mirror of
https://gitlab.gnome.org/jwestman/blueprint-compiler.git
synced 2025-05-04 15:59:08 -04:00
Tiny formatter improvements
This commit is contained in:
gregorni
James Westman
parent
e261180dcc
commit
e5cde71fc1
4 changed files with 155 additions and 166 deletions
|
|
@ -44,192 +44,183 @@ class LineType(Enum):
|
||||||
COMMENT = 4
|
COMMENT = 4
|
||||||
|
|
||||||
|
|
||||||
class Formatter:
|
def format(data, tab_size=2, insert_space=True):
|
||||||
def format(data, tab_size=2, insert_space=True):
|
indent_levels = 0
|
||||||
indent_levels = 0
|
tokens = tokenizer.tokenize(data)
|
||||||
tokens = tokenizer.tokenize(data)
|
end_str = ""
|
||||||
end_str = ""
|
last_not_whitespace = tokens[0]
|
||||||
last_not_whitespace = tokens[0]
|
current_line = ""
|
||||||
|
prev_line_type = None
|
||||||
|
is_child_type = False
|
||||||
|
indent_item = " " * tab_size if insert_space else "\t"
|
||||||
|
watch_parentheses = False
|
||||||
|
parentheses_balance = 0
|
||||||
|
bracket_tracker = [None]
|
||||||
|
last_whitespace_contains_newline = False
|
||||||
|
|
||||||
|
def commit_current_line(
|
||||||
|
line_type=prev_line_type, redo_whitespace=False, newlines_before=1
|
||||||
|
):
|
||||||
|
nonlocal end_str, current_line, prev_line_type
|
||||||
|
|
||||||
|
indent_whitespace = indent_levels * indent_item
|
||||||
|
whitespace_to_add = "\n" + indent_whitespace
|
||||||
|
|
||||||
|
if redo_whitespace or newlines_before != 1:
|
||||||
|
end_str = end_str.strip() + "\n" * newlines_before
|
||||||
|
if newlines_before > 0:
|
||||||
|
end_str += indent_whitespace
|
||||||
|
|
||||||
|
end_str += current_line + whitespace_to_add
|
||||||
|
|
||||||
current_line = ""
|
current_line = ""
|
||||||
prev_line_type = None
|
prev_line_type = line_type
|
||||||
is_child_type = False
|
|
||||||
indent_item = " " * tab_size if insert_space else "\t"
|
|
||||||
watch_parentheses = False
|
|
||||||
parentheses_balance = 0
|
|
||||||
bracket_tracker = [None]
|
|
||||||
last_whitespace_contains_newline = False
|
|
||||||
|
|
||||||
def commit_current_line(
|
for item in tokens:
|
||||||
line_type=prev_line_type, redo_whitespace=False, newlines_before=1
|
str_item = str(item)
|
||||||
):
|
|
||||||
nonlocal end_str, current_line, prev_line_type
|
|
||||||
|
|
||||||
indent_whitespace = indent_levels * indent_item
|
if item.type == TokenType.WHITESPACE:
|
||||||
whitespace_to_add = "\n" + indent_whitespace
|
last_whitespace_contains_newline = "\n" in str_item
|
||||||
|
continue
|
||||||
|
|
||||||
if redo_whitespace or newlines_before != 1:
|
whitespace_required = (
|
||||||
end_str = end_str.strip() + "\n" * newlines_before
|
str_item in WHITESPACE_BEFORE
|
||||||
if newlines_before > 0:
|
or str(last_not_whitespace) in WHITESPACE_AFTER
|
||||||
end_str += indent_whitespace
|
or (str_item == "(" and end_str.endswith(": bind"))
|
||||||
|
)
|
||||||
|
whitespace_blockers = (
|
||||||
|
str_item in NO_WHITESPACE_BEFORE
|
||||||
|
or str(last_not_whitespace) in NO_WHITESPACE_AFTER
|
||||||
|
or (str_item == "<" and str(last_not_whitespace) == "typeof")
|
||||||
|
)
|
||||||
|
|
||||||
end_str += current_line + whitespace_to_add
|
this_or_last_is_ident = TokenType.IDENT in (item.type, last_not_whitespace.type)
|
||||||
|
current_line_is_empty = len(current_line) == 0
|
||||||
|
is_function = str_item == "(" and not re.match(
|
||||||
|
r"^([A-Za-z_\-])+(: bind)?$", current_line
|
||||||
|
)
|
||||||
|
|
||||||
current_line = ""
|
any_blockers = whitespace_blockers or current_line_is_empty or is_function
|
||||||
prev_line_type = line_type
|
if (whitespace_required or this_or_last_is_ident) and not any_blockers:
|
||||||
|
current_line += " "
|
||||||
|
|
||||||
for item in tokens:
|
current_line += str_item
|
||||||
str_item = str(item)
|
|
||||||
|
|
||||||
if item.type == TokenType.WHITESPACE:
|
if str_item in ("[", "("):
|
||||||
last_whitespace_contains_newline = "\n" in str_item
|
bracket_tracker.append(str_item)
|
||||||
continue
|
elif str_item in ("]", ")"):
|
||||||
|
bracket_tracker.pop()
|
||||||
|
|
||||||
whitespace_required = (
|
needs_newline_treatment = (
|
||||||
str_item in WHITESPACE_BEFORE
|
str_item in NEWLINE_AFTER or item.type == TokenType.COMMENT
|
||||||
or str(last_not_whitespace) in WHITESPACE_AFTER
|
)
|
||||||
or (str_item == "(" and end_str.endswith(": bind"))
|
if needs_newline_treatment:
|
||||||
)
|
if str_item in OPENING_TOKENS:
|
||||||
whitespace_blockers = (
|
list_or_child_type = str_item == "["
|
||||||
str_item in NO_WHITESPACE_BEFORE
|
if list_or_child_type:
|
||||||
or str(last_not_whitespace) in NO_WHITESPACE_AFTER
|
is_child_type = current_line.startswith("[")
|
||||||
or (str_item == "<" and str(last_not_whitespace) == "typeof")
|
|
||||||
)
|
|
||||||
|
|
||||||
this_or_last_is_ident = (
|
if is_child_type:
|
||||||
item.type == TokenType.IDENT
|
if str(last_not_whitespace) not in OPENING_TOKENS:
|
||||||
or last_not_whitespace.type == TokenType.IDENT
|
end_str = (
|
||||||
)
|
end_str.strip() + "\n\n" + (indent_item * indent_levels)
|
||||||
current_line_is_empty = len(current_line) == 0
|
)
|
||||||
is_function = str_item == "(" and not re.match(
|
last_not_whitespace = item
|
||||||
r"^([A-Za-z_\-])+(: bind)?$", current_line
|
continue
|
||||||
)
|
|
||||||
|
|
||||||
any_blockers = whitespace_blockers or current_line_is_empty or is_function
|
indent_levels += 1
|
||||||
|
keep_same_indent = prev_line_type not in (
|
||||||
|
LineType.CHILD_TYPE,
|
||||||
|
LineType.COMMENT,
|
||||||
|
LineType.BLOCK_OPEN,
|
||||||
|
)
|
||||||
|
if keep_same_indent:
|
||||||
|
end_str = (
|
||||||
|
end_str.strip() + "\n\n" + indent_item * (indent_levels - 1)
|
||||||
|
)
|
||||||
|
commit_current_line(LineType.BLOCK_OPEN)
|
||||||
|
|
||||||
if (whitespace_required or this_or_last_is_ident) and not any_blockers:
|
elif str_item == "]" and is_child_type:
|
||||||
current_line += " "
|
commit_current_line(LineType.CHILD_TYPE, False)
|
||||||
|
is_child_type = False
|
||||||
|
|
||||||
current_line += str_item
|
elif str_item in CLOSING_TOKENS:
|
||||||
|
if str_item == "]" and last_not_whitespace != ",":
|
||||||
|
current_line = current_line[:-1]
|
||||||
|
commit_current_line()
|
||||||
|
current_line = "]"
|
||||||
|
elif str(last_not_whitespace) in OPENING_TOKENS:
|
||||||
|
end_str = end_str.strip()
