qpdf/generate_auto_job

#!/usr/bin/env python3
import os
import sys
import argparse
import hashlib
import re
import yaml

whoami = os.path.basename(sys.argv[0])
BANNER = f'''//
// This file is automatically generated by {whoami}.
// Edits will be automatically overwritten if the build is
// run in maintainer mode.
//'''


def warn(*args, **kwargs):
    print(*args, file=sys.stderr, **kwargs)


class Main:
    SOURCES = [whoami, 'job.yml']
    DESTS = {
        'decl': 'libqpdf/qpdf/auto_job_decl.hh',
        'init': 'libqpdf/qpdf/auto_job_init.hh',
    }
    SUMS = 'job.sums'

    def main(self, args=sys.argv[1:], prog=whoami):
        options = self.parse_args(args, prog)
        self.top(options)

    def parse_args(self, args, prog):
        parser = argparse.ArgumentParser(
            prog=prog,
            description='Generate files for QPDFJob',
        )
        mxg = parser.add_mutually_exclusive_group(required=True)
        mxg.add_argument('--check',
                         help='update checksums if files are not up to date',
                         action='store_true', default=False)
        mxg.add_argument('--generate',
                         help='generate files from sources',
                         action='store_true', default=False)
        return parser.parse_args(args)

    def top(self, options):
        if options.check:
            self.check()
        elif options.generate:
            self.generate()
        else:
            exit(f'{whoami} unknown mode')

    def get_hashes(self):
        hashes = {}
        for i in sorted([*self.SOURCES, *self.DESTS.values()]):
            m = hashlib.sha256()
            try:
                with open(i, 'rb') as f:
                    m.update(f.read())
                hashes[i] = m.hexdigest()
            except FileNotFoundError:
                pass
        return hashes

    def check(self):
        hashes = self.get_hashes()
        match = False
        try:
            old_hashes = {}
            with open(self.SUMS, 'r') as f:
                for line in f.readlines():
                    m = re.match(r'^(\S+) (\S+)\s*$', line)
                    if m:
                        old_hashes[m.group(1)] = m.group(2)
            match = old_hashes == hashes
        except Exception:
            pass
        if not match:
            exit(f'{whoami}: auto job inputs have changed')

    def update_hashes(self):
        hashes = self.get_hashes()
        with open(self.SUMS, 'w') as f:
            print(f'# Generated by {whoami}', file=f)
            for k, v in hashes.items():
                print(f'{k} {v}', file=f)

    def generate(self):
        warn(f'{whoami}: regenerating auto job files')

        with open('job.yml', 'r') as f:
            data = yaml.safe_load(f.read())
        self.validate(data)
        with open(self.DESTS['decl'], 'w') as f:
            print(BANNER, file=f)
            self.generate_decl(data, f)
        with open(self.DESTS['init'], 'w') as f:
            print(BANNER, file=f)
            self.generate_init(data, f)

        # Update hashes last to ensure that this will be rerun in the
        # event of a failure.
        self.update_hashes()
        # DON'T ADD CODE TO generate AFTER update_hashes

    def check_keys(self, what, d, exp):
        if not isinstance(d, dict):
            exit(f'{what} is not a dictionary')
        actual = set(d.keys())
        extra = actual - exp
        if extra:
            exit(f'{what}: unknown keys = {extra}')

    def validate(self, data):
        self.check_keys('top', data, set(['choices', 'options']))
        for o in data['options']:
            self.check_keys('top', o, set(
                ['table', 'prefix', 'bare', 'positional',
                 'optional_parameter', 'required_parameter',
                 'required_choices', 'from_table']))

    def to_identifier(self, label, prefix, const):
        identifier = re.sub(r'[^a-zA-Z0-9]', '_', label)
        if const:
            identifier = identifier.upper()
        else:
            identifier = identifier.lower()
            identifier = re.sub(r'(?:^|_)([a-z])',
                                lambda x: x.group(1).upper(),
                                identifier).replace('_', '')
        return prefix + identifier

    def generate_decl(self, data, f):
        for o in data['options']:
            table = o['table']
            if table in ('main', 'help'):
                continue
            i = self.to_identifier(table, 'O_', True)
            print(f'static constexpr char const* {i} = "{table}";', file=f)
        print('', file=f)
        for o in data['options']:
            table = o['table']
            prefix = 'arg' + o.get('prefix', '')
            if o.get('positional', False):
                print(f'void {prefix}Positional(char*);', file=f)
            for i in o.get('bare', []):
                identifier = self.to_identifier(i, prefix, False)
                print(f'void {identifier}();', file=f)
            for i in o.get('optional_parameter', []):
                identifier = self.to_identifier(i, prefix, False)
                print(f'void {identifier}(char *);', file=f)
            for i in o.get('required_parameter', {}):
                identifier = self.to_identifier(i, prefix, False)
                print(f'void {identifier}(char *);', file=f)
            for i in o.get('required_choices', {}):
                identifier = self.to_identifier(i, prefix, False)
                print(f'void {identifier}(char *);', file=f)
            if table not in ('main', 'help'):
                identifier = self.to_identifier(table, 'argEnd', False)
                print(f'void {identifier}();', file=f)

    def generate_init(self, data, f):
        for k, v in data['choices'].items():
            print(f'char const* {k}_choices[] = {{', file=f, end='')
            for i in v:
                print(f'"{i}", ', file=f, end='')
            print('0};', file=f)


if __name__ == '__main__':
    try:
        os.chdir(os.path.dirname(os.path.realpath(__file__)))
        Main().main()
    except KeyboardInterrupt:
        exit(130)