aspera-cli 4.0.0 → 4.1.0

data/bin/dascli

@@ -0,0 +1,13 @@
+#!/usr/bin/env bash
+tag=martinlaurent/ascli:latest
+case "$1" in install)
+docker pull $tag
+exit 0
+esac
+conf_host=${ASCLI_HOME:-$HOME/.aspera/ascli}
+if test ! -d $conf_host;then
+	echo creating foder: $conf_host
+	mkdir -p $conf_host
+fi
+conf_img=/usr/src/app/config
+exec docker run --tty --interactive --rm --env ASCLI_HOME="$conf_img" --volume "$conf_host:$conf_img" $tag ascli "$@"

data/docs/README.erb.md

@@ -15,27 +15,78 @@ prstt=opprst.capitalize # in title
 gemspec=Gem::Specification::load(ENV["GEMSPEC"]) or raise "error loading #{ENV["GEMSPEC"]}"
 geminstadd=gemspec.version.to_s.match(/\.[^0-9]/)?' --pre':''
 -%>
-# <%=tool%> : a Command Line for IBM Aspera products
+# <%=tool%> : Command Line Interface for IBM Aspera products
 
 Version : <%= gemspec.version.to_s %>
 
 _Laurent/2016-<%=Time.new.year%>_
 
-This gem provides a command line interface to Aspera Applications.
+This gem provides <%=tool%>: a command line interface to Aspera Applications.
 
-Location (once released):
-[<%= gemspec.metadata['rubygems_uri'] %>](<%= gemspec.metadata['rubygems_uri'] %>)
+<%=tool%> is a also great tool to learn Aspera APIs.
 
-Disclaimers:
+Ruby Gem: [<%= gemspec.metadata['rubygems_uri'] %>](<%= gemspec.metadata['rubygems_uri'] %>)
 
-* This has not yet been officially released so things may change
+Ruby Doc: [<%= gemspec.metadata['documentation_uri'] %>](<%= gemspec.metadata['documentation_uri'] %>)
 
-<%=tool%> is also a great tool to learn Aspera APIs.
+Ruby version must be >= <%= gemspec.required_ruby_version %>
+
+# <a name="when_to_use"></a>When to use and when not to use
+
+<%=tool%> is designed to be used as a command line tool to:
+
+* execute commands on Aspera products
+* transfer to/from Aspera products
+
+So it is designed for:
+
+* Interactive operations on a text terminal (typically, VT100 compatible)
+* Batch operations in (shell) scripts (e.g. cron job)
+
+<%=tool%> can be seen as a command line tool integrating:
+
+* a configuration file (config.yaml) and advanced command line options
+* cURL (for REST calls)
+* Aspera transfer (ascp)
+
+One might be tempted to use it as an integration element, e.g. by building a command line programmatically, and then executing it. It is generally not a good idea.
+For such integration cases, e.g. performing operations and transfer to aspera products, it is preferred to use [Aspera APIs](https://ibm.biz/aspera_api):
+
+* Product APIs (REST) : e.g. AoC, Faspex, node
+* Transfer SDK : with gRPC interface and laguage stubs (C, C++, Python, .NET/C#, java, ruby, etc...)
+
+Using APIs (application REST API and transfer SDK) will prove to be easier to develop and maintain.
+
+For scripting and ad'hoc command line operations, <%=tool%> is perfect.
+
+# Notations
 
 In examples, command line operations (starting with `$`) are shown using a standard shell: `bash` or `zsh`.
+Prompt `# ` refers to user `root`, prompt `xfer$ ` refer to user `xfer`.
 
 Command line parameters in examples beginning with `my_`, like `my_param_value` are user-provided value and not fixed value commands.
 
+# <a name="parsing"></a>Shell and Command line parsing
+
+<%=tool%> is typically executed in a shell, either interactively or in a script. <%=tool%> receives its arguments from this shell.
+
+On Linux and Unix environments, this is typically a POSIX shell (bash, zsh, ksh, sh). In this environment shell command line parsing applies before <%=tool%> (Ruby) is executed, e.g. [bash shell operation](https://www.gnu.org/software/bash/manual/bash.html#Shell-Operation). Ruby receives a list parameters and gives it to <%=tool%>. So special character handling (quotes, spaces, env vars, ...) is done in the shell.
+
+On Windows, `cmd` is typically used. Windows process creation does not receive the list of arguments but just the whole line. It's up to the program to parse arguments. Ruby follows the Microsoft C/C++ parameter parsing rules.
+
+* [Windows: How Command Line Parameters Are Parsed](https://daviddeley.com/autohotkey/parameters/parameters.htm#RUBY)
+* [Understand Quoting and Escaping of Windows Command Line Arguments](http://www.windowsinspired.com/understanding-the-command-line-string-and-arguments-received-by-a-windows-program/)
+
+In case of doubt of argument values after parsing test like this:
+
+```
+$ <%=cmd%> conf echo "Hello World" arg2 3
+"Hello World"
+ERROR: Argument: unprocessed values: ["arg2", "3"]
+```
+
+`echo` displays the value of the first argument using ruby syntax (strings get double quotes) after command line parsing (shell) and extended value parsing (ascli), next command line arguments are shown in the error message.
+
 # Quick Start
 
 This section guides you from installation, first use and advanced use.
@@ -111,77 +162,154 @@ Then, follow the section relative to the product you want to interact with ( Asp
 
 # <a name="installation"></a>Installation
 
-In order to use the tool or the gem, it is necessary to install those components:
+It is possible to install *either* as a docker container or step by step on the native host:
 
-* [Ruby](#ruby)
+* [Ruby](#ruby) version >= <%= gemspec.required_ruby_version %>
 * [<%= gemspec.name %>](#the_gem)
-* [FASP](#fasp_prot)
+* [Aspera SDK (ascp)](#fasp_prot)
 
 The following sections provide information on the installation.
 
