Amazon API Resources

Introduction

此次要使用的Amazon API全名"Amazon Product Advertising API", 原名"Associates Web Service （AWS）", 现在很多相关的库及教程仍使用AWS的称呼。

Python库

2. python-amazon-product-api

经试用可以进行连接与基本查询
documentation
试用时使用的code：
- 安装：$ sudo pip install python-amazon-product-api
- 创建amazon api credentials 文件： ```localhost:~ apple$ touch .amazon-product-api
```
localhost:~ apple$ vim .amazon-product-api
```
  在vim编辑器中编辑文件.amazon-product-api的内容，编辑好的内容如下： ```localhost:~ apple$ cat .amazon-product-api
```
[Credentials]
access_key = XXXX-XXXX-XXXX
secret_key = XXXXXXXXXXXXXXXXXXX
associate_tag = XXXXXXXXXXXXXX
```
- 进入python，试用python-amazon-product-api
```
>>> from amazonproduct import *
>>> api = API(locale='us')
>>> items = api.item_search('Books', Publisher="O'Reilly")
>>> items
>>> type(items)
```
  最后一条命令返回值为 "class 'amazonproduct.processors._lxml.SearchPaginator'"

3. Boto3

在aws.amazon.com上官方给出的AWS SDK(Software Development Kit)，但时还未尝试。

4. bootlenose

试用成功

>>> import bottlenose
>>> AWS_ACCESS_KEY_ID = 'XXXX-XXXX-XXXX' #my access key
>>> AWS_SECRET_ACCESS_KEY = 'XXXXXXXXXXXXXXXXXXX' #my secret access key
>>> AWS_ASSOCIATE_TAG = 'XXXXXXXXXXXXXX' #my associate tag
>>> amazon = bottlenose.Amazon(AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_ASSOCIATE_TAG)
>>> response = amazon.ItemLookup(ItemId="0596520999", ResponseGroup="Images",
... SearchIndex="Books", IdType="ISBN")
>>> type(response)

最后一条命令返回值为 "type 'str'". 这种返回值类型比python-amazon-product-api返回的'amazonproduct.processors._lxml.SearchPaginator' 更好处理。前者可直接查看，也可用标准xml parser处理；后者因为不是标准的xml文件，用xml.dom.minidom， xml.etree.ElementTree parse时均报错。

连接API所需的credentials

AWS Access Key ID
AWS Secret Key
- documentation
- 在[这个网址]中使用我原有的amazon购物用的邮箱与秘密登陆，然后在section"Your Security Credentials" 中找到 AWS Access Key ID
- As described in a previous announcement, you cannot retrieve the existing secret access keys for your AWS root account, though you can still create a new root access key at any time. 创立新access key id的时候可以看见对应的secret key。
AWS associate tage 即亚马逊联盟的ID，需申请。申请时需先选择所在地区，先申请了中国区的，美国区需电话验证所以目前为申请。 AWS associate tage申请入口： http://docs.aws.amazon.com/AWSECommerceService/latest/DG/becomingAssociate.html

the EditorialReview response group

The EditorialReview response group returns Amazon's review of the item, which appears on the Product Detail page for each item in the response.
Note: Copyrighted editorial reviews are not returned. For this reason, the reviews returned may be different than those returned by Amazon.com. This is a very point, see the example below: Use the following code to retrieve Editorial Reviews for book with id 031606792X: resp2 = amazon.ItemLookup(ItemId='031606792X', ResponseGroup='EditorialReview') Recall that resp2 is a string. looking for tag 'EditorialReviews' gives two editorial reviews: one is the production description itself, the other one is "Amazon.com Review".

now visit page http://www.amazon.com/gp/product/031606792X?ie=UTF8&isInIframe=1&n=283155&redirect=true&ref_=dp_proddesc_0&s=books&showDetailProductDesc=1#iframe-wrapper, we can see that there are three editorial reviews. The first one it "Amazon.com Review", which is the same as the API returns. However the other two editorial reviews "Form Publishers Weekly" and "From School Library Journal" can be seen on the webpage but is not returned by the API. The reason for this is that both editorial reviews end with "All rights reserved." and thus are copyrighted.

So we need to keep it in mind that copyrighted editorial reviews won't be return by the API.
Big difference between EditorialReview and Reviews, as response groups. For RG (abbreviation for Response Group) Reviews, the sole returned value is a URL to an iframe that contains customer reviews. The only way (among all ways I can think of) to view the actual content of reviews is to embed the iframe on a web page by doing <iframe src="reviews_iframe_url" />.

With the iframe URL, I can view the content of reviews in the browser as an iframe, but how can I scrape the text content of the reviews with the iframe URL? I did a lot of web searching but find no result.

Luckily, RG EditorialReviews returns readable texts other than iframe.

Still, I'd very much to learn if the reviews can be scrapped, with the iframe url in hand.

其他

1. Vim

想要在terminal中编辑.amazonXXX，使用text editor vim。关于vim的操作，参考了[这个资料] (http://stackoverflow.com/questions/5764071/how-to-use-vim-in-the-terminal)

You can pass a filename as an option and it will open that file, e.g. vim main.c.
Press "a" to start INSERT mode (in append after cursor mode, which is what you are probably used to).
To leave INSERT mode, press “Esc”.
To save your file, use “:w.” To save and exit, use “:x”.

Note for 12.7

现在觉得python-amazon-product-api返回变量类型难以操纵且缺少参考文档，反而时bottlenose返回文本更易读易操纵。
了解了在api中获得editorial review的方法。
需确认大牛的review不是copyrighted的，否则无法用api获得。

12.9: Email Amazon Product Advertising API

The EditorialReview Response Group does not return some uncopyrighted editorial reviews. I tried using two Python wrappers (python-amazon-product-api and bootlenose) to get editorial reviews of a book, given the book's itemID.

However, for book with ID 0787996491, on Amazon.com, there are editorial reviews from Gary Klein, Tom Peters, etc.. But the API, using EditorialReview as Response Group, did not return any of these reviews, it only returns the product description. The missing editorial reviews don't have "All rights reserved" at the end so seem not copyrighted.

Book with ID 0230613519 has the same issue. Editorial reviews from Dr. Martin Barnes, Scott Tibbitts, etc. can be seen on Amazon.com, are not copyrighted, but are not included in the API's return value.

Could you kindly confirm for me that all these missing editorial reviews are not available via the Product Advertising API, due to some limitations similar as the copyright issue?

Thanks a lot!

OMOOC2py the Book