I have been developing an app for an iDevice, and I wanted to create a .deb file of my app, so I created a simple app in Xcode to test things out, but I am having difficulty creating the .deb file.
I did the following steps:
1) deleted the .DS_Store files, *$ rm -rf .DS_Store*
2) $ dpkg-deb -b ~/repo/debs
When it creates the .deb file it is only 16KB which I know the test app should be a little bigger.
The directory structure for the app is as follows:
+- MyProgram
+- Applications
| +- MyProgram.app
| +- Info.plist
| +- MyProgram
| +- icon.png
+- DEBIAN
| +- control
+- System
+- Library
+- LaunchDaemons
+- com.saurik.MyProgram.plist
I didn't include a com.saurik.MyProgram.plist file because I didn't see a file resembling one in the build directory, when I built the app in Xcode.
Basically the app is a simple viewcontroller with a label. The app does contain a .storyboard file if that makes a difference.
I got this working, and created a shell script in the process. The shell script is as follows:
#!/bin/bash
echo "Update Repo Script Started"
cd ~/Projects/<ProjectName>
ldid -S app/<Project.app>/<Project>
cp -R app/<Project.app> ~/packages/<Project>/Applications/
cd ~/packages
echo $PWD
dpkg-deb -b <Project> ~/repo/debs/
cd ~/repo
dpkg-scanpackages debs / > Packages
bzip2 -fks Packages
cd ..
ssh -n <host> 'rm -rf /path/to/your/repo'
scp -r repo <host>:/home/<user>/www/
echo "Repo Updated"
Related
Here is the folder structure I have.
Workspace folder: D:\Node\MyNode\
When Jenkins build runs on a node, the files from scm gets downloaded to the following folder: D:\Node\MyNode\xx_development
I need to do cd to the folder "xx_development" and this name xx can change for different strasms (RTC) but "_development" remains same.
how can I do cd to a folder with (*development) using a pipeline script?
Edit: I am using windows Nodes for Jenkins.
To change the current directory to the folder with the pattern *_development, you can use the following script:
For Windows:
def folder = bat(returnStdout: true, script: 'dir /b /ad | findstr "_development"').trim()
bat "cd ${folder}"
dir /b /ad | findstr "_development" --> lists all directories in the current folder and filters them by the pattern _development.
/b --> to list only the directory names.
/ad --> to list only directories.
findstr --> to filter the output by the pattern _development.
The second line changes the current directory to the directory stored in the Folder variable.
For Linux:
def Folder = sh(returnStdout: true, script: 'ls -d */ | grep "_development"').trim()
sh "cd ${Folder}"
ls -d */ | grep "_development" --> lists all directories in the folder and filters by the pattern _development.
trim() --> If there are any leading or trailing whitespaces, they are removed using this command.
The second line changes the current directory to the folder stored in the Folder variable.
I am running WSL Ubuntu 20.04 (Version 2 with Docker Desktop Support) within Windows 10 Pro Version 21H1
The steps are as follows:
git clone https://github.com/textileio/powergate.git
cd powergate/
cd docker/
nano docker-compose.yaml where I added "["lotus", "daemon", "--import-snapshot", "https://fil-chain-snapshots-fallback.s3.amazonaws.com/mainnet/minimal_finality_stateroots_latest.car"]" between lines 32 and 33.
make up
Waited for the node to finish importing and then syncing.
^C then make down then deleted the line "["lotus", "daemon", "--import-snapshot", "https://fil-chain-snapshots-fallback.s3.amazonaws.com/mainnet/minimal_finality_stateroots_latest.car"]" from docker-compose.yaml
make up
Now that the node was running I typed cd .. so I was in the repo's root directory, then make install-pow
with the pow command in my GOPATH I typed pow to make sure pow was linked fine to powd. It was.
pow admin users create
copied the token and ran export POW_TOKEN=<token copied to here>
Then pow wallet addrs and funded the address
I went to the directory behind the folder of my static website which is about 5GB in size.
I typed pow data stage <my-static-site-folder>
After it was finished staging and printed out the CID I typed pow config apply --watch <CID waited a long time while it said the job was executing and then I got...
---------------------------------------+--------------------------------+-------+-------+--------------
<job id here> | JOB_STATUS_FAILED executing | | |
| cold-storage config: making | | |
| deal configs: getting miners | | |
| from minerselector: getting | | |
| miners from reputation | | |
| module: not enough miners from | | |
| reputation module to satisfy | | |
| the constraints | | |
I don't understand what the problem is. I repeated the pow config apply --watch <CID command each time adding the --override flag with several different modifications to a custom config file. The content did appear briefly on IPFS (not Filecoin), but after I continued running the config apply command the site went down from IPFS.
This problem can be fixed by adding miners to the "trustedMiner" entry in the config file because pow doesn't necessary detect miners that fit your specs.
I went to a Filecoin miner info aggregation site (I used "https://filrep.io/") and added miners to the trustedMiner section of the config file used in the apply command to start a Filecoin deal.