+An internet connection is required for the installation. If you dont have internet for the installation, refer to section [Installation without internet access](#offline_install).
+
+## Docker container
+
+This method installs a docker image that contains: Ruby, ascli and the FASP sdk.
+
+Ensure that you have Docker installed.
+
+```
+$ docker --version
+```
+
+Download the wrapping script:
+
+```
+$ curl -o <%=cmd%> https://raw.githubusercontent.com/IBM/aspera-cli/develop/bin/dascli
+$ chmod a+x <%=cmd%>
+```
+
+Install the container image:
+
+```
+$ ./<%=cmd%> install
+```
+
+Start using it !
+
+Note that the tool is run in the container.
+
+The wrapping script maps the container folder `/usr/src/app/config` to configuration folder `$HOME/.aspera/<%=cmd%>`.
+
+To transfer to/from the native host, you will need to map a volume in docker or use the config folder (already mapped).
+
 ## <a name="ruby"></a>Ruby
 
-A ruby interpreter is required to run the tool or to use the gem and tool.
+Use this method to install on the native host.
 
-Ruby minimum version: <%= gemspec.required_ruby_version %>
+A ruby interpreter is required to run the tool or to use the gem and tool.
 
-Ruby 3 is also supported.
+Ruby minimum version: <%= gemspec.required_ruby_version %>. Ruby version 3 is also supported.
 
-Any type of Ruby installation can be used (installer, rpm, rvm, ...).
+*Any type of Ruby installation can be used* : rpm, yum, dnf, rvm, brew, windows installer, ... .
 
 Refer to the following sections for a proposed method for specific operating systems.
 
-### macOS
+The recommended installation method is `rvm` for systems with "bash-like" shell (Linux, Macos, Windows with cygwin, etc...).
+If the generic install is not suitable (e.g. Windows, no cygwin), you can use one of OS-specific install method.
+If you have a simpler better way to install Ruby version >= <%= gemspec.required_ruby_version %> : use it !
 
-MacOS 10.13+ (High Sierra) comes with a recent Ruby, so you can use it directly, you will need to install <%= gemspec.name %> using `sudo` :
+### Generic: RVM: single user installation (not root)
+
+Use this method which provides more flexibility.
+
+Install "rvm": follow [https://rvm.io/](https://rvm.io/) :
+
+Install the 2 keys
 
 ```
-$ sudo gem install <%= gemspec.name %><%=geminstadd%>
+$ gpg2 --keyserver hkp://pool.sks-keyservers.net --recv-keys 409B6B1796C275462A1703113804BB82D39DC0E3 7D2BAF1CF37B13E2069D6956105BD0E739499BDB
 ```
 
-Alternatively, if you use [Homebrew](https://brew.sh/) already you can install Ruby with it:
+Execute the shell/curl command. As regular user, it install in the user's home: `~/.rvm` .
 
 ```
-$ brew install ruby
+$ \curl -sSL https://get.rvm.io | bash -s stable
 ```
 
-It is also possib le to use `rvm`
+If you keep the same terminal (ont needed if re-login):
 
-### Windows
+```
+$ source ~/.rvm/scripts/rvm
+```
 
-Install Latest stable Ruby using [https://rubyinstaller.org/](https://rubyinstaller.org/).
+It is advised to get one of the pre-compiled ruby version, you can list with: 
 
-Go to "Downloads".
+```
+$ rvm list --remote
+```
 
-Select the Ruby 2 version "without devkit", x64 corresponding to the one recommended "with devkit". Devkit is not needed.
+Install the chosen pre-compiled Ruby version:
+
+```
+$ rvm install 2.7.2 --binary
+```
 
-At the end of the installer uncheck the box to skip the installation of "MSys2".
+Ruby is now installed for the user, go on to Gem installation.
 
-### Linux
+### Generic: RVM: global installation (as root)
 
-Install Latest Ruby 2 using "rvm" [https://rvm.io/](https://rvm.io/) .
-It installs by default in /usr/local/rvm , but you can install in another location:
+Follow the same method as single user install, but execute as "root".
+
+As root, it installs by default in /usr/local/rvm for all users and creates `/etc/profile.d/rvm.sh`.
+One can install in another location with :
 
 ```
-curl -sSL https://get.rvm.io | bash -s -- --path /usr/local
+# curl -sSL https://get.rvm.io | bash -s -- --path /usr/local
 ```
 
-Once installed, you can install latest ruby:
+As root, make sure this will not collide with other application using Ruby (e.g. Faspex).
+If so, one can rename the login script: `mv /etc/profile.d/rvm.sh /etc/profile.d/rvm.sh.ok`.
+To activate ruby (and <%=cmd%>) later, source it:
 
 ```
-# rvm install ruby
+# source /etc/profile.d/rvm.sh.ok
+# rvm version
 ```
 
-If you dont want all users to have ruby by default,
-rename the file: `/etc/profile.d/rvm.sh` with another extension, and source it to get rvm.
+### Windows: Installer
+
+Install Latest stable Ruby using [https://rubyinstaller.org/](https://rubyinstaller.org/) :
+
+* Go to "Downloads".
+* Select the Ruby 2 version "without devkit", x64 corresponding to the one recommended "with devkit". Devkit is not needed.
+* At the end of the installer uncheck the box to skip the installation of "MSys2": not needed.
 