|
||||||
|
commit_current_line(LineType.BLOCK_CLOSE, True, 0)
|
||||||
|
|
||||||
if str_item in ["[", "("]:
|
indent_levels -= 1
|
||||||
bracket_tracker.append(str_item)
|
commit_current_line(LineType.BLOCK_CLOSE, True)
|
||||||
elif str_item in ["]", ")"]:
|
|
||||||
bracket_tracker.pop()
|
|
||||||
|
|
||||||
needs_newline_treatment = (
|
elif str_item == ";":
|
||||||
str_item in NEWLINE_AFTER or item.type == TokenType.COMMENT
|
line_type = LineType.STATEMENT
|
||||||
)
|
newlines = 1
|
||||||
if needs_newline_treatment:
|
|
||||||
if str_item in OPENING_TOKENS:
|
|
||||||
list_or_child_type = str_item == "["
|
|
||||||
if list_or_child_type:
|
|
||||||
is_child_type = current_line.startswith("[")
|
|
||||||
|
|
||||||
if is_child_type:
|
if len(current_line) == 1:
|
||||||
if str(last_not_whitespace) not in OPENING_TOKENS:
|
newlines = 0
|
||||||
end_str = (
|
line_type = LineType.BLOCK_CLOSE
|
||||||
end_str.strip()
|
elif prev_line_type == LineType.BLOCK_CLOSE:
|
||||||
+ "\n\n"
|
newlines = 2
|
||||||
+ (indent_item * indent_levels)
|
|
||||||
)
|
|
||||||
last_not_whitespace = item
|
|
||||||
continue
|
|
||||||
|
|
||||||
indent_levels += 1
|
commit_current_line(line_type, newlines_before=newlines)
|
||||||
keep_same_indent = not prev_line_type in [
|
|
||||||
LineType.CHILD_TYPE,
|
|
||||||
LineType.COMMENT,
|
|
||||||
LineType.BLOCK_OPEN,
|
|
||||||
]
|
|
||||||
if keep_same_indent:
|
|
||||||
end_str = (
|
|
||||||
end_str.strip() + "\n\n" + indent_item * (indent_levels - 1)
|
|
||||||
)
|
|
||||||
commit_current_line(LineType.BLOCK_OPEN)
|
|
||||||
|
|
||||||
elif str_item == "]" and is_child_type:
|
elif item.type == TokenType.COMMENT:
|
||||||
commit_current_line(LineType.CHILD_TYPE, False)
|
require_extra_newline = (
|
||||||
is_child_type = False
|
LineType.BLOCK_CLOSE,
|
||||||
|
LineType.STATEMENT,
|
||||||
|
LineType.COMMENT,
|
||||||
|
)
|
||||||
|
|
||||||
elif str_item in CLOSING_TOKENS:
|
single_line_comment = str_item.startswith("//")
|
||||||
if str_item == "]" and last_not_whitespace != ",":
|
newlines = 1
|
||||||
current_line = current_line[:-1]
|
if single_line_comment:
|
||||||
commit_current_line()
|
if not str_item.startswith("// "):
|
||||||
current_line = "]"
|
current_line = f"// {current_line[2:]}"
|
||||||
elif str(last_not_whitespace) in OPENING_TOKENS:
|
|
||||||
end_str = end_str.strip()
|
|
||||||
commit_current_line(LineType.BLOCK_CLOSE, True, 0)
|
|
||||||
|
|
||||||
indent_levels -= 1
|
if not last_whitespace_contains_newline:
|
||||||
commit_current_line(LineType.BLOCK_CLOSE, True)
|
current_line = " " + current_line
|
||||||
|
|
||||||
elif str_item == ";":
|
|
||||||
line_type = LineType.STATEMENT
|
|
||||||
if len(current_line) == 1:
|
|
||||||
newlines = 0
|
newlines = 0
|
||||||
line_type = LineType.BLOCK_CLOSE
|
|
||||||
elif prev_line_type == LineType.BLOCK_CLOSE:
|
elif prev_line_type == LineType.BLOCK_CLOSE:
|
||||||
newlines = 2
|
newlines = 2
|
||||||
else:
|
|
||||||
newlines = 1
|
|
||||||
|
|
||||||
commit_current_line(line_type, newlines_before=newlines)
|
elif prev_line_type in require_extra_newline:
|
||||||
|
newlines = 2
|
||||||
|
|
