Change History
- now recognizes oriental type thousand seperators like in
the lists done by and for Sabaithai.
- added special case recognition of spaces in filenames. The
special case is that the filename must have '.JPG' in it as
well as be the first item. I.e. for a then valid line it will
interpret everything from the first non-whitespace character
upto the first whitespace following the '.JPG' (case
independent) as the filename. This has been done to bring
textual file recognition more in line with other formats. Even
though it does have an inherent danger built in.
- 4th of July anniversary release with fireworks. Or at
least a close approximation thereof. It was an outstanding
request I had all but forgotten.
- changed the generated 'Get it at:' list header to point to a
bounce page so the actual site can be changed if needed
- changed recognizer so that it will first search for '.jpg'
and only when not found will search for first dot
- changed HTML generator to output <TD> instead of <TH> for
table data rows. Still <TH> for table headers.
- changed box text from miss: to want: to better indicate
what the count represents
- numeric recognition was too liberal
- When showing the verification list the option to limit the
list to the missing files has been changed to limit the list to
the wanted files. I.e. list those files from the verification
list that are not present or are not according to the verification
list in size or CRC.
- A popup menu for the files grid that has some options
for color handling as well as updating verification lists
- An extra menu to rename files to comply with verification list
- The directory reader thread displays its results as well
as a simple size-only verification. The JFIF parser thread
continues to read the files.
While the menus remain disabled while the JFIF parser thread is
running (and others like the sorting thread and such), it does
provide a quick (and dirty) feedback of the state of the collection
while not getting in the way of processing. Provided the
verification list has been read beforehand!
The reason for this limitation is shying away from complexity.
- ability to detect duplicates by name or by size
- several sorting options
- contrary to before not all new things have been tested.
I don't really like to release it publicly without at least
some measure of quality control, but. Since one way or the
other the QA releases have a tendency to show up all over the
place anyway, I decided I might as well.
However, since this is the first such non-QAd release it is
just item 3 that is relatively untested.