-Alternatively, only if you know what you do, on RPM based systems (CentOs, Redhat), install the ruby provided by yum which may be 2.0.
+### macOS: pre-installed or `brew`
+
+MacOS 10.13+ (High Sierra) comes with a recent Ruby. So you can use it directly. You will need to install <%= gemspec.name %> using `sudo` :
+
+```
+$ sudo gem install <%= gemspec.name %><%=geminstadd%>
+```
+
+Alternatively, if you use [Homebrew](https://brew.sh/) already you can install Ruby with it:
+
+```
+$ brew install ruby
+```
+
+### Linux: package
+
+If your Linux distribution provides a standard ruby package, you can use it provided that the version is compatible (check at beginning of section).
+
+Example:
 
 ```
 # yum install -y ruby rubygems ruby-json
 ```
 
-One can cleanup your whole yum-installed ruby environment like this to uninstall:
+One can cleanup the whole yum-installed ruby environment like this to uninstall:
 
 ```
 gem uninstall $(ls $(gem env gemdir)/gems/|sed -e 's/-[^-]*$//'|sort -u)
@@ -206,6 +334,50 @@ For instance to build from source, and install in `/opt/ruby` :
 # make install
 ```
 
+### <a name="offline_install"></a>Installation without internet access
+
+Note that currently no pre-packaged version exist yet.
+A method to build one provided here:
+
+On a server with the same OS version and with internet access follow the "Generic single user installation" method.
+
+Then create an archive:
+
+```
+$ cd
+$ tar zcvf rvm-<%=cmd%>.tgz .rvm
+```
+
+Get the Aspera SDK. Execute:
+
+```
+$ <%=cmd%> conf --show-config|grep sdk_url
+```
+
+Then download the SDK archive from that URL.
+
+Another method for the SDK is to install the SDK (`<%=cmd%> conf ascp install`) on the first system, and archive `$HOME/.aspera`.
+
+Transfer those 2 archives to the target system without internet access.
+
+On the target system:
+
+* Extract the RVM archive either in a global location, or in a user's home folder : `path_to_rvm_root`
+* in the user's `.profile` add this line: (replace `path_to_rvm_root` with the actual location)
+
+```
+source path_to_rvm_root/scripts/rvm
+rvm use 2.7.2
+```
+
+For the SDK, either install from archive:
+
+```
+$ <%=cmd%> conf ascp install --sdk-url=file:///SDK.zip
+```
+
+or restore the `$HOME/.aspera` folder for the user.
+
 ## <a name="the_gem"></a>`<%= gemspec.name %>` gem
 
 Once you have Ruby and rights to install gems: Install the gem and its dependencies:
@@ -220,21 +392,40 @@ To upgrade to the latest version:
 # gem update <%= gemspec.name %>
 ```
 
+<%=tool%> checks every week if a new version is available and notify the user in a WARN log. To de-activate this feature set the option `version_check_days` to `0`, or specify a different period in days.
+
+To check manually:
+
+```
+# <%=cmd%> conf check_update
+```
+
+
+
 ## <a name="fasp_prot"></a>FASP Protocol
 
-Most file transfers will be done using the FASP protocol. Only two additional files are required to perform
-an Aspera Transfer:
+Most file transfers will be done using the FASP protocol, using `ascp`.
+Only two additional files are required to perform an Aspera Transfer, which are part of Aspera SDK:
 
 * ascp
 * aspera-license (in same folder, or ../etc)
 
-This can be installed directly with
+This can be installed either be installing an Aspera transfer sofware, or using an embedded command:
 
