Compare the public API

Based on the original idea of ExamplesFx try to create the HTML of the examples from the Examples solution of the library

In the issue #39 there are some related discussions

Make it cleaner
Remove redundant text
Make use a slideshow for the images and example like: http://www.newtonsoft.com/json

Based on the idea of @mcavigelli we will be moving the new redesigned Home GItHub pages

Currently the engine.Options is partial working, we must to provide an isolation between engines, ie, if we create 2 engines with the same Type and change the options of one engine the other don't get changed

Add the list of all breaking changes from 2.0 to 3.1

The current implementation only support reading runtime records
We need a way to write non typed records (generic CSV for example)

Sometimes you can't edit an existing type to add the necessary attributes (DelimitedRecordAttribute or FixedLengthRecordAttribute) and this information could be specified through parameters.
If the type doesn't have any attribute, you could use a default value, if this information is not specified through a parameter.

I've been annoyed when I wanted to proceed to a quick dump of data to a csv file without wanting to create a new object & copy data in order to be able to use the API.

I tried all possible engines and all of them seem to require those attributes.
Even when it doesn't seem logical : CommonEngine.RecordsToDataTable() only works when the record type has those attributes. It's not logical because I don't think it need those attributes here.

It would also allow the use of anonymous types with the engines, like :

var itemList = data.GroupBy(item => new { item.NomApporteur }).Select(g => new { g.Key.NomApporteur, Somme = g.Sum(o => o.Montant) });
if (itemList.Count() > 0)
{
    var engine = new DelimitedFileEngine(itemList.First().GetType());
    engine.Options.Delimiter = ";";
    engine.WriteFile("", itemList);
}

Today, it seems that I have to write my own RecordsToDataTable() in order to bypass this limitation.

Add the ads at the bottom of the page

The class source don't show
Can't Change column names or types
Can't set field as optional

Something like a DataLink replacement but much more pluggable and that left to the user the core actions of assign fields and transform info, basicaly solve the Extract and Load operations and let the user to full the Transform part.

Integrated with the TPL DataFlow library

Don't use standard framework names like System.String, etc
Use the language specific ones: like string, int, long

No more forums, mails or other way

Currently there are a fair number of fixes that are available on github but not from the NuGet versions. It's possible to work around by cloning from github and building there and using the result, but especially for use cases like LINQPad scripts (where LINQPad has built-in NuGet integration), it would be far nicer to have a recent version on NuGet.

Making it available as a pre-release version of 3.0.0 would be an option, if there are already breaking changes and what's currently checked in can't/shouldn't be released without the major version change.

Thanks!

After Matthias creates it, add the list of people that help us a ton all these years

With @mcavigelli will be working on using the material design theme:

http://themeforest.net/item/materialize-material-design-admin-template/11446068

If we go ahead with the rewrite of the RecordInfo class we can take this into account

Now only supports AutoProperties

Today we throw BadUsageException in the engine creation but maybe the user only want to write records, so there is no problem if there is no constructor.

Throw the BadUsageException only when trying to read objects

That will also speed up a bit the initial setup process because don't need to use reflection to take the constructor pointer on engine creation, that will be delayed to first read :)

Configure Hosting at godaddy and subdomain

http://wizard.filehelpers.net

I would love to extend attributes like FieldAlignAttribute to match my current requirements and to avoid repeating [FieldAlign(AlignMode.Right, '0')].

I could imagine the reason for using sealed is about performance, but are there really any significant gains, to justify the removal of extensibility?

If there are other reasons I haven't thought of, please enlighten me :)

Allos add or remove fields and create the record class

Upload them to YouTube with 1080 resolution

Try to compile the lib with the Mono C# compiler, if not at least check if the dll is compatible

Allow to drop a txt and trigger the auto detection feature, reading the first lines

For example with 3 Combos with the values

I want to:

• Read
• Write
• Transform
• Sort
• Compare

a:

• File
• String
• Stream

with

• Delimited
• Fixed Length
• Dynamic

records

You change the combos and you get the right example

Also in the Tips

Create an empty MVC 6 app and mimic the layout, header and footer of the jekyll site

I have little background on working with advance mocking, mostly for mocking IO ops

So we need your help to identify the key points to make FileHelpers main classes mockeable

That addition must avoid breaking changes, in the worst case we can add some overloads but not change the way users work now with the library

Because are full of spam, I just disabled the Forums, we will review interesting posts and add them to the FAQ and Docs

As a long time user of 2.x - is there a 'conversion' document for

• What has been removed/changed
• Migration guide
• New features, etc?

