I want to insert UserProfile.nickname as text annotation into convert option.
This is what I have now.
How can I fix this?
user_profile.rb
before_save :text_to_insert
def text_to_insert
nickname = self.nickname
end
has_attached_file :user_avatar,
:styles => {
:thumb=> "100x100>",
:small => "400x400>" },
:convert_options => {
:small => lambda {|a| %Q{ -gravity south -pointsize 25 -fill black -annotate +0+0 "#{nickname}" -fill white -annotate +2+2 "#{nickname} "} } }
Try #{:nickname} or maybe #{self.nickname} instead.
I'm with same problem.
Usually they recomend you to create a custom paperclip processor to do it, but years ago i done something like that in a way which remembers the one o mentioned in your question, but i dont remember exactly how.
If you accomplished, please post your solution here and Let me know if works ;)
Related
I'm using Paperclip in my project but some of my users are complaining that it's incorrectly rotating some images.
For some reasons I can't even imagine I figured it out that some files are with wrong exif orientation attributes. I was looking and I saw that paperclip calls ImageMagick by default using -auto-orient. I saw that the Thumbnail processor has an option to turn auto-orient on or off.
But I couldn't find a way to pass this to the Processor.
This is the code I have:
has_attached_file :photo,
styles: { :square => "400x400#" }
Does anyone now how to do that?
Thanks!
In the end I created a new processor which extends from the paperclip default Thumbnail processor to send the correct options.
class WithouAutoOrientProcessor < Paperclip::Thumbnail
def initialize(file, options = {}, attachment = nil)
options[:auto_orient] = false
super
end
end
And in the model I added
has_attached_file :photo,
styles: { :square => "400x400#" },
processors: [:WithouAutoOrientProcessor]
Although it is a valid option to add your own processor, this is how you pass the option to the processor:
In your styles hash replace your dimension strings with another hash
Put your old dimensions in the key geometry into this hash
The other key/value pairs are the options passed to the processor
You can of course pass auto_orient: false, too
Applying this to your model's code:
has_attached_file :photo,
styles: { square: { geometry: "400x400#", auto_orient: false } }
Say there is a hash field that can have two possible value permutations, "foo" and "bar". How can I validate the hash value is one of the two?
class ValidateMe
validates :type => { :type => "foo" or :type => "bar" }
end
This results in an error. What is the proper way to handle this use case?
My actual case is using Paperclip to attach an image. I need to enforce the image is only .png or .jpg
class ValidateMe
validates_attachment :image,
presence => true,
:content_type => { :content_type => "image/png" }
end
Help with either code block is greatly appreciated. Thanks!
The best way to do this would be to pass an array of types to :content_type
class ValidateMe
validates_attachment :image,
presence => true,
:content_type => { :content_type => ['image/png', 'image/jpeg'] }
end
(My answer is based on code in Paperclip - Validate File Type but not Presence)
This can also be done using regular expressions. (Not as preferable)
class ValidateMe
validates_attachment :image,
presence => true,
:content_type => { :content_type => /^image\/(jpeg|png)$/ }
end
(source How can I restrict Paperclip to only accept images?)
I think Btuman's first answer would be considered canonical: The content_type key of content_type in validates_attachment can accept an array of valid content-types.
You can use :inclusion attribute for more see this http://guides.rubyonrails.org/active_record_validations_callbacks.html
I'm trying to insert the URL of the page into uploaded image.
I already have the code like this below but it doesn't work.
Is there something wrong in my model? How can I fix this?
My associations
User has_one :profile
Profile belongs_to :user
models/user.rb
before_save :text_to_insert?
def text_to_insert
nickname = self.profile.nickname
end
has_attached_file :user_avatar,
:styles => {
:thumb=> "100x100>",
:small => "400x400>" },
:convert_options => {
:small => '-fill white -undercolor "#00000080" -gravity South -annotate +0+5 " example.com/'+ nickname +' "' }
before saving , you are using text_to_insert? method which doesn't exist thatwhy it is returning false,so it fails to save .
It looks like typos ,try removing ? after :text_to_insert ie
before_save :text_to_insert
Please be sure that is valid self.profile.nickname
I want attachment_fu to resize my thumbnails in a similar way to how flickr, facebook and twitter handle this: If I want a 100x100 thumbnail I want the thumbnail to be exactly 100x100 with any excess cropped off so that the aspect ratio is preserved.
Any ideas?
To set up the 100x100 thumbnails, add the following to your model:
has_attachment :content_type => :image,
:storage => IMAGE_STORAGE,
:max_size => 20.megabytes,
:thumbnails => {
:thumb => '100x100>',
:large => '800x600>',
}
(In this example, I am creating a 100x100 thumbnail, and also an 800x600 'large' size, in additional to keeping the original size.)
Also, keep in mind that the thumbnail might not be exactly 100x100; it will have a maximum dimension of 100x100. This means that if the original has an aspect ration of 4:3, the thumbnail would be 100x75. I'm not exactly sure if that is what you meant by "exactly 100x100 with any excess cropped off so that the aspect ratio is preserved."
