file-parser/OpenFontFormat/table-order/tables.py

#!/usr/bin/env python3

from os import system as run
import os
from subprocess import check_output as run_and_read
import sys
from typing import Dict, Set

from OpenFont import parse_font_directory

def get_vendor(file: str) -> str:
	command = f"otfinfo -i {file} | grep 'Vendor ID' | awk '{{print $3}}'"
	return run_and_read(command, shell=True).decode(encoding='ascii').rstrip('\n')

def get_tables(file: str) -> list[str]:
	USE_OFTINFO = '--no-otfinfo' not in sys.argv
	if USE_OFTINFO:
		command = f"otfinfo -t {file} | awk '{{print $2}}'"
		return run_and_read(command, shell=True).decode(encoding='ascii').rstrip('\n').split('\n')
	else:
		with open(file, 'rb') as f:
			font_directory = parse_font_directory(f) # Since the font directory is always at the start of the file
			sorted_entries = sorted(font_directory.table_directory, key=lambda entry: entry.offset)
			return [entry.tableTag._value_ for entry in sorted_entries]


FILENAME = os.path.join(os.path.dirname(__file__), "tables")

graph: Dict[str, Set[str]] = {}

def verify_node(tag: str):
	if tag not in graph: graph[tag] = set()

def add_edge(tag1: str, tag2: str):
	verify_node(tag1)
	verify_node(tag2)
	graph[tag1].add(tag2)

MODE = None
if not sys.stdin.isatty():
	MODE = 'direct' if '--direct' in sys.argv else 'before'
	accumulator: Dict[tuple[str, str], int] = {} # acc[(tag1, tag2)] is number of times tag1 occured before tag2
	for file in sys.stdin:
		file = file.rstrip('\n')
		print(f"{get_vendor(file):<4} {file}")
		tables = get_tables(file)

		def add_to_acc(tag1: str, tag2: str):
			accumulator[(tag1, tag2)] = accumulator.get((tag1, tag2), 0)+1
		match MODE:
			case 'before':
				for i, tag1 in enumerate(tables):
					for tag2 in tables[i+1:]:
						add_to_acc(tag1, tag2)
			case 'direct':
				for i, tag1 in enumerate(tables[:-1]):
					tag2 = tables[i+1]
					add_to_acc(tag1, tag2)
			case _:
				assert False, f"Invalid mode: '{MODE}'"

	with open(f"{FILENAME}.txt", 'w') as f:
		f.write(f"mode: {MODE}\n")
		for (tag1, tag2) in accumulator:
			f.write(f"'{tag1:<4}', '{tag2:<4}', {accumulator[(tag1, tag2)]}, {accumulator.get((tag2, tag1), 0)}\n")

	for (tag1, tag2) in accumulator: add_edge(tag1, tag2)

else:
	with open(f"{FILENAME}.txt", 'r') as f:
		for i, line in enumerate(f.readlines()):
			if i == 0:
				MODE = line[6:-1]
				continue
			tag1, tag2 = line[1:5], line[9:13]
			add_edge(tag1, tag2)
assert MODE, "Unreachable"

UNTRANSITIVITY = '--untrans' in sys.argv
def untransitivity() -> None:
	to_remove: Dict[str, Set[str]] = {tag: set() for tag in graph}
	for tag1 in graph:
		for tag2 in graph[tag1]:
			for tag3 in graph[tag2]:
				if tag3 in graph[tag1]: # find a->b, b->c where a->c
					to_remove[tag1].add(tag3)

	# to_remove = {
	# 	tag1: {tag3 for tag3 in graph[tag2] for tag2 in graph[tag1] if tag3 in graph[tag1]}
	# 	for tag1 in graph
	# }

	for tag1 in to_remove:
		for tag3 in to_remove[tag1]:
			graph[tag1].remove(tag3)

if UNTRANSITIVITY: untransitivity()

GENERATE = '--svg' in sys.argv
def generate_svg() -> int:
	def node_name(tag: str) -> str:
		return '"'+tag+'"'

	with open(f"{FILENAME}.dot", 'w') as f:
		f.write("digraph {\n")
		f.write(f"\tlayout=dot\n")
		for node in graph:
			f.write(f"\t{node_name(node)}\n")
			for neighbour in graph[node]:
				f.write(f"\t{node_name(node)} -> {node_name(neighbour)};\n")
		f.write("}")

	return run(f"dot -Tsvg {FILENAME}.dot > {FILENAME}.svg")

if GENERATE: generate_svg()

print(f"{MODE = }\n{UNTRANSITIVITY = }\n{GENERATE = }")