Once we migrated all in interested pages, we can safely remove the old html and images from the master branch at:

https://github.com/MarcosMeli/FileHelpers/tree/master/Doc/Include

Reported in the old issue tracker

Error Converting 'HEADERTEXT' to type: 'Int32'.   at FileHelpers.ConvertHelpers+Int32Converter.ParseString (System.String from) [0x00000] in <filename unknown>:0 
  at FileHelpers.ConvertHelpers+CultureConverter.StringToField (System.String from) [0x00000] in <filename unknown>:0 
  at FileHelpers.FieldBase.AssignFromString (ExtractedInfo fieldString, FileHelpers.LineInfo line) [0x00000] in <filename unknown>:0

when using with

    [DelimitedRecord(";"), IgnoreFirst]
    public class DataClass


    var engine = new FileHelperEngine<T>();

    if (!File.Exists(file.FullName))
    {
        throw new NotSupportedException("file doesn't exist");
    }
    return engine.ReadFile(file.FullName);

After finishing the redesign point to the GitHub pages the main domain

Here are a list of small pending changes in different pages:

✅ Unify some docs in Must Read page
✅ Add some tips about how to use the library
✅ Add info about Mono support
✅ Review the changelog to add links and more items
◽ Add new attributes
◽ Add small description to attributes

Create a Roslyn Analizer to evaluate library usage and suggest best practices

Deploy it to Nuget and recommend user to install it

For Example:

[DelimiterRecords(";")] 
class internalClass 
{ 
string A; 
string B; 
[FieldConverter(ConverterKind.Date, "dd.MM.yy")] 
DateTime C; 

[FieldIgnored]
string D; 
[FieldIgnored]
int whatEver; 
... 
} 

[DelimiterRecords(";")] 
class MyClass 
{ 
string someOtherInfo; 
string maybeAnotherThing; 

[FieldAsClassConverter] 
[FieldArrayLength(10)] 
internalClass [] internalInfo; 
}

Quick summary of the current attributes indicating when can be used (Read or Write) and for what kind of record types (FixedLength, Delimited)

Let discuss what is the best approach to make the library Task friendly and Thread Safe for concurrent work

Add it to the FileHelpers.Example project and create a pull request
Later we will be publishing it to the site

In the left nav bar

We own the new .net domain and we need to redirect the traffic to the new github pages

Currently at: http://www.codeproject.com/Articles/12170/FileHelpers-v2-0-Delimited-CSV-or-Fixed-Data-Impor

Hallo there!
II would like to suggest a new feature.
Assumed there are a csv File with the semicolon as a delimiter. But some of the fields includes multiple values delimited with other signs. Example:

• Alex;Kovalsky;19.02.2015;50.1
• john;Smith;04.04.2015:13.02.2014;50.1:16.3
• Max;Mustermann;;

I would expect such file to be parsed into

first_name:       Alex
last_name:       Kovalsky
Checkout Date: 19.02.2015
Checkout Sum: 50.1

first_name:       john
last_name:       Smith
Checkout Date: 04.04.2015
Checkout Sum: 50.1

first_name:       john
last_name:       Smith
Checkout Date: 13.02.2014
Checkout Sum: 16.3

first_name:       Max
last_name:       Mustermann
Checkout Date: null
Checkout Sum: 0.0

Or with arrays:

first_name:       Alex
last_name:       Kovalsky
Checkout Date[]: {19.02.2015}
Checkout Sum[]: {50.1}

first_name:       john
last_name:       Smith
Checkout Date[]: {04.04.2015, 13.02.2014}
Checkout Sum[]: {50.1, 16.3}

first_name:       Max
last_name:       Mustermann
Checkout Date[]: {}
Checkout Sum[]: {}