:11.0) Gecko/20120314 Thunderbird/11.0
Brilliant work! That was some fast effective refactoring. So aside from
pure amazement at how awesome everything worked, here are the issues I
ran into:
Entering a new search term and pressing next results in the previous
search term being sought. This may be intentional, but if doesn't feel
like rule of least surprise to me.
There is no way (that I've yet found) to un-highlight the search
results. Whatever is highlighted remains highlighted until a new search
is initiated.
There are also some points where the previous search term becomes lost
if it is deleted from the search box. For example, when a search is
made, then the search term is erased, then the document is rotated,
zoomed, or the view type is switched. In that case the search will be
nullified, while not erasing the search term will result in the search
remaining active.
There may be more, but that is as much as I have time to test right now.
On 03/22/2012 11:50 AM, Adam Reichold wrote:
Hello.
Because of the problems encountered when implementing the requested
additions to the search feature, I started rather extensive
refactoring to improve the structure of the code base. With regards to
functionality this should be the same as trunk revision 81 but the
code changed considerably and is hopefully much easier to understand.
(Well the problems with search results being discarded after a change
of display mode should be gone.)
If this did not introduce serious regressions, I will start improving
the source code documentation. But to find out, this needs to be
tested. Because the reshuffling of code is so extensive, I would ask
you to test all the functionality again if you have the time. It is
probably not necessary to open bug reports for every regression
introduced, just post to this mailing list thread. Thank you for your
help.
Best regards, Adam.
P.S.: One nice side-effect of this is that I also have a litte crazy
branch where poppler is replaced by fitz, the PDF rendering library
used by mupdf which seems considerably more lightweight and has proper
support for multi-threading. (And seems to have even worse
documentation that poppler. :-()
Follow ups
References