My first question

[Sead Sejo Gicić]

What is the difference between ordinary txt file and csv file for using in Python programming?

[Liu Zheng]

Hi. Doesn't sound like a Django question, but I assume you came across this question when writing Django codes:) So anyway, let me try to see whether I could provide any useful information here.

This question is a bit too broad, so let me try to answer it in several levels.

First, python doesn't care the format of any file. From python's point of view, any file is just a sequence of binary data. On the lowest level, python only two types of operations: read N bytes from the file starting from a position; write N bytes to the file starting from a position. That's it. What each byte mean is up to the programmers. So on the level of python programming, there's no difference between a txt file and a csv file.

Having said the first point, there are standard formats defined by programmers, to make communication easier. First, among all binary files, there's special type of files called "text files". It's special because every byte of a text file must be a value from a certain table of symbols (think of the ASCII table). Not all files are text files. But if a file is a text file, then most text editors can handle that file well. (Python also have native support on the language level, that is, you open a file in the text mode if you use "open(<file name>, 'r')" intead of "open(<file name>, 'rb')". However, this is only for sake of convenience of human readability. It doesn't make a text file fundamentally different from other files. In fact, you can open a text file in the binary mode using 'rb'. Python will simply forget that the file is a text file.

Next, text files still form a big family. By assigning meanings to different symbols and creating rules of how the content should look like, there are more types under text files. csv is a special type of text file. (Example of othe types are .py, .html, etc). Since it's just a text file, you can open a csv file using "open(<csv file name>, 'r')" and do read/write operations like any other text file. However, since it has its own rules, there are also some helper libraries to assist you handling csv files, too. Examples are "csv" package, "pandas" package, etc. But again, since it's a text file, you can ignore those packages and write your codes using string functions directly. For example:

```

s = open("sample.csv", 'r').read()

lines = s.split('\n')

first_line = lines[0]

columns_lables = first_line.split(',')

......

```

Such a piece of codes will handle a very simple and standard csv file already. However, csv has more complications than this. It even has different variations (called dialets). If treating a csv file as an ordinary text file, you are basically re-inventing the wheel. (There's nothing wrong with that. Just tedious.) Those standard packages can handle those staff and let you focus on your business logic. That's all.

In the end, I want to emphasize that files' extension names doesn't mean anything. When I said a csv file above, I don't mean a file whose file name ends with ".csv". I mean a file whose content complies the csv rules. By creating a file called "something.csv", I'm merely giving a promise to anyone who open it that "I promise you that the content in this file satisfies the csv rules. You have my words!" Whether I keep my promise or not is totally up to me (indeed, I can rename an image file "pic1.jpg" to "pic1.csv".) In addition, you can save some csv data into a .txt files. For example, consider the following data:

Name,Job Title,Phone Number

Alice,Manager,23121212

Bob,Director,12345678

Chales,Developer,41212113

You can store the data into "contacts.txt". I will still call it a csv file even though it has ".txt" extension. Therefore, I can handle this file using any csv package. For example:

```

import pandas as pd

df = pd.read_csv("contacts.txt")

...

```

Pandas will not complain just because its extension is not ".csv". However, it will be a good practice to choose the extension which best describes the nature of the file's contents. In this way, you'll bring less confusion to the users of this file (including the future yourself).

[Sead Sejo Gicić]

Thanks very much

--
You received this message because you are subscribed to the Google Groups "Django users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to django-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/django-users/7a9b233b-a61b-4866-8383-44b4b60ac736o%40googlegroups.com.

[Derek]

This would make a great blog post!  And in future, we can all refer directly to it, when someone asks a similar question.

[o1bigtenor]

Appreciate very much your informative post. Question below - - - 

On Sat, Jul 18, 2020 at 9:14 AM Liu Zheng <firstd...@gmail.com> wrote:

Hi. Doesn't sound like a Django question, but I assume you came across this question when writing Django codes:) So anyway, let me try to see whether I could provide any useful information here.

This question is a bit too broad, so let me try to answer it in several levels.

First, python doesn't care the format of any file. From python's point of view, any file is just a sequence of binary data. On the lowest level, python only two types of operations: read N bytes from the file starting from a position; write N bytes to the file starting from a position. That's it. What each byte mean is up to the programmers. So on the level of python programming, there's no difference between a txt file and a csv file.