 ```
 $ <%=cmd%> conf ascp install
 ```
 
-Those can be found in one of IBM Aspera transfer server or client with its license file (some are free):
+If a local SDK installation is prefered instead of fetching from internet: one can specify the location of the SDK file:
+
+```
+$ curl -Lso SDK.zip https://ibm.biz/aspera_sdk
+$ <%=cmd%> conf ascp install --sdk-url=file:///SDK.zip
+```
+
+The format is: `file:///<path>`, where `<path>` can be either a relative path (not starting with `/`), or an absolute path.
+
+If the embedded method is not used, the following packages are also suitable:
 
 * IBM Aspera Connect Client (Free)
 * IBM Aspera Desktop Client (Free)
@@ -243,7 +434,7 @@ Those can be found in one of IBM Aspera transfer server or client with its licen
 * IBM Aspera High Speed Transfer EndPoint (Licensed)
 
 For instance, Aspera Connect Client can be installed
-by visiting the page: [http://downloads.asperasoft.com/connect2/](http://downloads.asperasoft.com/connect2/).
+by visiting the page: [https://www.ibm.com/aspera/connect/](https://www.ibm.com/aspera/connect/).
 
 <%=tool%> will detect most of Aspera transfer products in standard locations and use the first one found.
 Refer to section [FASP](#client) for details on how to select a client or set path to the FASP protocol.
@@ -251,6 +442,33 @@ Refer to section [FASP](#client) for details on how to select a client or set pa
 Several methods are provided on how to start a transfer. Use of a local client is one of them, but
 other methods are available. Refer to section: [Transfer Agents](#agents)
 
+## <a name="offline_install"></a>Offline Installation (without internet)
+
+The procedure consists in:
+
+* Follow the non-root installation procedure with RVM, including gem
+* archive (zip, tar) the main RVM folder (includes <%=cmd%>):
+
+```
+$ cd ~
+$ tar zcvf rvm_<%=cmd%>.tgz .rvm
+```
+
+* retrieve the SDK:
+
+```
+$ curl -Lso SDK.zip https://ibm.biz/aspera_sdk
+```
+
+* on the system without internet access:
+
+```
+$ cd ~
+$ tar zxvf rvm_<%=cmd%>.tgz
+$ source ~/.rvm/scripts/rvm
+$ <%=cmd%> conf ascp install --sdk-url=file:///SDK.zip
+```
+
 # <a name="cli"></a>Command Line Interface: <%=tool%>
 
 The `<%= gemspec.name %>` Gem provides a command line interface (CLI) which interacts with Aspera Products (mostly using REST APIs):
@@ -431,6 +649,7 @@ By default, a table output will display one line per entry, and columns for each
 * a,b,c : the list of attributes specified by the comma separated list
 * Array extended value: for instance, @json:'["a","b","c"]' same as above
 * +a,b,c : add selected properties to the default selection.
+* -a,b,c : remove selected properties from the default selection.
 
 ## <a name="extended"></a>Extended Value Syntax
 
@@ -444,25 +663,25 @@ The extended value syntax is:
 
 The difference between reader and decoder is order and ordinality. Both act like a function of value on right hand side. Decoders are at the beginning of the value, followed by a single optional reader, followed by the optional value.
 
-The following "readers" are supported:
+The following "readers" are supported (returns value in []):
 
-* @val:VALUE , prevent further special prefix processing, e.g. `--username=@val:laurent` sets the option `username` to value `laurent`.
-* @file:PATH , read value from a file (prefix "~/" is replaced with the users home folder), e.g. --key=@file:~/.ssh/mykey
-* @path:PATH , performs path expansion (prefix "~/" is replaced with the users home folder), e.g. --config-file=@path:~/sample_config.yml
-* @env:ENVVAR , read from a named env var, e.g.--password=@env:MYPASSVAR
-* @stdin: , read from stdin (no value on right)
-* @preset:NAME , get whole <%=opprst%> value by name
+* @val:VALUE : [String] prevent further special prefix processing, e.g. `--username=@val:laurent` sets the option `username` to value `laurent`.
+* @file:PATH : [String] read value from a file (prefix "~/" is replaced with the users home folder), e.g. --key=@file:~/.ssh/mykey
+* @path:PATH : [String] performs path expansion (prefix "~/" is replaced with the users home folder), e.g. --config-file=@path:~/sample_config.yml
+* @env:ENVVAR : [String] read from a named env var, e.g.--password=@env:MYPASSVAR
+* @stdin: : [String] read from stdin (no value on right)
+* @preset:NAME : [Hash] get whole <%=opprst%> value by name
 
 In addition it is possible to decode a value, using one or multiple decoders :
 
-* @base64: decode a base64 encoded string
-* @json: decode JSON values (convenient to provide complex structures)
-* @zlib: uncompress data
-* @ruby: execute ruby code
-* @csvt: decode a titled CSV value
-* @lines: split a string in multiple lines and return an array
-* @list: split a string in multiple items taking first character as separator and return an array
-* @incps: include values of presets specified by key include_presets in hash
+* @base64: [String] decode a base64 encoded string
+* @json: [any] decode JSON values (convenient to provide complex structures)
+* @zlib: [String] uncompress data
+* @ruby: [any] execute ruby code
+* @csvt: [Array] decode a titled CSV value
+* @lines: [Array] split a string in multiple lines and return an array
+* @list: [Array] split a string in multiple items taking first character as separator and return an array
+* @incps: [Hash] include values of presets specified by key `incps` in input hash
 
 To display the result of an extended value, use the `config echo` command.
 
@@ -494,7 +713,7 @@ $ <%=cmd%> config echo @csvt:@file:test.csv
 :......:.....................:
 ```
 
-Example: create a hash and include values from preset named "config" of config file
+Example: create a hash and include values from preset named "config" of config file in this hash
 
 ```
 $ <%=cmd%> config echo @incps:@json:'{"hello":true,"incps":["config"]}'
@@ -533,9 +752,9 @@ It can be overriden using the envinonment variable `<%=evp%>HOME`.
 Example (Windows):
 
 ```
-$ set <%=evp%>HOME=C:\Users\Kenji\.aspera\ascli
+$ set <%=evp%>HOME=C:\Users\Kenji\.aspera\<%=cmd%>
 $ <%=cmd%> config folder
-C:\Users\Kenji\.aspera\ascli
+C:\Users\Kenji\.aspera\<%=cmd%>
 ```
 
 ## <a name="configfile"></a>Configuration file
@@ -983,9 +1202,11 @@ If it possible to send using a HTTP gateway, in case FASP is not allowed.
 Example:
 