||||||
elif item.type == TokenType.COMMENT:
|
commit_current_line(LineType.COMMENT, newlines_before=newlines)
|
||||||
require_extra_newline = [
|
|
||||||
LineType.BLOCK_CLOSE,
|
|
||||||
LineType.STATEMENT,
|
|
||||||
LineType.COMMENT,
|
|
||||||
]
|
|
||||||
|
|
||||||
single_line_comment = str_item.startswith("//")
|
else:
|
||||||
newlines = 1
|
commit_current_line()
|
||||||
if single_line_comment:
|
|
||||||
if not str_item.startswith("// "):
|
|
||||||
current_line = f"// {current_line[2:]}"
|
|
||||||
|
|
||||||
if not last_whitespace_contains_newline:
|
elif str_item == "(" and (
|
||||||
current_line = " " + current_line
|
re.match(r"^([A-Za-z_\-])+\s*\(", current_line) or watch_parentheses
|
||||||
newlines = 0
|
):
|
||||||
elif prev_line_type == LineType.BLOCK_CLOSE:
|
watch_parentheses = True
|
||||||
newlines = 2
|
parentheses_balance += 1
|
||||||
|
|
||||||
elif prev_line_type in require_extra_newline:
|
elif str_item == ")" and watch_parentheses:
|
||||||
newlines = 2
|
parentheses_balance -= 1
|
||||||
|
all_parentheses_closed = parentheses_balance == 0
|
||||||
|
if all_parentheses_closed:
|
||||||
|
commit_current_line(
|
||||||
|
newlines_before=2 if prev_line_type == LineType.BLOCK_CLOSE else 1
|
||||||
|
)
|
||||||
|
watch_parentheses = False
|
||||||
|
|
||||||
commit_current_line(LineType.COMMENT, newlines_before=newlines)
|
tracker_is_empty = len(bracket_tracker) > 0
|
||||||
|
if tracker_is_empty:
|
||||||
|
last_in_tracker = bracket_tracker[-1]
|
||||||
|
is_list_comma = last_in_tracker == "[" and str_item == ","
|
||||||
|
if is_list_comma:
|
||||||
|
last_was_list_item = end_str.strip()[-1] not in ("[", ",")
|
||||||
|
if last_was_list_item:
|
||||||
|
end_str = end_str.strip()
|
||||||
|
commit_current_line()
|
||||||
|
|
||||||
else:
|
last_not_whitespace = item
|
||||||
commit_current_line()
|
last_whitespace_contains_newline = False
|
||||||
|
|
||||||
elif str_item == "(" and (
|
return end_str.strip() + "\n"
|
||||||
re.match("^([A-Za-z_\-])+\s*\(", current_line) or watch_parentheses
|
|
||||||
):
|
|
||||||
watch_parentheses = True
|
|
||||||
parentheses_balance += 1
|
|
||||||
|
|
||||||
elif str_item == ")" and watch_parentheses:
|
|
||||||
parentheses_balance -= 1
|
|
||||||
all_parentheses_closed = parentheses_balance == 0
|
|
||||||
if all_parentheses_closed:
|
|
||||||
commit_current_line(
|
|
||||||
newlines_before=2
|
|
||||||
if prev_line_type == LineType.BLOCK_CLOSE
|
|
||||||
else 1
|
|
||||||
)
|
|
||||||
watch_parentheses = False
|
|
||||||
|
|
||||||
tracker_is_empty = len(bracket_tracker) > 0
|
|
||||||
if tracker_is_empty:
|
|
||||||
last_in_tracker = bracket_tracker[-1]
|
|
||||||
is_list_comma = last_in_tracker == "[" and str_item == ","
|
|
||||||
if is_list_comma:
|
|
||||||
last_was_list_item = end_str.strip()[-1] not in ["[", ","]
|
|
||||||
if last_was_list_item:
|
|
||||||
end_str = end_str.strip()
|
|
||||||
commit_current_line()
|
|
||||||
|
|
||||||
last_not_whitespace = item
|
|
||||||
last_whitespace_contains_newline = False
|
|
||||||
|
|
||||||
return end_str.strip() + "\n"
|
|
||||||
|
|
|
||||||
|
|
@ -24,11 +24,10 @@ import traceback
|
