combine recursive and move options in camel - apache-camel

I'm using recursive=true option in File Component route.
I want to use the move option in order to move the file to sufolder of the current folder , and also rename the file by adding a "backup_" prefix.
I notice that the recursive make a problem as it create directories structure for the move file and also add the prefix to the directories name.
How can I solve it and combine recursive and move options ?

Related

How to rename a file while using "move" in URL in apache camel

I have an URL like
url = "file:D:/inputFolder?move=D:/outputFolder". we are making this url dynamically.
I want to rename the file while moving, So I made it something like this
url = "file:D:/inputFolder?move=D:/outputFolder&fileName=abc.txt". But I think move and fileName do not work together, it is not renaming.
Is there any alternative to do it? Please remember I want with "move" only.
I cannot use .setHeader(..) also.
Thanks,
Hy,
as far as I understand you, your trying to move the file in one single uri.
That is not really how camel works.
The idea of camel is to have a "consumer" and a "producer", where the consumer loads data (e.g. your file) and the producer puts the data somewhere (e.g. save the file into a folder)
That being said, here is what worked for me with a java route:
from("file:/home/chris/temp/camel/in")
.to("file:/home/chris/temp/camel/out/?fileName=test.txt");
The from part configures the folder where camel looks for new files. A few notes on that:
The file component checks the folder each 0.5 sec for new files. This can be changed with the delay parameter
The option noop configures, if the file is being moved or copied. By default it is set to false, which means it is moved
In the to part you configure, where the file is supposed to be moved. Here you can use the fileName parameter to rename the file.
Be careful with this though, because setting an option in the uri directly does make it "static".
What I mean by that is, that the only way of changing the parameter is by completely reconfiguring the route or by restarting it, where neither is something you would want to do normally.
Note 1:
Moving all files that are put into one folder into the same file always overrides the previous file by default.
You could, for example, use the fileExists parameter to always just append the content of the file: fileExists=Append (See camel file docu for details)
Note 2:
There is an option in the file component to not "move" the file, but copy, rename and delete it, which sometimes is necessary, when you want to move it onto a different drive and a simple copy does not work.
Also see the docu for the camel file component for details on that.
Note 3:
You can have multiple to() statements in the same route to have the file moved to multiple locations. For example:
from("file:/home/chris/temp/camel/in")
.to("file:/home/chris/temp/camel/out/?fileName=test.txt")
.to("smtp:....");
Hope I could help you and answer you question.
Greets
Chris
Two possible ways to achieve your goal.
Use both "consumer" and "producer"
Using this way, you are free to control where and how your destination can be set and has great freedom to control filename with the use of a processor/bean.
from("file:D:/inputFolder")
.to("file:D:/outputFolder?fileName=abc.txt")
Use "consumer" only
Using this way, you are treating your work as source data control. This can be use when your file is going to move within same drive. The drawback is the filename rename pattern is limited (refer to camel file language)
from("file:D:/inputFolder?move=${file:parent}/../outputFolder/abc.txt")

Kettle Get Files Names - listing files from a list of folder names

I am trying to something fairly simple: I have a first "Get File Names" step that generates a list of folder names and a second "Get File Names" step that should take the folder names from the first and list the files in that folder. However, I have trouble getting the second step to work.
Example below
Anyway it is not recursive. It just make what u described. Only files in second level directory.
If u need recursive just use JavaScript Step and Java API to get recursive list of files.

Loop through a directory in Talend

I have a Directory with many sub-directories in it including one named as OLD. This OLD folder could also be inside any of the sub directories and contains archived files.
Root Directory A
SUB-DIRECTORY A
file1.txt
file2.txt
SUB-DIRECTORY B
file1.txt
file2.txt
OLD
SUB-DIRECTORY C
file1.txt
file2.txt
SUB-DIRECTORY D
file1.txt
OLD
SUB-SUB-DIRECTORY E
file7.txt
OLD
I need to create a job in Talend which shall look for all OLD folders (in main and in sub directory both) and delete the files from that folder. I can use tFilelist to and mask the files to be deleted. But unable to figure out how to configure the job to look for OLD folder in all sub directories and delete those files also.
What you need to do is in the tFileList put your main folder choose check box include sub directories and in the FileList type drop down list choose Directories. Your file mask should be "OLD" or if it is anything more "OLD".
Iterate and use the parameter ((String)globalMap.get("tFileList_1_CURRENT_FILEPATH"))
to capture your sub directories. Now you can use this folder path in a tFileDelete Which can also delete folders
follow below approach..
Add tFileList and configure to Travers over all the directories.
now use if conndition connection from tFileList
Add tJava and connect with iterator connection
Add tFileDelete after tJava and connect with IF condition.
Add below condition inside if condition.
((String)globalMap.get("tFileList_2_CURRENT_FILEPATH")).contains("OLD")
Now you will get the all files from all the directories above code will give pass to the files which has the "OLD" in there file path.
I have not tested but you can try it.

Conditionally ignore path from Subversion?

Is it possible to globally ignore a folder IF it is a child of a folder having a specific name? For example...
Exclude:
client/vendor
... or ...
app/vendor
But never exclude a "vendor" folder if it appears anywhere else?
I'm working on an AngularJS project and the "vendor" folder is common for client-side files. However, theoretically, it is possible that "vendor" may have another meaning in future projects and, if it does, it would generally be in another path.
The docs on this are a bit misleading (to me, anyway). It says to use the svn:ignore property but no examples anywhere show how to specify the conditional parent folder. They all appear to be manually ignoring a specific folder every time... via a command line.
Per the TortoiseSVN docs:
No Paths in Global Ignore List (Link here) You should not include
path information in your pattern. The pattern matching is intended to
be used against plain file names and folder names. If you want to
ignore all CVS folders, just add CVS to the ignore list. There is no
need to specify CVS */CVS as you did in earlier versions. If you want
to ignore all tmp folders when they exist within a prog folder but not
within a doc folder you should use the svn:ignore property instead.
There is no reliable way to achieve this using global ignore patterns.

How do I (recursively?) monitor contents of new directories using inotify?

Firstly, I want to start by using inotify to monitor a specific directory (the main directory) for files and sub-directories. If a new directory is added into this main directory, how would I make sure to monitor this sub-directory with inotify? How would I monitor a new directory within this sub-directory of the main directory?
I think adding it to the watch is easy by using the inotify_add_watch() function but I do not know how to get the correct relative path address of files and directories within sub-directories (to use for like Dropbox-like syncing in a different location while maintaining the correct directory tree, for example).
Well the fastest to implement (but not the fastest in reality) would be to:
Create the initial tree of directories by recursively exploring the children; An example in C/Linux can be found here:
http://www.lemoda.net/c/recursive-directory/
Add a watch for each subdirectory; When something has been modified or changed you can parse all children recursively and see the differences. Something similar was discussed here:
How to monitor a folder with all subfolders and files inside?
If this solution doesn't appeal to you, you might try to do a polling mechanism such that you must re-check the whole structure using a thread at a certain time interval.
Hope it helps!

Resources