Python CSV output does not work.

Incidentally, in my environment, click() with driver.find_element_by_xpath() immediately after while True:, exception ended without any processing.
try:...except:break can now be moved to the end of the while True: loop.For your information.

Other major fixes are as follows:

• The main reason is that we open a new 'w' csv file every time in the loop for each article (remove and recreate any files with the same name already).
  I can only do the file containing the data for the last article.
  If you want to open the csv file, do it before the per-article loop for article in articles: and open it with an additional write of 'a' instead of new.

The

• csv write object is csv.writer() in normal simple text, but the write process for each article is not correct because it tries to write dictionary data.
  Use csv.DictWriter().
  If so, use DictWriter.writeheader() to write the header.
  However, if you have opened the file with 'a' additional writes, and data already exists, the header will always be added as the middle row data, so try to get the current location with the f.tell() function when you are not doing anything right after opening the file.

• On a small note, the fourth "published" in the header item name string is misspelled.The correct one is "published".

• In the per-article loop for article inarticles:, initialize records as a list each time, add only one data extracted from the article as dictionary data records.append({...}) and immediately write the list as a csvrd.
  If you want to write to csv and leave it as a list for later processing, it's weird to initialize every time in the loop for each article, and if you write the entire number of articles in csv every time, the same data will be written multiple times.You can write multiple lines (records) using a function with s ending with s.
  Add the data extracted from the current article as dictionary data to the list, stop writing the entire list to csv for each article, create one dictionary data record for each article, and write it to csv with writer.writerow(record)

• The if records: after the per-article loop for article in articles: and the pass processing when it is established are not meaningful.There may be something left to do that doesn't matter what you ask when you write a question article.

• open() the csv file in the per-article loop but f.close() at the end of the loop does not match the processing nesting level.
  The first reason is that open() moves the open() to the front of the article-by-article inarticles:, so the processing level will be fine, but if eof flag: and break are not clear after processing the last article.
  So let's move f.close() before if eof_flag:.

The main reason is that we open a new 'w' csv file every time in the loop for each article (remove and recreate any files with the same name already).
I can only do the file containing the data for the last article.
If you want to open the csv file, do it before the per-article loop for article in articles: and open it with an additional write of 'a' instead of new.

The csv write object is csv.writer() in normal simple text, but the write process for each article is not correct because it tries to write dictionary data.
Use csv.DictWriter().
If so, use DictWriter.writeheader() to write the header.
However, if you have opened the file with 'a' additional writes, and data already exists, the header will always be added as the middle row data, so try to get the current location with the f.tell() function when you are not doing anything right after opening the file.

The fourth "published" in the header item name string is misspelled.The correct one is "published".

records=[] initializes each article with records as a list in the per-article loop for article in articles:, adding only one data extracted from the article as dictionary data, and immediately writing the list as row data in csv.
If you want to write to csv and leave it as a list for later processing, it's weird to initialize every time in the loop for each article, and if you write the entire number of articles in csv every time, the same data will be written multiple times.You can write multiple lines (records) using a function with s ending with s.
Add the data extracted from the current article as dictionary data to the list, stop writing the entire list to csv for each article, create one dictionary data record for each article, and write it to csv with writer.writerow(record)

The if records: after the per-article loop for article in articles: and the pass processing when it is established are not meaningful.There may be something left to do that doesn't matter what you ask when you write a question article.

open() the csv file is open() in the per-article loop, but f.close() at the end of the loop does not match the nested level of the process.
The first reason is that open() moves the open() to the front of the article-by-article inarticles:, so the processing level will be fine, but if eof flag: and break are not clear after processing the last article.
So let's move f.close() before if eof_flag:.