Regular Expressions 101

Community Patterns

Community Library Entry

0

Regular Expression
Python

r"
(?P<field_with_params>[^{ }]+(?P<parenth_open>\()(?P<keyword>[^\(\s]\w+)\s?(?P<colon>\:)\s?(?P<value>(?P<list_open>\[?).+?(?P<list_close>\]?))\s?(?P<parenth_close>\)))|(?P<BEGIN>^\{)|(?P<open>\{)|(?P<END>\}$)|(?P<close>\})|(?P<fragment>(?P<dots>\.{3})\s*?(?P<on>on)\s+?(?P<frag_field>\w+))|(?P<parent_field>(?!(?P=field_with_params)|(?P=fragment))\w+)(?=\s*?\{)|(?P<child_field>(?!(?P=field_with_params)|(?P=fragment))\w+)
"
gm

Description

Parse Graphql queries

Changelog (versions):

  1. Initial
  2. Version 2 is much improved and intended to integrate well with customized highlighting with the help of the awesome Rich library.
  3. Bug fixes and added rich example.
  4. Bugfixes and now the regex is actually recognizing the different parts of field_with_params. And an improved example to go with that.

Todo (maybe):

  • Improve the beginning, opening, closing and ending curly bracket recognition with better support for malformed queries and make sure there's only one BEGIN and one END in situations where newlines are present.
  • Also differentiate between arbitrary text and actual graphql query language content. For example, the string "asdf" is currently recognized and categorized as a child_field when it shouldn't be recognized at all.

Version 4

def pretty_query(query_string: str) -> str:
    import re

    from rich.console import Console
    from rich.highlighter import JSONHighlighter
    from rich.theme import Theme, DEFAULT_STYLES

    # this is a string, not a tuple
    field_with_params_rgx = (
        r'(?P<field_with_params>[^{ }]+' # fields with parameters
        r'(?P<parenth_open>\()'          # opening parenthesis
        r'(?P<keyword>[^\(\s]\w+)\s?'    # keywords and colons
        r'(?P<colon>\:)\s?'              # keywords and colons
        r'(?P<value>'                    # parameter values
        r'(?P<list_open>\[?).+?'         # opening square bracket
        r'(?P<list_close>\]?))\s?'       # closing square bracket
        r'(?P<parenth_close>\)))'        # closing parenthesis
    )

    regexes = [
        field_with_params_rgx,
        r'(?P<BEGIN>^\{)|(?P<open>\{)|(?P<END>\}$)|(?P<close>\})',                  # curly brackets
        r'(?P<fragment>(?P<dots>\.{3})\s*?(?P<on>on)\s+?(?P<frag_field>\w+))',      # fragments
        r'(?P<parent_field>(?!(?P=field_with_params)|(?P=fragment))\w+)(?=\s*?\{)', # parent fields
        r'(?P<child_field>(?!(?P=field_with_params)|(?P=fragment))\w+)',            # child fields
    ]

    indent = 2
    count = 0
    parts = []

    for x in re.finditer('|'.join(rgx for rgx in regexes), query_string):
        data = x.groupdict()
        string = ''

        BEGIN = data['BEGIN']
        open = data['open']
        parent_field = data['field_with_params'] or data['parent_field']
        child_field = data['child_field']
        fragment = data['fragment']
        close = data['close']
        END = data['END']


        if BEGIN:
            string += BEGIN
            count += indent

        if open:
            count += indent
            continue

        if parent_field:
            string += '{0: >{fill}}{value} {{'.format(' ', value=parent_field, fill=count)

        if child_field:
            string += '{0: >{fill}}{value}'.format(' ', value=child_field, fill=count)

        if fragment:
            string += '{0: >{fill}}{value} {{'.format(' ', value=fragment, fill=count)

        if close:
            count -= indent
            string += '{0: >{fill}}{value}'.format(' ', value=close, fill=count)

        if END:
            string += END
            count -= indent
        parts.append(string)

    class GQLHighlighter(JSONHighlighter):
        base_style = "gql."
        highlights = ['|'.join(rgx for rgx in regexes)] + JSONHighlighter.highlights

    theme = Theme({
        **{
            f'gql.{k}': DEFAULT_STYLES[f'json.{k}']
            for k in ['brace', 'bool_true', 'bool_false', 'null', 'number', 'str', 'key']
        },
        'gql.BEGIN': 'bold green',
        'gql.END': 'bold green',
        'gql.open': 'bold yellow',
        'gql.close': 'bold blue',
        'gql.field_with_params': 'bold blue',
        'gql.parent_field': 'bold white',
        'gql.child_field': 'italic green',
        'gql.fragment': 'bold yellow',
        'gql.dots': 'bold white',
        'gql.on': 'green',
        'gql.frag_field': 'bold magenta italic',
        'gql.parenth_open': 'blue',
        'gql.parenth_close': 'green',
        'gql.keyword': 'yellow',
        'gql.colon': 'blue',
        'gql.list_open': 'bold yellow',
        'gql.list_close': 'bold yellow',
        'gql.value': 'cyan',
    })

    console = Console(highlighter=GQLHighlighter(), theme=theme)

    console.print('\n'.join(parts))

    return '\n'.join(parts)
Submitted by iwconfig - 2 years ago