Skip to content

removing unesserary imports and implementing add_macro and add_type #41310

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
rashadalsharpini wants to merge 1 commit into
base: develop
Choose a base branch
from
+84 −57
Open

removing unesserary imports and implementing add_macro and add_type #41310

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
141 changes: 84 additions & 57 deletions src/sage_setup/autogen/flint/reader.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,24 +2,22 @@
Extraction of function, macros, types from flint documentation.
"""

#*****************************************************************************
# *****************************************************************************
# Copyright (C) 2023 Vincent Delecroix <20100.delecroix@gmail.com>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 2 of the License, or
# (at your option) any later version.
# http://www.gnu.org/licenses/
#*****************************************************************************

import os
from .env import FLINT_INCLUDE_DIR, FLINT_DOC_DIR
# *****************************************************************************


class Extractor:
r"""
Tool to extract function declarations from a flint .rst file
"""

NONE = 0
DOC = 1
FUNCTION_DECLARATION = 2
Expand All @@ -28,16 +26,20 @@ class Extractor:

def __init__(self, filename):
self.filename = filename
if not filename.endswith('.rst'):
if not filename.endswith(".rst"):
raise ValueError

# Attributes that are modified throughout the document parsing
self.state = self.NONE # position in the documentation
self.section = None # current section
self.content = {} # section -> list of pairs (function signatures, func documentation)
self.functions = [] # current list of pairs (function signatures, func documentation)
self.signatures = [] # current list of function/macro/type signatures
self.doc = [] # current function documentation
self.state = self.NONE # position in the documentation
self.section = None # current section
self.content = (
{}
) # section -> list of pairs (function signatures, func documentation)
self.functions = (
[]
) # current list of pairs (function signatures, func documentation)
self.signatures = [] # current list of function/macro/type signatures
self.doc = [] # current function documentation

with open(filename) as f:
text = f.read()
Expand Down Expand Up @@ -68,7 +70,7 @@ def clean_doc(self):

for i, line in enumerate(self.doc):
# To make sage linter happier
line = line.replace('\\choose ', 'choose ')
line = line.replace("\\choose ", "choose ")
self.doc[i] = line

@staticmethod
Expand All @@ -78,51 +80,64 @@ def has_boolean_return_type(func_signature):

If so, it will be declared in Cython as `bint` rather than `int`.
"""
if func_signature.count('(') != 1 or func_signature.count(')') != 1:
if func_signature.count("(") != 1 or func_signature.count(")") != 1:
return False

j = func_signature.index('(')
j = func_signature.index("(")
func_name = func_signature[:j].strip().split()
if len(func_name) != 2:
return False

return_type = func_name[0]
if return_type != 'int':
if return_type != "int":
return False

func_name = func_name[1]

return func_name.startswith('is_') or \
'_is_' in func_name or \
func_name.endswith('_eq') or \
func_name.endswith('_ne') or \
func_name.endswith('_lt') or \
func_name.endswith('_le') or \
func_name.endswith('_gt') or \
func_name.endswith('_ge') or \
'_contains_' in func_name or \
func_name.endswith('_contains') or \
'_equal_' in func_name or \
func_name.endswith('_equal') or \
func_name.endswith('_overlaps')
return (
func_name.startswith("is_")
or "_is_" in func_name
or func_name.endswith("_eq")
or func_name.endswith("_ne")
or func_name.endswith("_lt")
or func_name.endswith("_le")
or func_name.endswith("_gt")
or func_name.endswith("_ge")
or "_contains_" in func_name
or func_name.endswith("_contains")
or "_equal_" in func_name
or func_name.endswith("_equal")
or func_name.endswith("_overlaps")
)

def clean_signatures(self):
if (self.state & self.FUNCTION_DECLARATION) or (self.state & self.MACRO_DECLARATION):
if (self.state & self.FUNCTION_DECLARATION) or (
self.state & self.MACRO_DECLARATION
):
for i, func_signature in enumerate(self.signatures):
replacement = [('(void)', '()'), (' enum ', ' ')]
replacement = [("(void)", "()"), (" enum ", " ")]
for bad_type, good_type in replacement:
func_signature = func_signature.replace(bad_type, good_type)

bad_arg_names = [('in', 'input'), ('lambda', 'lmbda'), ('iter', 'it'), ('is', 'iis')]
replacements = [(pattern.format(bad), pattern.format(good)) for pattern in [' {},', ' {})', '*{},', '*{})'] for bad, good in bad_arg_names]
bad_arg_names = [
("in", "input"),
("lambda", "lmbda"),
("iter", "it"),
("is", "iis"),
]
replacements = [
(pattern.format(bad), pattern.format(good))
for pattern in [" {},", " {})", "*{},", "*{})"]
for bad, good in bad_arg_names
]
for bad_form, good_form in replacements:
func_signature = func_signature.replace(bad_form, good_form)