||||||
import typing as T
|
import typing as T
|
||||||
from difflib import SequenceMatcher
|
from difflib import SequenceMatcher
|
||||||
|
|
||||||
from . import decompiler, parser, tokenizer, utils, xml_reader
|
from . import decompiler, formatter, parser, tokenizer, utils, xml_reader
|
||||||
from .ast_utils import AstNode
|
from .ast_utils import AstNode
|
||||||
from .completions import complete
|
from .completions import complete
|
||||||
from .errors import CompileError, MultipleErrors
|
from .errors import CompileError, MultipleErrors
|
||||||
from .formatter import Formatter
|
|
||||||
from .lsp_utils import *
|
from .lsp_utils import *
|
||||||
from .outputs.xml import XmlOutput
|
from .outputs.xml import XmlOutput
|
||||||
from .tokenizer import Token
|
from .tokenizer import Token
|
||||||
|
|
@ -292,7 +291,7 @@ class LanguageServer:
|
||||||
return
|
return
|
||||||
|
|
||||||
try:
|
try:
|
||||||
formatted_blp = Formatter.format(
|
formatted_blp = formatter.format(
|
||||||
open_file.text,
|
open_file.text,
|
||||||
params["options"]["tabSize"],
|
params["options"]["tabSize"],
|
||||||
params["options"]["insertSpaces"],
|
params["options"]["insertSpaces"],
|
||||||
|
|
|
||||||
|
|
@ -24,9 +24,8 @@ import os
|
||||||
import sys
|
import sys
|
||||||
import typing as T
|
import typing as T
|
||||||
|
|
||||||
from . import interactive_port, parser, tokenizer
|
from . import formatter, interactive_port, parser, tokenizer
|
||||||
from .errors import CompileError, CompilerBugError, PrintableError, report_bug
|
from .errors import CompileError, CompilerBugError, PrintableError, report_bug
|
||||||
from .formatter import Formatter
|
|
||||||
from .gir import add_typelib_search_path
|
from .gir import add_typelib_search_path
|
||||||
from .lsp import LanguageServer
|
from .lsp import LanguageServer
|
||||||
from .outputs import XmlOutput
|
from .outputs import XmlOutput
|
||||||
|
|
@ -211,7 +210,7 @@ class BlueprintApp:
|
||||||
except:
|
except:
|
||||||
errored = True
|
errored = True
|
||||||
|
|
||||||
formatted_str = Formatter.format(data, opts.spaces_num, not opts.tabs)
|
formatted_str = formatter.format(data, opts.spaces_num, not opts.tabs)
|
||||||
|
|
||||||
if data != formatted_str:
|
if data != formatted_str:
|
||||||
happened = "Would format"
|
happened = "Would format"
|
||||||
|
|
|
||||||
|
|
@ -21,7 +21,7 @@
|
||||||
import unittest
|
import unittest
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
|
|
||||||
from blueprintcompiler.formatter import Formatter
|
from blueprintcompiler import formatter
|
||||||
|
|
||||||
|
|
||||||
class TestFormatter(unittest.TestCase):
|
class TestFormatter(unittest.TestCase):
|
||||||
|
|
@ -38,7 +38,7 @@ class TestFormatter(unittest.TestCase):
|
||||||
) as f:
|
) as f:
|
||||||
expected = f.read()
|
expected = f.read()
|
||||||
|
|
||||||
actual = Formatter.format(input_data)
|
actual = formatter.format(input_data)
|
||||||
self.assertEqual(actual, expected)
|
self.assertEqual(actual, expected)
|
||||||
|
|
||||||
def test_formatter(self):
|
def test_formatter(self):
|
||||||
|
|
|
||||||
Loading…
Add table
Add a link
Reference in a new issue