mikrosimage / sequenceparser Goto Github PK

The file describes a build system using Boost.Build, whereas the real build system uses sconsProject.
Maybe we can change to CMake ?
Or keep things like that, update INSTALL file, and create a submodule for sconsProject ?

in the examples folder:

CMakeLists.txt

java/lsFile.java
java/lsFolder.java
java/lsSequence.java

cpp/lsFile.cpp
cpp/lsFolder.cpp
cpp/lsSequence.cpp

python/lsFile.py
python/lsFolder.py
python/lsSequence.py

See the same kind of issue for avTranscoder:
mikrosimage/avTranscoder#258

It's currently supported in the file/sequence name, but not in the directory.

/path/to/*/all/directories
/path/to/*/all/directories/seq.####.jpg

It cool be better if this library can give detailed information about file:

• creation date
• last modification date
• UID
• GID
• Size (or estimated file size for sequences)

Need to up CMake minimum required to v2.8.11... But this breaks the Travis build: they tell they will fix it.

https://www.ibm.com/support/knowledgecenter/ssw_ibm_i_71/apis/lstat.htm
http://man7.org/linux/man-pages/man2/stat.2.html

If the method returns -1, lstat() was not successful. The errno global variable is set to indicate the error.

Instead of having it with the build system: https://github.com/mikrosimage/sequenceparser/blob/develop/CMakeLists.txt#L6

Because it could be great to get the version in bindings.

The case:

sam ls *.jpg

Returns:

foo_####.jpg 
foo_####.jpg 
foo_####.jpg 
foo_####.jpg 
foo_####.jpg 
foo_####.jpg 
foo_####.jpg 
foo_####.jpg 
foo_####.jpg 

Should it be nice to convert a sequence to Items ?

And a shortcut to get the list of missing frames / files directly.

sequenceParser detects 2 sequences like this:

sequence.####.jpg      [100:138] 39 files
sequence.###.jpg        [50:99] 50 files

Whereas rvls detects one sequence:

sequence.50-138@@@.jpg

Can we have an EDetection method to choose 'eDetectionSequenceCanHaveVariablePadding'?

pyTest.testBrowse.testBrowse ... python: /home/travis/build/mikrosimage/sequenceparser/src/sequenceParser/FrameRange.hpp:38: sequenceParser::Time sequenceParser::FrameRange::getNbFrames() const: Assertion `step >= 1' failed.

The path could be a file/folder/link/sequence.

Hey.

it looks like there is a bad cast or overflow when parsing some files :
I am seeing this :

$> ./lss /path/to/testimage
Listing 1 items took 0 milliseconds
UNITFILE "/path/to/testimage/0891634363_3b0ebecbb3_o.jpg"

when parsing a file named 5186601659_3b0ebecbb3_o.jpg

$> ls /path/to/testimage
5186601659_3b0ebecbb3_o.jpg

Item::explode:
if it's not a sequence, add current item to the vector returned.

Considering the following folder:

Pano_1-1.jpg
Pano_1-2.jpg
Pano_2-1.jpg
Pano_2-2.jpg

This call to browse:

browse(dir=folder, detectOptions=eDetectionDefault, filters=[Pano_1-@.jpg])

Will output:

[email protected] 
[email protected]

Any hint?

Seems to be:

lstat("/path/to/folder/"); // stat of folder
lstat("/path/to/folder"); // stat of link

Check the python major version.

Could be great to add a way to iter through existing files of a sequence (in C++ and Python). Sequence object could become a standard iterable:

for i in mySeq:
print("existing file:", i)

for( Sequence::const_iterator it = mySeq.begin(), itEnd = mySeq.end(); it != itEnd; ++it )
{
std::cout << *it << std::endl;
}

The _sequenceParser.so* files should be inside pySequenceParser.

mv _sequenceParser.so* python2.7.2/site-packages/pySequenceParser/

Branch: tmp_dude
Commit: 651aea9

Since
commit 20e2c1e

Bad rpath on dynamic libraries.
No symbolic links.

The goal is to recognize only some sequences.

test1.maya // directory
test2.maya // file without extension

Results: 1 sequence of folders (and test2.maya is a folder).
Expected: 1 folder and 1 file.
Solution: discover sequences based on pattern AND type.

When we do an item stat, return group_name and user_name based on user_id and group_id.

Hi,

The previous version of the sequenceParser currently used into tuttle is able to manage muttiple syntaxes. So it could be used to convert a pattern. Is it available in the new version too?

tuttleofx/TuttleOFX#25

the parsing will not found no strict padding:
for example a directory with files: f8.png, f9.png, f10.png, f11.png list:

"/datas/tt/toto/ft#.fits" [8:9]
"/datas/tt/toto/ft##.fits" [10:11]

wishes:
"/datas/tt/toto/ft#.fits" [8:11]

and if just these f9.png, f10.png files are present, it detect like file, and print nothing.
wishes:
"/datas/tt/toto/ft#.fits" [9:10]

Under Linux, with the comand line: lss /
give this result:
Listing 25 items took 1 millisecond
UNITFILE "opt"
UNITFILE "srv"
UNITFILE "usr"
UNITFILE "proc"
UNITFILE "lost+found"
UNITFILE "tmp"
UNITFILE "selinux"
UNITFILE "lib64"
UNITFILE "home"
UNITFILE "mnt"
UNITFILE "success"
UNITFILE "dev"
UNITFILE "root"
UNITFILE "run"
UNITFILE "sys"
UNITFILE ".config"
UNITFILE "var"
UNITFILE "windows"
UNITFILE ".readahead"
UNITFILE "boot"
UNITFILE "media"
UNITFILE "etc"
UNITFILE "bin"
UNITFILE "lib"
UNITFILE "sbin"

They are dectected as UnitFile, not Folders.