Skip to content

extract: "Deeply nested" gettext calls aren't extracted #1125

Open
@oomsveta

Description

@oomsveta

Overview

#488 enabled the python extractor to look at nested calls for strings to extract, but the way it does it ignores nested gettext calls when there's more than one level of nesting.

This is because scanning sources for nested calls is only done when call_stack is 0:

elif funcname and call_stack == 0:
nested = (tok == NAME and value in keywords)

call_stack starts at -1, and is incremented each time an opening parenthesis is encountered within a gettext call:
elif tok == OP and value == '(':
if in_def:
# Avoid false positives for declarations such as:
# def gettext(arg='message'):
in_def = False
continue
if funcname:
message_lineno = lineno
call_stack += 1

In the case of a "deeply nested" call (a call with more than one level of nesting), the value of call_stack will be greater than 0, meaning that the part of the code that handles the nested call will never be reached, and the contents of that call won't be exported.

Steps to Reproduce

Copy-paste this to your python REPL:

from io import BytesIO
from babel.messages.extract import extract

file = b"""_(
    "dummy",
    nested_call(
        _("deeply nested call")
    )
)"""

list(extract("python", BytesIO(file)))  # -> [(4, 'dummy', [], None)]

Actual Results

[(4, 'dummy', [], None)]

"deeply nested call" is missing.
Note: the lineno of "dummy" is also wrong, see this other issue.

Expected Results

[(1, 'dummy', [], None), (4, 'deeply nested call', [], None)]

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions