parse_text_interpolation.py

parse_text_interpolation.py

"""
pip install parsimonious
"""

import dataclasses as _dc
from parsimonious.grammar import Grammar
from parsimonious.nodes import NodeVisitor

TEXT = """
Here are some {variable}s. I don't care about spaces in the {  interpolations}.
More lines...
{and_one_more}
Linebreaks are not ok in the curlies: { 
i_will_not_be_recognized_as_var}
"""

main_grammar = Grammar(
    r"""
    expr          = (interpolation / text)*
    interpolation = LPAR var_usage RPAR
    text          = TEXT / LPAR / RPAR
    var_usage     = space* VAR_NAME space*
    space         = " "
    VAR_NAME      = ~"[a-zA-Z0-9_]+"
    TEXT          = ~"[^{}]+"
    LPAR          = ~"{"
    RPAR          = "}"
    """
)


class MainVisitor(NodeVisitor):
    def visit_expr(self, node, visited_children):
        return Node(children=list(flatten_list(visited_children)))

    def visit_text(self, node, visited_children):
        return TextNode(node.text)

    def visit_interpolation(self, node, visited_children):
        # according to grammar contains exactly one "var_name"
        return InterpolationNode(var_name=[child.name for child in visited_children if isinstance(child, VarName)][0])

    def visit_var_usage(self, node, visited_children):
        # according to grammar contains exactly one "var_name"
        return [child for child in visited_children if isinstance(child, VarName)][0]

    def visit_VAR_NAME(self, node, visited_children):
        return VarName(name=node.text)

    def generic_visit(self, node, visited_children):
        return visited_children or node

def flatten_list(lst):
    for item in lst:
        if isinstance(item, list):
            yield from flatten_list(item)
        else:
            yield item

@_dc.dataclass
class Node:
    children: list

    def flatten(self):
        return [child.flatten() for child in self.children]


@_dc.dataclass
class TextNode:
    text: str

    def flatten(self):
        return self


@_dc.dataclass
class InterpolationNode:
    var_name: str

    def flatten(self):
        return self

@_dc.dataclass
class VarName:
    name: str


tree = main_grammar.parse(TEXT)
out = MainVisitor().visit(tree)
print(out.flatten())
print("Used vars")
print([node.var_name for node in out.flatten() if isinstance(node, InterpolationNode)])

parser.pegjs

/*
Paste the grammar below in the online peggy parser builder: https://peggyjs.org/online.html
And use this as the test text:
Here are some {variable}s. I don't care about spaces in the {  interpolations}.
More lines...
{and_one_more}
Linebreaks are not ok in the curlies: { 
i_will_not_be_recognized_as_var}
*/

expr
  = one_expr *
one_expr
  = interp:interpolation { return interp } 
  / text { return {type:"text", text: text()} }
interpolation
  = LPAR var_use:var_usage RPAR { return var_use }
text
  = TEXT / LPAR / RPAR { return text() }
var_usage
  = space* var_name:VAR_NAME space* { return {type:"var", var: var_name} }
space
  = " "
VAR_NAME
  = [a-zA-Z0-9_] + { return text() }
TEXT
  = [^{}]+
LPAR
  = "{"
RPAR
  = "}"