 ```
-$ <%=cmd%> faspex package recv --id=323 --transfer=httpgw --transfer-info=@json:'{"url":"https://eudemo.asperademo.com:9443/aspera/http-gwy/v1"}'
+$ <%=cmd%> faspex package recv --id=323 --transfer=httpgw --transfer-info=@json:'{"url":"https://asperagw.example.com:9443/aspera/http-gwy/v1"}'
 ```
 
+Note that the gateway only supports transfers authorized with a token.
+
 ## <a name="transferspec"></a>Transfer Specification
 
 Some commands lead to file transfer (upload/download), all parameters necessary for this transfer
@@ -1032,8 +1253,8 @@ The option `to_folder` provides an equivalent and convenient way to change this
 
 ### List of files for transfers
 
-When uploading, downloading or sending files, the user must specify
-the list of files to transfer. Most of the time, the list of files to transfer will be simply specified on the command line:
+When uploading, downloading or sending files, the user must specify the list of files to transfer.
+Most of the time, the list of files to transfer will be simply specified on the command line:
 
 ```
 $ <%=cmd%> server upload ~/mysample.file secondfile
@@ -1047,7 +1268,8 @@ $ <%=cmd%> server upload --sources=@args ~/mysample.file secondfile
 
 More advanced options are provided to adapt to various cases. In fact, list of files to transfer are conveyed using the [_transfer-spec_](#transferspec) using the field: "paths" which is a list (array) of pairs of "source" (mandatory) and "destination" (optional).
 
-Note that this is different from the "ascp" command line. The paradigm used by <%=tool%> is: all transfer parameters are kept in [_transfer-spec_](#transferspec) so that execution of a transfer is independent of the transfer agent. Note that other IBM Aspera interfaces use this: connect, node, transfer sdk.
+Note that this is different from the "ascp" command line. The paradigm used by <%=tool%> is:
+all transfer parameters are kept in [_transfer-spec_](#transferspec) so that execution of a transfer is independent of the transfer agent. Note that other IBM Aspera interfaces use this: connect, node, transfer sdk.
 
 For ease of use and flexibility, the list of files to transfer is specified by the option `sources`. Accepted values are:
 
@@ -1290,7 +1512,7 @@ updated: my_aoc_org
 Define this <%=prst%> as default configuration for the `aspera` plugin:
 
 ```
-$ <%=cmd%> config id default set aspera my_aoc_org
+$ <%=cmd%> config id default set aoc my_aoc_org
 ```
 
 Note: Default `auth` method is `web` and default `redirect_uri` is `http://localhost:12345`. Leave those default values.
@@ -1664,7 +1886,7 @@ Then, create two shared folders located in two regions, in your files home, in a
 Then, transfer between those:
 
 ```
-$ <%=cmd%> -Paoc_show aspera files transfer --from-folder='IBM Cloud SJ' --to-folder='AWS Singapore' 100GB.file --ts=@json:'{"target_rate_kbps":"1000000","multi_session":10,"multi_session_threshold":1}'
+$ <%=cmd%> -Paoc_show aoc files transfer --from-folder='IBM Cloud SJ' --to-folder='AWS Singapore' 100GB.file --ts=@json:'{"target_rate_kbps":"1000000","multi_session":10,"multi_session_threshold":1}'
 ```
 
 * create registration key to register a node
@@ -2065,23 +2287,67 @@ $ <%=cmd%> node access_key create --value=@json:'{"id":"eudemo-sedemo","secret":
 
 # Plugin: IBM Aspera Faspex
 
-Note that the command "v4" requires the use of APIv4, refer to the Faspex Admin manual on how to activate.
+Notes:
+
+* the command "v4" requires the use of APIv4, refer to the Faspex Admin manual on how to activate.
+* for full details on Faspex API, refer to: [Reference on Developer Site](https://www.ibm.com/products/aspera/developer)
 
-## Sending a Package
+## Faspex 5 Beta1
+
+As the web UI does not yet allow adding API client yet, the way to use CLI is:
+
+* open a browser
+* start developer mode
+* login to faspex 5
+* find the first API call with `Authorization` token, and copy it (kind of base64 long string)
+
+Use it as password and use `--auth=boot`.
+
+```
+$ <%=cmd%> conf id f5boot update --url=https://localhost/aspera/faspex --auth=boot --password=ABC.DEF.GHI...
+```
+
+An JWT client can then be created with a private key:
 
-Provide delivery info in JSON, example:
+```
+$ jsonk=$(openssl rsa -in  ~/.aspera/ascli/aspera_on_cloud_key -pubout 2> /dev/null | sed -e :a -e N -e '$!ba' -e 's/\n/\\n/g')
+$ <%=cmd%> faspex5 -Pf5boot auth_client create --value=@json:'{"name":"hello","client_type":"public","redirect_uris":["https://localhost:12345"],"allow_jwt_grant":true,"public_key":"'$jsonk'"}'
+```
+
+or deleted by name:
 
 ```
---delivery-info=@json:'{"title":"my title","recipients":["laurent.martin.aspera@fr.ibm.com"]}'
+$ id=$(<%=cmd%> faspex5 auth_client list --select=@json:'{"name":"hello"}' --fields=client_id --format=csv)
+$ <%=cmd%> faspex5 auth_client delete --id=$id
+```
+
+Once the API client is created with a client_id and secret (result of create command), create a configuration:
+
 ```
+$ <%=cmd%> conf id f5 update --url=https://localhost/aspera/faspex --auth=jwt --client-id=abcd --client-secret=def --username=pierre@example.com --private-key=@val:@file:~/.aspera/ascli/aspera_on_cloud_key
+$ <%=cmd%> conf id default set faspex5 f5
+``` 
+
+Ready to use Faspex5 with CLI.
+
+Once the graphical registration form exist, ther bootstrap method can be removed.
+
+## Sending a Package
+
+The command is `faspex package send`. Package information (title, note, metadata, options) is provided in option `delivery_info`. (Refer to Faspex API).
+
+Example:
 