Add this to your model
protected
# Override image resizing method
def resize_image(img, size)
# resize_image take size in a number of formats, we just want
# Strings in the form of "crop: WxH"
if (size.is_a?(String) && size =~ /^crop: (\d*)x(\d*)/i) ||
(size.is_a?(Array) && size.first.is_a?(String) &&
size.first =~ /^crop: (\d*)x(\d*)/i)
img.crop_resized!($1.to_i, $2.to_i)
# We need to save the resized image in the same way the
# orignal does.
self.temp_path = write_to_temp_file(img.to_blob)
else
super # Otherwise let attachment_fu handle it
end
end
and change the thumbnail size to:
:thumbnails => {:thumb => 'crop: 100x100' }
source:
http://stuff-things.net/2008/02/21/quick-and-dirty-cropping-images-with-attachment_fu/
There's a cropping directive that can be given in the specification:
has_attachment :content_type => :image,
:thumbnails => {
:thumb => '100x100#'
}
Memonic: '#' looks like the crop tool.
Edit: Correction
has_attachment :content_type => :image,
:thumbnails => {
:thumb => '100x100!'
}
The previous method was for Paperclip which has a different notation.
My solution was to delve into the attachment_fu plugin folder (vendor/plugins) and edit the rmagick_processor.rb file. First I renamed resize_image to resize_image_internal, then added:
def resize_image(img, size)
# resize_image take size in a number of formats, we just want
# Strings in the form of "square: WxH"
if (size.is_a?(String) && size =~ /^square: (\d*)x(\d*)/i) ||
(size.is_a?(Array) && size.first.is_a?(String) &&
size.first =~ /^square: (\d*)x(\d*)/i)
iw, ih = img.columns, img.rows
aspect = iw.to_f / ih.to_f
if aspect > 1
shave_off = (iw - ih) / 2
img.shave!(shave_off, 0)
else
shave_off = (ih-iw) / 2
img.shave!(0, shave_off)
end
resize_image_internal(img, "#{$1}x#{$2}!")
else
resize_image_internal(img, size) # Otherwise let attachment_fu handle it
end
end
I can now use 'square: 100x100' as my geometry string. Note that the above code assumes the required output is square.
I'm looking for a way to determine image orientation preferably with Paperclip, but is it even possible or do I need to user RMagick or another image library for this?
Case scenario: When a user uploads an image i want to check the orientation/size/dimensions to determine if the image is in portrait/landscape or square and save this attribute to the model.
Here's what I generally do in my image models. Perhaps it will help:
I use IM's -auto-orient option when converting. This ensures images are always rotated properly after upload
I read the EXIF data after processing and get the width and height (among other things)
You can then just have an instance method that outputs an orientation string based on width and height
has_attached_file :attachment,
:styles => {
:large => "900x600>",
:medium => "600x400>",
:square => "100x100#",
:small => "300x200>" },
:convert_options => { :all => '-auto-orient' },
:storage => :s3,
:s3_credentials => "#{RAILS_ROOT}/config/s3.yml",
:s3_permissions => 'public-read',
:s3_protocol => 'https',
:path => "images/:id_partition/:basename_:style.:extension"
after_attachment_post_process :post_process_photo
def post_process_photo
imgfile = EXIFR::JPEG.new(attachment.queued_for_write[:original].path)
return unless imgfile
self.width = imgfile.width
self.height = imgfile.height
self.model = imgfile.model
self.date_time = imgfile.date_time
self.exposure_time = imgfile.exposure_time.to_s
self.f_number = imgfile.f_number.to_f
self.focal_length = imgfile.focal_length.to_s
self.description = imgfile.image_description
end
Thanks for the answer jonnii.
Although I did find what I was looking for in the PaperClip::Geometry module.
This worked find:
class Image < ActiveRecord::Base
after_save :set_orientation
has_attached_file :data, :styles => { :large => "685x", :thumb => "100x100#" }
validates_attachment_content_type :data, :content_type => ['image/jpeg', 'image/pjpeg'], :message => "has to be in jpeg format"
private
def set_orientation
self.orientation = Paperclip::Geometry.from_file(self.data.to_file).horizontal? ? 'horizontal' : 'vertical'
end
end
This of course makes both vertical and square images have the vertical attribute but that's what I wanted anyway.
When I take a photo with my camera the dimensions of the image are the same regardless if the photo is landscape or portrait. However, my camera is smart enough to rotate the image for me! How thoughtful! The way this work is that is uses something called exif data which is meta data placed on the image by the camera. It includes stuff like: the type of camera, when the photo was taken, orientation etc...
With paperclip you can set up callbacks, specifically what you'll want to do is have a callback on before_post_process that checks the orientation of the image by reading the exif data using a library (you can find a list here: http://blog.simplificator.com/2008/01/14/ruby-and-exif-data/), and then rotating the image clockwise or counterclockwise 90 degrees (you won't know which way they rotated the camera when they took the photo).
I hope this helps!