Custom Grunt Build Script for Documents

Cancelado Publicado Mar 23, 2014 Pagado a la entrega
Cancelado Pagado a la entrega

Custom Grunt Build "Text-Converter"

Hi, I have hundreds of book-sized documents of various types and want to create a Grunt web-development build process to convert to HTML5 and DB store the documents. I'm a novice with Grunt and don't have time to go this deep right now -- but if you are a wizard with Grunt and Grunt tasks, this might be pretty easy for you.

I've been using various Yeoman generated Grunt configurations and they actually do a lot of what I want but I'm too novice to make many changes to [url removed, login to view] without just breaking everything. What I have in mind could be a modified and extended version of the basic Yeoman "generator-webapp".

I just need a place where I can plug in my code for converting documents. The grunt commands I need are:

grunt (default) - rebuild entire output folder, rebuild index file

grunt serve - open index file & watch for changes with LiveReload

changes rebuild the changed document and the index file

grunt deploy - push changed files to CouchDB

Notes:

* You don't have to actually write the document conversion code, just the Grunt build & deploy processes, leaving a convenient way for me to add the complex conversion steps myself later. That could be one or more custom grunt tasks applied to all documents which I will extend later.

I'll want to be able to pull in other libraries to help me convert many document types such as .doc, .html, .epub, .jade, .md, .txt, .rtf, .pdf. I want it to be easy to add new conversion types as well as sub-types (such as a custom versions of text documents). Sub-types will be detected by the conversion script or determined by an optional metadata JSON file accompanying each document.

* I don't need testing frameworks, since these are just document conversions. You can yank all the moca and phantom stuff out (if you start from the Yeoman project "generator-webapp".)

* The build process should always generate a master index file which provides a file list of all documents so that one can find and view a specific document while editing. The build process should rebuild this file from a template in the app folder -- inserting a JSON object with an array of book metadata such as:

{ documents: [

language: "English:en", bookshelf: "Fiction", title: "Moby-Dick",

author: "Herman Melville", modified: "2013-08-31T09:05:[url removed, login to view]",

size: "102456", crc32: "461707669",

uri: "English/Fiction/Herman Melville/Moby-Dick/[url removed, login to view]"

],

}

The idea is that I should be able to browse each book using the URI. All assets are stored as attachments in a single CouchDB record with the relative path as filename -- so links to resources inside the documents (pictures etc) do not need to be modified to be viewed online. The build script should not care where such files are located but simply move them to a parallel location in the document's output folder.

* Deploy to CouchDB:

Each document gets two CouchDB JSON records. Only changed files need be pushed to the DB.

1) One record contains all the standalone files as attachments (including the generated .html files).

2) A second record contains metadata for each document and a list of associated files from the first record (perhaps just a copy of the first record's _attachments block).

* Use a PouchDB wrapper to communicate with CouchDB (for consistency with my other code)

More: [url removed, login to view]

JavaScript Node.js

Nº del proyecto: #5708850

Sobre el proyecto

1 propuesta Proyecto remoto Activo Mar 27, 2014

1 freelancer está ofertando el promedio de $2000 para este trabajo

ranganathp

Can help... I am an Expert... Lets Start! Please start a Discussion with me and we can get started from there... Please check the past projects I have handled and check my reviews for what employers have to say about Más

$2000 USD en 30 días
(16 comentarios)
5.3