-a note can be added: `"note":"Please ..."`
+```
+$ <%=cmd%> faspex package send --delivery-info=@json:'{"title":"my title","recipients":["laurent.martin.aspera@fr.ibm.com"]}' --url=https://faspex.corp.com/aspera/faspex --username=foo --password=bar /tmp/file1 /home/bar/file2
+```
 
-metadata: `"metadata":{"Meta1":"Val1","Meta2":"Val2"}`
+If the recipient is a dropbox, just provide the name of the dropbox in `recipients`: `"recipients":["My Dropbox Name"]`
 
+Additional optional parameters in `delivery_info`:
 
-Note for full details, refer to:
-[Reference on Developer Site](https://developer.asperasoft.com/web/faspex/sending)
+* Package Note: : `"note":"note this and that"`
+* Package Metadata: `"metadata":{"Meta1":"Val1","Meta2":"Val2"}`
 
 ## operation on dropboxes
 
@@ -2144,6 +2410,8 @@ $ for p in 1 2 3;do <%=cmd%> shares2 admin users list --value=@json:'{"page":'$p
 # Plugin: IBM Cloud Object Storage
 
 The IBM Cloud Object Storage provides the possibility to execute transfers using FASP.
+It uses the same transfer service as Aspera on Cloud.
+see [https://status.aspera.io](https://status.aspera.io)
 
 Required options are either:
 
@@ -2200,7 +2468,14 @@ Endpoints for regions can be found by querying the `endpoints` URL.
 For convenience, let us create a default configuration, for example:
 
 ```
-$ <%=cmd%> conf id mycos update --service-credentials=@val:@json:@file:$HOME/service_creds.json  --region=us-south --bucket=laurent
+$ <%=cmd%> conf id mycos update --bucket=laurent --service-credentials=@val:@json:@file:~/service_creds.json --region=us-south
+$ <%=cmd%> conf id default set cos mycos
+```
+
+or using direct parameters:
+
+```
+$ <%=cmd%> conf id mycos update --bucket=mybucket --endpoint=https://s3.us-east.cloud-object-storage.appdomain.cloud --apikey=abcdefgh --crn=crn:v1:bluemix:public:iam-identity::a/xxxxxxx
 $ <%=cmd%> conf id default set cos mycos
 ```
 
@@ -2266,6 +2541,7 @@ The tool requires the following external tools available in the `PATH`:
 * OptiPNG : `optipng`
 * FFmpeg : `ffmpeg` `ffprobe`
 * Libreoffice : `libreoffice`
+* ruby gem `mimemagic`
 
 Here shown on Redhat/CentOS.
 
@@ -2277,6 +2553,26 @@ To check if all tools are found properly, execute:
 $ <%=cmd%> preview check
 ```
 
+### mimemagic
+
+To benefit from extra mime type detection install gem mimemagic:
+
+```
+# gem install mimemagic
+```
+
+or to install an earlier version if any problem:
+
+```
+# gem install mimemagic -v '~> 0.3.0'
+```
+
+To use it, set option `mimemagic` to `yes`: `--mimemagic=yes`
+
+If not used, Mime type used for conversion is the one provided by the node API.
+
+If used, it the `preview` command will first analyse the file content using mimemagic, and if no match, will try by extension.
+
 ### Image: Imagemagick and optipng
 
 ```
@@ -2301,90 +2597,110 @@ The generation of preview in based on the use of `unoconv` and `libreoffice`
 # dnf install unoconv
 ```
 
+* Amazon Linux
+
+```
+# amazon-linux-extras enable libreoffice
+# yum clean metadata
+# yum install libreoffice-core libreoffice-calc libreoffice-opensymbol-fonts libreoffice-ure libreoffice-writer libreoffice-pyuno libreoffice-impress
+# wget https://raw.githubusercontent.com/unoconv/unoconv/master/unoconv
+# mv unoconv /usr/bin
+# chmod a+x /usr/bin/unoconv
+```
 
 ## Configuration
 
-Like any <%=tool%> commands, parameters can be passed on command line or using a configuration <%=prst%>. Note that if you use the <%=tool%> run as `xfer` user, like here, the configuration file must be created as the same user. Example using a <%=prst%> named `my_preset_name` (choose any name relevant to you, e.g. the AoC node name, and replace in the following lines):
+The preview generator is run as a user, preferably a regular user (not root). When using object storage, any user can be used, but when using local storage it is usually better to use the user `xfer`, as uploaded files are under this identity: this ensures proper access rights. (we will assume this)
+
+Like any <%=tool%> commands, parameters can be passed on command line or using a configuration <%=prst%>.  The configuration file must be created with the same user used to run so that it is properly used on runtime.
+
+Note that the `xfer` user has a special protected shell: `aspshell`, so changing identity requires specification of alternate shell:
 
