Google recently changed the way they present the image data, and so the links were no longer being scraped.
I figured out how to get the image urls with the new system and made the appropriate changes so it would work. 

Unfortunately, google no longer provides file format data so I had to try and retrieve it from the url of the image, which does not work in some cases.

By filtering out the image objects which had data[0]==2, I have removed the null items and it will no longer give the error: "TypeError: 'NoneType' object is not subscriptable".

This system is not very flexible, it seems google does not keep the same positions of target items, so sometimes it doens't work. I added a try-except just in case there are more problems

It is based on patch by https://github.com/Joeclinton1, but for some
reason we get escaped string when getting the results page directly
(limit < 101) and unescaped one when getting the results page using
selenium. This is not the most elegant solution, but it works for me.

Merged master and patch-1

…load into patch-1

Previously the end_object for the data pack was found by searching for '</script>' and then going 4 characters back, however google in a recent update has added , 'sideChannel: {}});' to the end of the data pack, which throws it off. To fix this the end_object finding script first searches for '</script>' and then searches for the first ']' to the left of that closing script tag. This should be more flexible.

Previously if the data unpacking failed it would tell the user that the URL could not be opened. But this is the wrong exception. So i fixed this by splitting up the data un packing and url opening into seperate parts so each can have their own exception. This should make it easier to identify what has gone wrong.

just added to wrong directory by accident

Adding offset

Fix clicking on the "Show more results" button with Selenium.

- The button has no more "smb" id
- We need to do more scroll down before clicking

This may have been caused by Google changing their Ajax response. Looking at the response, lines[4] only contained a single number and not any JSON. Removing it and simply pulling from lines[3] seems to fix the issue. The problem only manifested when downloading more than 100 images, which required launching ChromeDriver.

We extracted images from json.loads(data)[31][0]... because in json.loads(data)[31] was a list of 1 value.
Now json.loads(data)[31] is a list of 2 values and we want the last.
So replacing 0 by -1 manage this new case and the old one if Google revert this change.

The time range feature has changed, I used this tweet thread to fix it : https://twitter.com/i/events/1174066444029419520.

We can imagine work on the time_range format to avoid changing the "API".

It is not very useful to have the time range expression in the image directory names.

Get more than 400 images

Fix JSONDecodeError: Extra Data

Manage API change

Fix time_range argument

…fo object.

Update the url building to the new way to get the exact image size thanks to this article :
https://www.labnol.org/internet/google-image-size-search/26902/

Fix exact_size parameter

Support Firefox

Fix issue #20

upstream updates

fix breaking change due to google's response format