In the second tutorial in this two-part series about terminal programs with great command-line interfaces, we'll discuss Prompt, Toolkit, Click, Pygments, and Fuzzy Finder.

This is the second tutorial in my two-part series about terminal programs with great command line interfaces. In the first article, we discussed some of the features that make command line applications a joy to use. In this second article, we’ll look at how to implement these features using some of Python’s libraries.

I plan to implement it in less than 20 lines of Python code. let's start.

I am accustomed to calling this library the Swiss Army Knife of command line applications. It can be used as a replacement for readline, curses, etc. Let's first install this library and then start the tutorial:

pip install prompt_toolkit

We start with a simple REPL (LCTT Annotation: REPL - Read-Eval-Print Loop, interactive development environment). A typical REPL receives user input, performs an operation, and outputs the result. For example, in our example, we are going to implement a REPL with "echo" functionality. It simply prints out the user's input as is:

from prompt_toolkit import prompt

while 1:
    user_input = prompt('>')
    print(user_input)

This is all the code to implement REPL. It reads the user's input and then prints out the user's input. The prompt function used in this code comes from the prompt_toolkit library, which is a replacement for the readline library.

To enhance the functionality of our REPL, we can add command history:

from prompt_toolkit import prompt
from prompt_toolkit.history import FileHistory

while 1:
    user_input = prompt('>', 
                        history=FileHistory('history.txt'),
                       )
    print(user_input)

We just added a persistent command history to the REPL. Now we can use the up/down arrows to browse the command history and Ctrl-R to search the command history. It meets the basic guidelines of the command line.

In the first tutorial, one of the discoverability techniques I talked about was automatically recommending historical commands. (I first saw this feature in fish shell) Let's add this feature to our REPL:

from prompt_toolkit import prompt
from prompt_toolkit.history import FileHistory
from prompt_toolkit.auto_suggest import AutoSuggestFromHistory

while 1:
    user_input = prompt('>', 
                        history=FileHistory('history.txt'),
                        auto_suggest=AutoSuggestFromHistory(),
                       )
    print(user_input)

We just need to add a new parameter to the prompt() API call. Now we have a fish shell-style REPL that automatically recommends historical commands.

Now, let’s enhance tab completion with autocompletion. It pops up possible command recommendations as the user starts typing.

REPL How to make recommendations? We use a dictionary to recommend possible items.

For example, let’s implement a REPL for SQL. We can store SQL keywords in the auto-complete dictionary. Let's see how this is done:

from prompt_toolkit import prompt
from prompt_toolkit.history import FileHistory
from prompt_toolkit.auto_suggest import AutoSuggestFromHistory
from prompt_toolkit.contrib.completers import WordCompleter

SQLCompleter = WordCompleter(['select', 'from', 'insert', 'update', 'delete', 'drop'],
                             ignore_case=True)

while 1:
    user_input = prompt('SQL>', 
                        history=FileHistory('history.txt'),
                        auto_suggest=AutoSuggestFromHistory(),
                        completer=SQLCompleter,
                        )
    print(user_input)

Again, we simply used prompt-toolkit's built-in completion feature called WordCompleter, which matches user input with possible recommended dictionaries. , and then provide a list.

Now, we have a REPL with auto-completion, fish shell-style historical command recommendations, and up/down browsing history. Implementing these features took less than 10 lines of actual code.

Click is a command line creation toolkit that makes it easier to parse command line option parameters and constants for programs. We will not discuss how to use Click as a parameter parser here. Instead, we'll look at some of the features that Click comes with.

Install Click:

pip install click

Paginators are utilities on Unix systems that are capable of displaying very long output one page at a time. Some examples of pagers include less, more, most, etc. Displaying the output of a command through a pager is not only a friendly design, but also a necessity.

讓我們進(jìn)一步改進(jìn)前面的例子。我們不再使用默認(rèn)print()語句，取而代之的是click.echo_via_pager()。它將會(huì)把輸出通過分頁器發(fā)送到標(biāo)準(zhǔn)輸出。這是平臺(tái)無關(guān)的，因此在 Unix 系統(tǒng)或 Windows 系統(tǒng)上均能工作。如果必要的話，click_via_pager 會(huì)嘗試使用一個(gè)合適的默認(rèn)分頁器來輸出，從而能夠顯示代碼高亮。

from prompt_toolkit import prompt
from prompt_toolkit.history import FileHistory
from prompt_toolkit.auto_suggest import AutoSuggestFromHistory
from prompt_toolkit.contrib.completers import WordCompleter
import click

SQLCompleter = WordCompleter(['select', 'from', 'insert', 'update', 'delete', 'drop'],
                             ignore_case=True)

while 1:
    user_input = prompt(u'SQL>',
                        history=FileHistory('history.txt'),
                        auto_suggest=AutoSuggestFromHistory(),
                        completer=SQLCompleter,
                        )
    click.echo_via_pager(user_input)