 ```
 # su -s /bin/bash - xfer
-$ <%=cmd%> config id my_preset_name update --url=https://localhost:9092 --username=my_access_key --password=my_secret --skip-types=office --lock-port=12346
-$ <%=cmd%> config id default set preview my_preset_name
+$ <%=cmd%> config id previewconf update --url=https://localhost:9092 --username=my_access_key --password=my_secret --skip-types=office --lock-port=12346
+$ <%=cmd%> config id default set preview previewconf
 ```
 
-Here we assume that Office file generation is disabled, else remove the option. For the `lock_port` option refer to a previous section in thsi manual.
+Here we assume that Office file generation is disabled, else remove the option. `lock_port` prevents concurrent execution of generation when using a scheduler.
 
 Once can check if the access key is well configured using:
 
 ```
-$ <%=cmd%> -Pmy_preset_name node browse /
+$ <%=cmd%> -Ppreviewconf node browse /
 ```
 
 This shall list the contents of the storage root of the access key.
 
 ## Execution
 
-The tool intentionally supports only a "one shot" mode in order to avoid having a hanging process or using too many resources (calling REST api too quickly during the scan or event method).
-It needs to be run regularly to create or update preview files. For that use your best
+The tool intentionally supports only a "one shot" mode (no infinite loop) in order to avoid having a hanging process or using too many resources (calling REST api too quickly during the scan or event method).
+It needs to be run on a regular basis to create or update preview files. For that use your best
 reliable scheduler. For instance use "CRON" on Linux or Task Scheduler on Windows.
 
-Typically, for "Access key" access, the system/transfer is `xfer`. So, in order to be consiustent have generate the appropriate access rights, the generation process
-should be run as user `xfer`.
+Typically, for "Access key" access, the system/transfer is `xfer`. So, in order to be consistent have generate the appropriate access rights, the generation process should be run as user `xfer`.
 
 Lets do a one shot test, using the configuration previously created:
 
 ```
 # su -s /bin/bash - xfer
-$ <%=cmd%> preview scan --overwrite=always
+xfer$ <%=cmd%> preview scan --overwrite=always
 ```
 
 When the preview generator is first executed it will create a file: `.aspera_access_key`
-which contains the access key used.
+in the previews folder which contains the access key used.
 On subsequent run it reads this file and check that previews are generated for the same access key, else it fails. This is to prevent clash of different access keys using the same root.
 
 ## Configuration for Execution in scheduler
 
-Here is an example of configuration for use with cron on Linux. Adapt the scripts to your own preference.
+Here is an example of configuration for use with cron on Linux.
+Adapt the scripts to your own preference.
 
 We assume here that a configuration preset was created as shown previously.
 
-Here the cronjob is created for `root`, and changes the user to `xfer`, also overriding the shell which should be `aspshell`. (adapt the command below, as it would override existing crontab). It is also up to you to use directly the `xfer` user's crontab. This is an example only.
+Lets first setup a script that will be used in the sceduler and sets up the environment.
+
+Example of startup script `cron_<%=cmd%>`, which sets the Ruby environment and adds some timeout protection:
+
+```
+#!/bin/bash
+# set a timeout protection, just in case
+case "$*" in *trev*) tmout=10m ;; *) tmout=30m ;; esac
+. /etc/profile.d/rvm.sh
+rvm use 2.6 --quiet
+exec timeout ${tmout} <%=cmd%> "${@}"
+```
+
+Here the cronjob is created for user `xfer`.
 
