Workbook

Goal of this gem is to make working with workbooks (spreadsheets) as programmer friendly as possible. Not reinventing a totally new DSL or all kinds of new methodnames, but just borrowing from known concepts such as hashes and arrays (much like (Faster)CSV does)). Workbook is a gem that mimicks a typical spreadsheet, a bundle of sheets, bundled in a workbook. A sheet may contain one or more tables (which might be the multi table sheets of Apple Numbers or Excel ranges). Basically:

Book
- Sheet (one or more)
  - Table (one or more)

Subsequently a table consists of:

Table
- Row (one or more)
  - Cell ( wich has may have a (shared) Format )

Book, Sheet, Table and Row inherit much of the behaviours from the Array and Hash classes, it tries to walk and quack as such. The row is extended with hashlike lookups (row[:id]) and writers (row[:id]=). Values are converted to ruby native types, and optional parsers can be added to improve recognition.

In addition to offering you this plain structure it allows for importing .xls, .csv, .xlsx, .txt files (more to come), writing .xls, and .csv (more to come) and includes several utilities to easily create an overview of the differences between two tables and output basic cell-styling properties as css.

The Basics

Simply initialize a simple spreadsheet using:

b = Workbook::Book.new

b = Workbook::Book.open filename

Calling

s = b.sheet
t = s.table

will give you the first Sheet and Table (if one doesn't exist it is created on the fly).

You can initialize with simple 2-d array like this:

b = Workbook::Book.new [['a','b'],[1,2],[3,4],[5,6]]
t = s.sheet.table

Subsequently you look up values in the table like this:

t[1][:b]
# returns <Workbook::Cel @value=2>

which is equivalent to

t[1][1]

Of course you'll be able to write a new value back to it. If you just enter a value, formatting of the original cell will be maintained.

t[1][:b] = 5

Alternatively (more spreadsheet like) you can read cells like this (writing to be supported, not implemented yet)

t['A2']

If you want to use an existing file as a template (which you can create in Excel to create nice looking templates), simply clone the row, and add it back:

b = Workbook::Book.open("template.xls")
table = b.sheet.table
template_row = table[1]            # can be any, but I typically have a well
                                # formatted header row + an example template
                                # row for the data
[1,2,3,4].each do |v|
  new_row = template_row.clone
  table << new_row              # to use the symbol style header references,
                                # the row first needs to be added back to the
                                # table
  new_row[:a] = v
end
table.delete(template_row)      # you don't want the template to show up
                                # in the endresult
b.write("result.xls")           # write it!

Another typical use case is exporting a list of ActiveRecord-objects to xls (it is assumed that the headers of the excel-table correspond (like "Total order price" and total_order_price match) to the headers of the database-table ):

b = Workbook::Book.open("template.xls")
table = b.sheet.table
template_row = table[1]         # see above
Order.where("created_at > ?", Time.now - 1.week).each do |order|
  new_row = template_row.clone
  new_row.table = table
  order.to_hash.each{|k,v| row[k]=v}
end
table.delete(template_row)      # you don't want the template to show up
                                # in the endresult
b.write("recent_orders.xls")    # write it!

Utilities

Sorting

Sorting leaves the header alone, if it exists, and doesn't complain about comparing strings with dates with floats (Ever found OpenOffice Calc or Excel complainging about its inability to compare integers and strings? We're talking spreadsheet here). When classes are different the following (default) order is used: Numbers, Strings, Dates and Times, Booleans and Nils (empty values).

t.sort

To implement:

To some extent, sort_by works, it doesn't, however, adhere to the header settings...

t.sort_by {|r| r[:b]}

Comparing tables or entire workbooks

Simply call on a Workbook::Table

t1.diff t2

And a new book with a new table will be returned containing the differences between the two tables.

Alternatively you can run the same command on workbooks, which will compare sheet by sheet and return a new Workbook

Writing

Currently writing is limited to the following formats. Templating support is still limited.

b.to_xls                  # returns a spreadsheet workbook
b.write_to_xls filename   # writes to filename
t.(write_)to_csv          # returns a csv-string (called on tables)
b.(write_)to_html         # returns a clean html-page with all tables; unformatted, format-names are used in the classes
t.(write_)to_json         # returns the values of a table in json
t.(write_)to_xlsx         # returns/writes using RubyXL to XLS (unstable, work in progress)

In case you want to display a formatted table in HTML, some conversion is offered to convert text/background properties to css-entities. Internally the hash storing style elements tries to map to CSS where possible.

Compatibility

Workbook is automatically tested. Check Travis for Workbook's current build status of the supported ruby versions .

Future

Column support, e.g. t[:b] could then return Workbook::Column<[<Workbook::Cel @value=2>,<Workbook::Cel @value=4>,<Workbook::Cel @value=6>]>
In the future I hope to return the cell value as inheriting from the original value's class, so you don't have to call #value as often.
xlsx support definitly needs to be improved. Especially template based support.

Alternatives

The ruby toolbox lists plenty of alternatives, that just didn't suit my needs.

License

Workbook uses the following gems:

FasterCSV Used for reading CSV (comma separated text) and TXT (tab separated text) files (Copyright © James Edward Gray II; GPL2 & Ruby License)

Subclassing Workbook

One thing I was mulling over was how to subclass Workbook in order to add domain-specific functionality to the classes. I didn't want to monkey-patch your classes, as we may need more than one domain with different functionality, so subclassing seems the way to go.

The issue I was running into was that if I subclass Workbook::Book and Workbook::Sheet, for example, my Workbook::Book subclass will still use the super-classes, such as creating a Workbook::Sheet in #initialize or #push (instead of my new Sheet subclass). Short of overriding every method that references one of the other Workbook classes (ugh), there doesn't appear to be any way to easily do this right now (maybe I'm missing something obvious?).

Two solutions spring to mind: 1) create a new module for each domain, with classes that inherit from Workbook classes. And modify Workbook so that every reference to the Workbook constant is changed to something like self.class.const_get(self.class.name.split('::')[0..-2].join), which would get the module for self. We'd probably put that code in a helper method.

Or 2) replace (in Workbook) every reference to a Workbook class constant (e.g., Workbook::Sheet) with a call to a helper method that returns the class constant. Then when you subclass a Workbook class, you just need to override those helper methods to point to your subclasses instead of the base Workbook classes.

These options are fairly similar; option 1 has the advantage of putting the subclasses into a new module (although you could do that with option 2 also), and it doesn't require you to override any methods. option 2 is probably a bit simpler.

Any ideas or feedback?

murb / workbook Goto Github PK

workbook's Introduction

Workbook

The Basics

Utilities

Sorting

Comparing tables or entire workbooks

Writing

Compatibility

Future

Alternatives

License

workbook's People

Contributors

Stargazers

Watchers

Forkers

workbook's Issues

Recommend Projects

Recommend Topics

Recommend Org

Jobs