在我前面的文章中一個(gè)值得一提的細(xì)節(jié)是，當(dāng)命令過于復(fù)雜的時(shí)候進(jìn)入編輯器來編輯。Click 有一個(gè)簡單的 API 能夠打開編輯器，然后把在編輯器中輸入的文本返回給應(yīng)用。

import click
message = click.edit()

Fuzzy Finder 是一種通過少量輸入來為用戶減少推薦的方法。幸運(yùn)的是，有一個(gè)庫可以實(shí)現(xiàn) Fuzzy Finder 。讓我們首先安裝這個(gè)庫：

pip install fuzzyfinder

Fuzzy Finder的 API 很簡單。用戶向它傳遞部分字符串和一系列可能的選擇，然后，F(xiàn)uzzy Finder將會(huì)返回一個(gè)與部分字符串匹配的列表，這一列表是通過模糊算法根據(jù)相關(guān)性排序得出的。比如：

>>> from fuzzyfinder import fuzzyfinder

>>> suggestions = fuzzyfinder('abc', ['abcd', 'defabca', 'aagbec', 'xyz', 'qux'])

>>> list(suggestions)
['abcd', 'defabca', 'aagbec']

現(xiàn)在我們有了fuzzyfinder>，讓我們把它加入到我們的 SQL REPL 中。方法是我們自定義一個(gè)completer而不是使用來自prompt-toolkit庫的WordCompleter。比如：

from prompt_toolkit import prompt
from prompt_toolkit.history import FileHistory
from prompt_toolkit.auto_suggest import AutoSuggestFromHistory
from prompt_toolkit.completion import Completer, Completion
import click
from fuzzyfinder import fuzzyfinder

SQLKeywords = ['select', 'from', 'insert', 'update', 'delete', 'drop']

class SQLCompleter(Completer):
    def get_completions(self, document, complete_event):
        word_before_cursor = document.get_word_before_cursor(WORD=True)
        matches = fuzzyfinder(word_before_cursor, SQLKeywords)
        for m in matches:
            yield Completion(m, start_position=-len(word_before_cursor))

while 1:
    user_input = prompt(u'SQL>',
                        history=FileHistory('history.txt'),
                        auto_suggest=AutoSuggestFromHistory(),
                        completer=SQLCompleter(),
                        )
    click.echo_via_pager(user_input)

現(xiàn)在，讓我們給用戶輸入添加語法高亮。我們正在搭建一個(gè) SQL REPL，如果具有彩色高亮的 SQL 語句，這會(huì)很棒。

Pygments是一個(gè)提供語法高亮的庫，內(nèi)建支持超過 300 種語言。添加語法高亮能夠使應(yīng)用變得彩色化，從而能夠幫助用戶在執(zhí)行程序前發(fā)現(xiàn) SQL 中存在的錯(cuò)誤，比如拼寫錯(cuò)誤、引號(hào)不匹配或括號(hào)不匹配。

首先，安裝Pygments：

pip install pygments

讓我們使用Pygments來為 SQL REPL 添加顏色：

from prompt_toolkit import prompt
from prompt_toolkit.history import FileHistory
from prompt_toolkit.auto_suggest import AutoSuggestFromHistory
from prompt_toolkit.completion import Completer, Completion
import click
from fuzzyfinder import fuzzyfinder
from pygments.lexers.sql import SqlLexer

SQLKeywords = ['select', 'from', 'insert', 'update', 'delete', 'drop']

class SQLCompleter(Completer):
    def get_completions(self, document, complete_event):
        word_before_cursor = document.get_word_before_cursor(WORD=True)
        matches = fuzzyfinder(word_before_cursor, SQLKeywords)
        for m in matches:
            yield Completion(m, start_position=-len(word_before_cursor))

while 1:
    user_input = prompt(u'SQL>',
                        history=FileHistory('history.txt'),
                        auto_suggest=AutoSuggestFromHistory(),
                        completer=SQLCompleter(),
                        lexer=SqlLexer,
                        )
    click.echo_via_pager(user_input)

Prompt Toolkit能夠和Pygments一同很好的工作。我們把Pygments提供的SqlLexer加入到來自prompt-toolkit的prompt中?，F(xiàn)在，所有的用戶輸入都會(huì)被當(dāng)作 SQL 語句，并進(jìn)行適當(dāng)著色。

我們的“旅途”通過創(chuàng)建一個(gè)強(qiáng)大的 REPL 結(jié)束，這個(gè) REPL 具有常見的 shell 的全部特性，比如歷史命令，鍵位綁定，用戶友好性比如自動(dòng)補(bǔ)全、模糊查找、分頁器支持、編輯器支持和語法高亮。我們僅用少于 20 行 Python 代碼就實(shí)現(xiàn)了這個(gè) REPL 。

不是很簡單嗎？現(xiàn)在，你沒有理由不會(huì)寫一個(gè)自己的命令行應(yīng)用了。

The above is the detailed content of A library for building high-quality Python command-line user interfaces. For more information, please follow other related articles on the PHP Chinese website!