 ```
-# crontab<<EOF
-2-59 * * * * su -s /bin/bash - xfer -c 'nice +10 timeout 10m <%=cmd%> preview event --log-level=info --logger=syslog --iteration-file=/tmp/preview_restart.txt'
-0 * * * *    su -s /bin/bash - xfer -c 'nice +10 timeout 30m <%=cmd%> preview scan  --log-level=info --logger=syslog'
+xfer$ crontab<<EOF
+0    * * * *  /home/xfer/cron_<%=cmd%> preview scan --logger=syslog --display=error
+2-59 * * * *  /home/xfer/cron_<%=cmd%> preview trev --logger=syslog --display=error
 EOF
 ```
 
-Nopte that the options here may be located in the config preset, but it was left on the command line to keep stdout for command line execution of preview.
+Note that the loging options are kept in the cronfile instead of conf file to allow execution on command line with output on command line.
 
 ## Candidate detection for creation or update (or deletion)
 
-The tool will find candidates for preview generation using three commands:
+The tool generates preview files using those commands:
 
 * `trevents` : only recently uploaded files will be tested (transfer events)
 * `events` : only recently uploaded files will be tested (file events: not working)
-* `scan` : deeply scan all files under the access key&apos;s "storage root"
-* `folder` : same as `scan`, but only on the specified folder&apos;s "file identifier"
-* `file` : for an individual file generation
-
-Note that for the `event`, the option `iteration_file` should be specified so that
-successive calls only process new events. This file will hold an identifier
-telling from where to get new events.
-
-It is also possible to test a local file, using the `test` command.
+* `scan` : recursively scan all files under the access key&apos;s "storage root"
+* `test` : test using a local file
 
 Once candidate are selected, once candidates are selected,
 a preview is always generated if it does not exist already,
 else if a preview already exist, it will be generated
-using one of three overwrite method:
+using one of three values for the `overwrite` option:
 
 * `always` : preview is always generated, even if it already exists and is newer than original
 * `never` : preview is generated only if it does not exist already
 * `mtime` : preview is generated only if the original file is newer than the existing
 
-Deletion of preview for deleted source files: not implemented yet.
+Deletion of preview for deleted source files: not implemented yet (TODO).
 
 If the `scan` or `events` detection method is used, then the option : `skip_folders` can be used to skip some folders. It expects a list of path relative to the storage root (docroot) starting with slash, use the `@json:` notation, example:
 
@@ -2618,6 +2934,34 @@ $ <%=cmd%> server upload source_hot --to-folder=/Upload/target_hot --lock-port=1
 
 The local (here, relative path: source_hot) is sent (upload) to basic fasp server, source files are deleted after transfer. growing files will be sent only once they dont grow anymore (based ona 8 second cooloff period). If a transfer takes more than the execution period, then the subsequent execution is skipped (lock-port).
 
+# Aspera Health check and Nagios
+
+Each plugin provide a `health` command that will check the health status of the application. Example:
+
+```
+$ <%=cmd%> console health
++--------+-------------+------------+
+| status | component   | message    |
++--------+-------------+------------+
+| ok     | console api | accessible |
++--------+-------------+------------+
+```
+
+Typically, the health check uses the REST API of the application with the following exception: the `server` plugin allows checking health by:
+
+* issuing a transfer to the server
+* checking web app status with `asctl all:status`
+* checking daemons process status
+
+<%=tool%> can be called by Nagios to check the health status of an Aspera server. The output can be made compatible to Nagios with option `--format=nagios` :
+
+```
+$ <%=cmd%> server health transfer --to-folder=/Upload --format=nagios --progress=none
+OK - [transfer:ok]
+$ <%=cmd%> server health asctlstatus --cmd_prefix='sudo ' --format=nagios
+OK - [NP:running, MySQL:running, Mongrels:running, Background:running, DS:running, DB:running, Email:running, Apache:running]
+```
+
 # Module: `Aspera`
 
 Main components:
@@ -2637,6 +2981,16 @@ This sample code shows some example of use of the API as well as
 REST API.
 Note: although nice, it's probably a good idea to use RestClient for REST.
 
+Example of use of the API of Aspera on Cloud:
+
+```
+require 'aspera/aoc'
+
+aoc=Aspera::AoC.new(url: 'https://sedemo.ibmaspera.com',auth: :jwt, scope: 'user:all', private_key: File.read(File.expand_path('~/.aspera/<%=cmd%>/aspera_on_cloud_key')),username: 'laurent.martin.aspera@fr.ibm.com',subpath: 'api/v1')
+
+aoc.read('self')
+```
+
 # History
 
 When I joined Aspera, there was only one CLI: `ascp`, which is the implementation of the FASP protocol, but there was no CLI to access the various existing products (Server, Faspex, Shares). Once, Serban (founder) provided a shell script able to create a Faspex Package using Faspex REST API. Since all products relate to file transfers using FASP (ascp), I thought it would be interesting to have a unified CLI for transfers using FASP. Also, because there was already the `ascp` tool, I thought of an extended tool : `eascp.pl` which was accepting all `ascp` options for transfer but was also able to transfer to Faspex and Shares (destination was a kind of URI for the applications).
@@ -2653,11 +3007,25 @@ So, it evolved into <%=tool%>:
 * supports transfers with multiple [Transfer Agents](#agents), that&apos;s why transfer parameters moved from ascp command line to [_transfer-spec_](#transferspec) (more reliable , more standard)
 * `ruby` is consistent with other Aspera products
 
+# Changes (Release notes)
 
+* 4.x
 
-# Release Notes
+  	* fix: remove keys from transfer spec and command line when not needed
+  	* fix: default to create_dir:true so that sending single file to a folder does not rename file if folder does not exist 
+  	* new: update documentation with regard to offline and docker installation
+ 	* new: renamed command `nagios_check` to `health`
+	* new: agent `http_gw` now supports upload
+	* new: added option `sdk_url` to install SDK from local file for offline install
+	* new: check new gem version periodically
+	* new: the --fields= option, support -_fieldname_ to remove a field from default fields
+	* new: Oauth tokens are discarded automatically after 30 minutes (useful for COS delegated refresh tokens)
+	* new: mimemagic is now optional, needs manual install for `preview`, compatible with version 0.4.x
+	* new: AoC a password can be provided for a public link
+	* new: `conf doc` take an optional parameter to go to a section
+	* new: initial support for Faspex 5 Beta 1
 
-* 4.0.0.pre2
+* 4.0.0
 
 	* now available as open source at [<%= gemspec.homepage %>](<%= gemspec.homepage %>) with general cleanup
 	* changed default tool name from `mlia` to `ascli`
@@ -2785,7 +3153,7 @@ So, it evolved into <%=tool%>:
 
 * 0.10.6
 
-	* FaspManager: transfer spec `authentication` no more needed for local tranfer to use aspera public keys. public keys will be used if there is a token and no key or password is provided.
+	* FaspManager: transfer spec `authentication` no more needed for local tranfer to use Aspera public keys. public keys will be used if there is a token and no key or password is provided.
 	* gem version requirements made more open
 
 * 0.10.5