if self.has_boolean_return_type(func_signature):
func_signature = func_signature.strip()
if not func_signature.startswith('int '):
if not func_signature.startswith("int "):
raise RuntimeError
func_signature = 'b' + func_signature
func_signature = "b" + func_signature

self.signatures[i] = func_signature

Expand All @@ -144,10 +159,10 @@ def add_function(self):
# Drop va_list argument
signatures = []
for func_signature in self.signatures:
if '(' not in func_signature or ')' not in func_signature:
if "(" not in func_signature or ")" not in func_signature:
raise RuntimeError(func_signature)
elif 'va_list ' in func_signature:
print('Warning: va_list unsupported {}'.format(func_signature))
elif "va_list " in func_signature:
print("Warning: va_list unsupported {}".format(func_signature))
else:
signatures.append(func_signature)
self.signatures = signatures
Expand All @@ -156,11 +171,14 @@ def add_function(self):
self.functions.append((tuple(self.signatures), tuple(self.doc)))

def add_macro(self):
# TODO: we might want to support auto-generation of macros
self.clean_doc()
self.clean_signatures()
self.functions.append((tuple(self.signatures), tuple(self.doc)))
return

def add_type(self):
# TODO: we might want to support auto-generation of types
self.clean_doc()
self.functions.append((tuple(self.signatures), tuple(self.doc)))
return

def process_line(self):
Expand All @@ -170,38 +188,47 @@ def process_line(self):
if self.i >= len(self.lines):
return 0

if bool(self.state & self.FUNCTION_DECLARATION) + bool(self.state & self.MACRO_DECLARATION) + bool(self.state & self.TYPE_DECLARATION) > 1:
raise RuntimeError('self.state = {} and i = {}'.format(self.state, self.i))
if (
bool(self.state & self.FUNCTION_DECLARATION)
+ bool(self.state & self.MACRO_DECLARATION)
+ bool(self.state & self.TYPE_DECLARATION)
> 1
):
raise RuntimeError("self.state = {} and i = {}".format(self.state, self.i))

line = self.lines[self.i]
if line.startswith('.. function::'):
if line.startswith(".. function::"):
self.add_declaration()
line_rest = line.removeprefix('.. function::')
if not line_rest.startswith(' '):
print('Warning: no space {}'.format(line))
line_rest = line.removeprefix(".. function::")
if not line_rest.startswith(" "):
print("Warning: no space {}".format(line))
self.state = self.FUNCTION_DECLARATION
self.i += 1
signature = line_rest.strip()
while signature.endswith('\\'):
signature = signature.removesuffix('\\').strip() + ' ' + self.lines[self.i].strip()
while signature.endswith("\\"):
signature = (
signature.removesuffix("\\").strip()
+ " "
+ self.lines[self.i].strip()
)
self.i += 1
self.signatures.append(signature)
elif line.startswith('.. macro::'):
elif line.startswith(".. macro::"):
self.add_declaration()
if line[10] != ' ':
print('Warning no space{}'.format(line))
if line[10] != " ":
print("Warning no space{}".format(line))
self.signatures.append(line[10:].strip())
self.state = self.MACRO_DECLARATION
self.i += 1
elif line.startswith('.. type::'):
elif line.startswith(".. type::"):
# type
# NOTE: we do nothing as the documentation duplicates type declaration
# and lacks the actual list of attributes
self.add_declaration()
self.state = self.TYPE_DECLARATION
self.i += 1
elif self.state == self.FUNCTION_DECLARATION:
if len(line) > 14 and line.startswith(' ' * 14):
if len(line) > 14 and line.startswith(" " * 14):
# function with similar declaration
line = line[14:].strip()
if line:
Expand All @@ -214,7 +241,7 @@ def process_line(self):
else:
raise ValueError(line)
elif self.state == self.MACRO_DECLARATION:
if len(line) > 10 and line.startswith(' ' * 10):
if len(line) > 10 and line.startswith(" " * 10):
# macro with similar declaration
line = line[10:].strip()
if line:
Expand All @@ -226,18 +253,18 @@ def process_line(self):
self.i += 1
else:
raise ValueError(line)
elif (self.state & self.DOC) and line.startswith(' '):
elif (self.state & self.DOC) and line.startswith(" "):
# function doc
line = line.strip()
if line:
self.doc.append(line)
self.i += 1
elif self.i + 1 < len(self.lines) and self.lines[self.i + 1].startswith('----'):
elif self.i + 1 < len(self.lines) and self.lines[self.i + 1].startswith("----"):
# new section
self.add_declaration()
if self.functions:
self.update_section()
section = line
# section = line
self.i += 2
elif not line:
self.i += 1
Expand Down