All Projects → tschf → atom-build-oracle

tschf / atom-build-oracle

Licence: MIT License
Oracle compilation plugin for atom-build

Programming Languages

javascript
184084 projects - #8 most used programming language
PLSQL
303 projects

Oracle compiler for Atom

apm apm apm

Uses the atom-build package to execute Oracle compilations in the Atom editor.

This package requires atom-build to be installed.

Prerequisites

To be able to compile your code, you need either SQL*Plus or SQLcl available on your system - preferably in your PATH. SQL*Plus requires more steps to get set up (such as installing the instant client), but you will get much faster compile times so that is the preferred interpreter (SQLcl is Java based, so launching the JVM each time adds to the compile time).

SQL*Plus set up

OS X

You may face issues depending on how you set up the instant client. The typical guide involves setting an environment variable DYLD_LIBRARY_PATH. Unforunately, in OS X El Capitan, this variable is not passed in as an environment variable into Atom (this is a part of security measure known as System Integrity Protection). You can also see that this is not output if you run the env command on the terminal.

If you followed those general steps, you will likely be getting this error (or very similar):

dyld: Library not loaded: /ade/dosulliv_sqlplus_mac/oracle/sqlplus/lib/libsqlplus.dylib
  Referenced from: /opt/Oracle/instantclient_11_2/sqlplus
  Reason: image not found
Trace/BPT trap: 5

Two ways around this.

  • In the settings for this plugin, set the path where your client files are (copy what you have for your existing DYLD_LIBRARY_PATH).

  • Or, instead of setting DYLD_LIBRARY_PATH, you can make all the libaries available in a location that SQL*Plus knows where to look for them. Two common locations are ~/lib or /usr/local/lib. So, assuming you placed the instant client at /opt/Oracle/instantclient_11_2, run the following:
sudo mkdir -p /usr/local/lib
sudo ln -s /opt/Oracle/instantclient_11_2/libclntsh.dylib.11.1 /usr/local/lib/libclntsh.dylib.11.1
sudo ln -s /opt/Oracle/instantclient_11_2/libnnz11.dylib /usr/local/lib/libnnz11.dylib
sudo ln -s /opt/Oracle/instantclient_11_2/libociei.dylib /usr/local/lib/libociei.dylib
sudo ln -s /opt/Oracle/instantclient_11_2/libsqlplus.dylib /usr/local/lib/libsqlplus.dylib
sudo ln -s /opt/Oracle/instantclient_11_2/libsqlplusic.dylib /usr/local/lib/libsqlplusic.dylib

Ensure your PATH is set up with the instant client directory on it so that sqlplus is on it. In your ~/.bash_profile add the line:

export PATH=/opt/Oracle/instantclient_11_2:$PATH

I went with /usr/local/lib rather than ~/lib so to not fill up my home directory. Aside from that difference, this is much as what is described on the download page of the instant client (down the bottom of the page).

If you previously set up DYLD_LIBRARY_PATH you can unset that and you should be good to go.

Ubuntu/Linux

To be able to run SQL*Plus from Atom, Atom needs to have all the relevant enviornment variables set up. I followed the installation from this guide. The guide mentions to set environment variables in a system wide location or ~/.bash_profile. The guide suggests to install it in a system wide location since you typically only have one Oracle client installed per system. That is a good approach also for Atom, as it will ensure Atom picks up the relevant environment variables. I found that environment variables set in ~/.bash_profile are not picked up by the (GUI) terminal, and upon further digging I found that ~/.bash_profile is only picked up when launching a virtual console - see this post. The other location you may be tempted to use is ~/.bashrc, however this is also not a good place as Atom doesn't seem to pick up the environment set up in there, see this comment.

So, to sum it up. Either place your Oracle variables in a script at /etc/profile.d/oracle.sh (globally set), or place them in ~/.profile which will get picked up by Atom.

SQLcl set up

Windows

An issue you may face on Windows is if you try and compile code containing unicode characters, they won't persist when compiling with SQLcl. The reason is that the JVM doesn't default to UTF-8 encoding on Windows.

To get around this, ensure you point to sql.bat (instead of sql.exe) and also that you have checked "Enforce UTF-8" in the package settings, similar to:

screenshot from 2016-08-28 17-49-57

When Enforce UTF-8 encoding is on, the environment variable JAVA_OPTS will be set to -Dfile.encoding=UTF-8. This is also discussed on Stack overflow: http://stackoverflow.com/questions/361975/setting-the-default-java-character-encoding

Installation

Install through apm or in Atom itself, where the name of the package is build-oracle.

apm install build-oracle

Once installed, you will want to access the settings and set the relevant path for your SQL interpreter. That will be either SQLcl or SQL*Plus. For me, I have sqlplus in my path so no change is necessary (there is an issue where PATH may not be correctly set in Atom, in that case you should set the full path of sqlplus). If I wanted to point it to the binary for SQLcl, which isn't in my path, I would set the configuration to /opt/sqlcl/bin/sql.

Usage instructions

In your project root directory, add a JSON file named .atom-build-oracle.json, where you can define build targets for your project. This contents of the file expects an array of objects, which supports the following parameters:

Property name Description Example
targetName A descriptive name of the build target. This is so you know to which environment you'll be compiling to. hr: dev
connectString The connect string you would used to connect to sqlplus on the command line.
note 1: You must specify a password at this stage.
note 2: connectString is the preferred field to use.
- hr/hr@//server:port/sid
- hr/hr@XE1
host The server of the database example.com
port The database port 1521 (default)
sid The database SID or service name XE
user The schema name you connect to hr
password The password of the schema oracle

note: connectString is built from the other server details if you ommit that setting.

Examples:

Connect string:

Suppose I have the following TNS entry defined:

XE1 =
    (DESCRIPTION =
        (ADDRESS_LIST =
            (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.56.101)(PORT = 1521))

        )
        (CONNECT_DATA =
            (SERVICE_NAME = XE)
        )
    )

My build file becomes:

[
    {
        "targetName" : "hr: dev",
        "connectString" : "hr/hr@XE1"
    }
]

Individual details:

[
    {
        "targetName" : "hr: dev",
        "host" : "example.com",
        "sid" : "xe",
        "port" : 1521,
        "user" : "hr",
        "password" : "hr"
    }
]

Where you can specify any number of build targets. If your database server uses service names instead of SID, you can enter the service name in the property "sid" nonetheless.

Since this file contains sensitive information (password) you will likely also want to add an entry to your .gitignore file so this is not published.

.atom-build-oracle.json

It's worth noting, that to get the build system to recognise the above configuration, you will need to reload atom (only the first time). This can be done in the menu View->Reload or with the keyboard shortcut ctrl+alt+r.

This plugin is an extension to the build system, providing the necessary scripts to compile Oracle code. The build triggers are determined as part of the over-arching build package. To build the active file is done with the keyboard shortcut ctrl+alt+b or cmd+alt+b or f9. By default, this will be the first entry in your build config file. The active build target can be switched with f7, which will also trigger a build on the active file:

The build panel (with build output) can be toggled with f8.

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].