Having said the first point, there are standard formats defined by programmers, to make communication easier. First, among all binary files, there's special type of files called "text files". It's special because every byte of a text file must be a value from a certain table of symbols (think of the ASCII table). Not all files are text files. But if a file is a text file, then most text editors can handle that file well. (Python also have native support on the language level, that is, you open a file in the text mode if you use "open(<file name>, 'r')" intead of "open(<file name>, 'rb')". However, this is only for sake of convenience of human readability. It doesn't make a text file fundamentally different from other files. In fact, you can open a text file in the binary mode using 'rb'. Python will simply forget that the file is a text file.

Next, text files still form a big family. By assigning meanings to different symbols and creating rules of how the content should look like, there are more types under text files. csv is a special type of text file. (Example of othe types are .py, .html, etc). Since it's just a text file, you can open a csv file using "open(<csv file name>, 'r')" and do read/write operations like any other text file. However, since it has its own rules, there are also some helper libraries to assist you handling csv files, too. Examples are "csv" package, "pandas" package, etc. But again, since it's a text file, you can ignore those packages and write your codes using string functions directly. For example:

```

s = open("sample.csv", 'r').read()

lines = s.split('\n')

first_line = lines[0]

columns_lables = first_line.split(',')

......

```

Such a piece of codes will handle a very simple and standard csv file already. However, csv has more complications than this. It even has different variations (called dialets).

Did you mean dialects instead of dialets by any chance?

 

If treating a csv file as an ordinary text file, you are basically re-inventing the wheel. (There's nothing wrong with that. Just tedious.) Those standard packages can handle those staff and let you focus on your business logic. That's all.

In the end, I want to emphasize that files' extension names doesn't mean anything. When I said a csv file above, I don't mean a file whose file name ends with ".csv". I mean a file whose content complies the csv rules. By creating a file called "something.csv", I'm merely giving a promise to anyone who open it that "I promise you that the content in this file satisfies the csv rules. You have my words!" Whether I keep my promise or not is totally up to me (indeed, I can rename an image file "pic1.jpg" to "pic1.csv".) In addition, you can save some csv data into a .txt files. For example, consider the following data:

Name,Job Title,Phone Number

Alice,Manager,23121212

Bob,Director,12345678

Chales,Developer,41212113

You can store the data into "contacts.txt". I will still call it a csv file even though it has ".txt" extension. Therefore, I can handle this file using any csv package. For example:

```

import pandas as pd

df = pd.read_csv("contacts.txt")

...

```

Pandas will not complain just because its extension is not ".csv". However, it will be a good practice to choose the extension which best describes the nature of the file's contents. In this way, you'll bring less confusion to the users of this file (including the future yourself).

On Saturday, July 18, 2020 at 2:25:57 PM UTC+8, Sead Sejo Gicić wrote:

What is the difference between ordinary txt file and csv file for using in Python programming?

--

[Liu Zheng]

Hi Derek. I just copy pasted the answer to my blog (haven't written any blog in recent months). Will improve it into a readable blog post with better presentaion. Thanks for suggesting this. Acknowledgement was given in the post, but let me know if you are not comfortable of being mentioned :)

[Liu Zheng]

Hi. Yes. You are absolutely right. I'm new to google groups (in fact, this was my first post). I thought I could edit it after posting it out. Was a bit in a hurry when writing it down. :O

Also a bit wordy...

To unsubscribe from this group and stop receiving emails from it, send an email to django...@googlegroups.com.

[George]

Consider reading this article on how to read csv files in Python.

[firstd...@gmail.com]

Re-formatted the answer and it's now available at https://medium.com/@liuzheng900204/as-a-programmer-you-need-to-understand-these-about-files-b542c60d9285?sk=00c805001cd05feb7d501f6e7e1bbd04

[sakshi jain]

questions bank..tu phele mere swal ka jwab de

--

You received this message because you are subscribed to the Google Groups "Django users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to django-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/django-users/58184c88-fe45-4d3e-b75b-c77cf88a4719n%40googlegroups.com.