For example the "trustedMiners" line in your config file should look like this:
"trustedMiners": ["<Miner Id>", "<Miner Id>","<Miner Id>", "<Miner Id>", ...],
with however many miners you want to add.
Then you would execute the command:
pow config apply --watch <CID> -o -c new-config-file.json
Btw the --watch flag is optional as it just allows you to see the status of the deal in real time.
I work on a Maven project that need to copy mode than 10 GB of artifacts in a target repository from a maven local repository (after downloaded them).
In some cases (e.g. for tests), I'd like to replace this copy by a symlink creation in order to save few minutes.
My question is: Is there a way to ask to plugin maven-dependency-plugin goal copy-dependencies to create a symlink OR is there any maven plugin that can do it.
The copy-dependencies goal cannot, to my knowledge, do this out of the box. However, you can use a shell script:
#!/bin/sh
outputDir=target/dependency
mkdir -p "$outputDir"
mvn dependency:resolve |
grep ':\(compile\|runtime\)' | sed 's/\[INFO\] *//' |
while read gav
do
case "$gav" in
*:*:*:*:*:*) # G:A:P:C:V:S
g="${gav%%:*}"; remain="${gav#*:}"
a="${remain%%:*}"; remain="${remain#*:}"
p="${remain%%:*}"; remain="${remain#*:}"
c="${remain%%:*}"; remain="${remain#*:}"
v="${remain%%:*}"
s="${remain#*:}"
;;
*:*:*:*:*) # G:A:P:V:S
g="${gav%%:*}"; remain="${gav#*:}"
a="${remain%%:*}"; remain="${remain#*:}"
p="${remain%%:*}"; remain="${remain#*:}"
c=""
v="${remain%%:*}"
s="${remain#*:}"
;;
esac
g=$(echo "$g" | sed 's/\./\//g')
test -n "$c" && artName="$a-$v-$c" || artName="$a-$v"
ln -s "$HOME/.m2/repository/$g/$a/$v/$artName.$p" "$outputDir"
done
I could do this for .zip files in the folder using the command below:
for f in "!"; do unzip -d "${f%*.zip}" "$f"; done
The above command extracts all .zip files in a given folder to subfolders, having content and name of respective .zip files.
But I couldn't find a command that would do the same for .tar files. Please help.
Btw, I am trying to do this on a remote server using WinSCP/putty. So, I cannot use a GUI software. I need a command, thus the question.
After a bit of fiddling I came up with for f in $(find -maxdepth 1 | grep .tar); do mkdir ${f%.tar}; tar -xaf $f -C ${f%.tar} ; done appears to work, so long as the file name does not contain any spaces. I assume you wanted the directory from foo.tar to be named foo (no file extension). If you want the directory to be named foo.tar (with file extension) then try using for f in $(find -maxdepth 1 | grep .tar); do mkdir $f ; tar -xaf $f -C $f ; done.
IIRC, the remote access client Cyberduck can handle compressed files in a GUI - so you can try that if you're fine with a GUI solution.
I've fought against Xcode with regards to this before, where I want to add a list of files and directories to be copied into the built app-package, and XCode only wants to let me add entire folders. Now I need a proper solution...
I have a workspace with multiple targets, one per application. I have a directory structure with lots of assets/data files structured a bit like this:
- Data
|- Common
| |-Scripts
| |-Images
|- AppA
| |-Scripts
| |-Images
|- AppB
| |-Scripts
| |-Images
I want to add Data/Common/* to my targets AppA & AppB, and then Data/AppA/* to AppA, etc.
What I find is if I add a folder reference to Data to my XCode project, I cannot select workspaces - I only can set which targets Data is associated with.
I could add folder references to each subfolder individually but then I think this would break the directory structure I want to achieve. Also, it just seems to get messy... say I don't want all of Common in both apps, but to cherry-pick certain sub-dirs/files for each app?
So, is there a more arbitrary way in XCode[4] to tell it which files go where? I'm aware I can write a custom bash-script build phase, I used to do that in fact but it was really bad for build performance.
Option 1) Create a bash installer and hardcode the paths inside the main project:
## Compression Script
mkdir -vp installer/payload
cd installer/
tar tvf files.tar
echo "Running Installer"
mkdir $HOME/files
tar ./files.tar -C $HOME/files
## Decompression Script
#!/bin/bash
echo ""
echo "Self Extracting Installer"
echo ""
export TMPDIR=`mktemp -d /tmp/selfextract.XXXXXX`
ARCHIVE=`awk '/^__ARCHIVE_BELOW__/ {print NR + 1; exit 0; }' $0`
tail -n+$ARCHIVE $0 | tar xzv -C $TMPDIR
CDIR=`pwd`
cd $TMPDIR
./installer
cd $CDIR
rm -rf $TMPDIR
exit 0
__ARCHIVE_BELOW__
Option 2) Use pkgbuild, productbuild and pkgutil like so:
Making OS X Installer Packages like a Pro - Xcode Developer ID ready pkg