edi-parser v1.0.1
This project provide EDI parsing and has been tested and used on large source file (over 2Gb).
- Support delimiter, quote and escape characters
- Line breaks discovery: line breaks in source are detected and reported to destination
- Data transformation
- Asynch and event based
- Support for large datasets
- Complete test coverage as sample and inspiration
Quick example
Using the library is a 4 steps process:
- Create a source
- Create a destination (optional)
- Transform the data (optional)
- Listen to events (optional)
Here is a example:
// node samples/sample.js
var edi = require('edi');
edi()
.fromPath(__dirname+'/sample.in')
.toPath(__dirname+'/sample.out')
.transform(function(data){
data.unshift(data.pop());
return data;
})
.on('data',function(data,index){
console.log('#'+index+' '+JSON.stringify(data));
})
.on('end',function(count){
console.log('Number of lines: '+count);
})
.on('error',function(error){
console.log(error.message);
});
// Print sth like:
// #0 ["2000-01-01","20322051544","1979.0","8.8017226E7","ABC","45"]
// #1 ["2050-11-27","28392898392","1974.0","8.8392926E7","DEF","23"]
// Number of lines: 2
Installing
Via git (or downloaded tarball):
$ git clone http://github.com/rinie/node-edi-parser.git
Then, simply copy or link the ./lib/edi.js file into your $HOME/.node_libraries folder or inside a declared path folder.
Via npm:
$ npm install edi
Reading API
The following method are available:
fromPath(data, options)
Take a file path as first argument and optionally on object of options as a second arguments.fromStream(readStream, options)
Take a readable stream as first argument and optionally on object of options as a second arguments.from(data, options)
Take a string, a buffer, an array or an object as first argument and optionally some options as a second arguments.
Options are:
delimiter
Set the field delimiter, one character only, default to comma.quote
Set the field delimiter, one character only, default to double quotes.escape
Set the field delimiter, one character only, default to double quotes.columns
List of fields or true if autodiscovered in the first edi line, impact thetransform
argument and thedata
event by providing an object instead of an array, order matters, see the transform and the columns section below.encoding
Default to 'utf8', apply when a readable stream is created.trim
If true, ignore whitespace immediately around the delimiter, default to false.ltrim
If true, ignore whitespace immediately following the delimiter (i.e. left-trim all fields), default to false.rtrim
If true, ignore whitespace immediately preceding the delimiter (i.e. right-trim all fields), default to false.
Writing API
The following method are available:
write(data, preserve)
Take a string, an array or an object, implementation of the StreamWriter API.end()
Terminate the stream, implementation of the StreamWriter API.toPath(path, options)
Take a file path as first argument and optionally on object of options as a second arguments.toStream(writeStream, options)
Take a readable stream as first argument and optionally on object of options as a second arguments.
Options are:
delimiter
Default to the delimiter read option.quote
Default to the quote read option.escape
Default to the escape read option.columns
List of fields, apply whentransform
return an object, order matters, see the transform and the columns sections below.encoding
Default to 'utf8', apply when a writable stream is created.header Display the column names on the first line if the columns option is provided.
lineBreaks
String used to delimit record rows or a special value; special values are 'auto', 'unix', 'mac', 'windows', 'unicode'; default to 'auto' (discovered in source or 'unix' if no source is specified).flags
Default to 'w', 'w' to create or overwrite an file, 'a' to append to a file. Apply when using thetoPath
method.bufferSize
Internal buffer holding data before being flush into a stream. Apply when destination is a stream.end
Prevent callingend
on the destination, so that destination is no longer writable, similar to passing{end: false}
option instream.pipe()
.newColumns If the
columns
option is not specified (which means columns will be taken from the reader options, will automatically append new columns if they are added duringtransform()
.
Transforming data
- transform(callback) User provided function call on each line to filter, enriche or modify the dataset. The callback is called asynchronously.
The contract is quite simple, you receive an array of fields for each record and return the transformed record. The return value may be an array, an associative array, a string or null. If null, the record will simply be skipped.
Unless you specify the columns
read option, data
are provided as arrays, otherwise they are objects with keys matching columns names.
When the returned value is an array, the fields are merge in order. When the returned value is an object, it will search for the columns
property in the write or in the read options and smartly order the values. If no columns
options are found, it will merge the values in their order of appearance. When the returned value is a string, it is directly sent to the destination source and it is your responsibility to delimit, quote, escape or define line breaks.
Example of transform returning a string
// node samples/transform.js
var edi = require('edi');
edi()
.fromPath(__dirname+'/transform.in')
.toStream(process.stdout)
.transform(function(data,index){
return (index>0 ? ',' : '') + data[0] + ":" + data[2] + ' ' + data[1];
});
// Print sth like:
// 82:Zbigniew Preisner,94:Serge Gainsbourg
Events
By extending the Node EventEmitter
class, the library provide a few useful events:
data (function(data, index){}) Thrown when a new row is parsed after the
transform
callback and with the data being the value returned bytransform
. Note however that the event won't be call if transform returnnull
since the record is skipped. The callback provide two arguments:data
is the edi line being processed (by default as an array)index
is the index number of the line starting at zeroend In case your redirecting the output to a file using the
toPath
method, the event will be called once the writing process is complete and the file closed.error Thrown whenever an error is captured.
Columns
Columns names may be provided or discovered in the first line with the read options columns
. If defined as an array, the order must match the one of the input source. If set to true
, the fields are expected to be present in the first line of the input source.
You can define a different order and even different columns in the read options and in the write options. If the columns
is not defined in the write options, it will default to the one present in the read options.
When working with fields, the transform
method and the data
events receive their data
parameter as an object instead of an array where the keys are the field names.
// node samples/column.js
var edi = require('edi');
edi()
.fromPath(__dirname+'/columns.in',{
columns: true
})
.toStream(process.stdout,{
columns: ['id', 'name']
})
.transform(function(data){
data.name = data.firstname + ' ' + data.lastname
return data;
});
// Print sth like:
// 82,Zbigniew Preisner
// 94,Serge Gainsbourg
Running the tests
Tests are executed with expresso. To install it, simple use npm install -g expresso
.
To run the tests expresso test
Contributors
- David Worms : https://github.com/wdavidw
- Will White : https://github.com/willwhite
- Justin Latimer : https://github.com/justinlatimer
- jonseymour : https://github.com/jonseymour
- pascalopitz : https://github.com/pascalopitz
- Josh Pschorr : https://github.com/jpschorr
- Elad Ben-Israel: https://github.com/eladb
Related projects
- Pavel Kolesnikov "ya-edi": http://github.com/koles/ya-edi
- Chris Williams "node-edi": http://github.com/voodootikigod/node-edi