"""Wrapper around djvulibre to access djvu files properties and content."""
#
# (C) Pywikibot team, 2015-2024
#
# Distributed under the terms of the MIT license.
#
from __future__ import annotations
import os
import re
import subprocess
from collections import Counter
import pywikibot
def _call_cmd(args, lib: str = 'djvulibre') -> tuple:
"""Tiny wrapper around subprocess.Popen().
:param args: same as Popen()
:type args: str or typing.Sequence[string]
:param lib: library to be logged in logging messages
:return: returns a tuple (res, stdoutdata), where
res is True if dp.returncode != 0 else False
"""
# upcast any param in sequence args to str
cmd = ' '.join(str(a) for a in args) if not isinstance(args, str) else args
dp = subprocess.Popen(args, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
stdoutdata, stderrdata = dp.communicate()
if dp.returncode != 0:
pywikibot.error(f'{lib} error; {cmd}')
pywikibot.error(str(stderrdata))
return (False, stdoutdata)
pywikibot.log(f'SUCCESS: {cmd} (PID: {dp.pid})')
return (True, stdoutdata)
[docs]
class DjVuFile:
"""Wrapper around djvulibre to access djvu files properties and content.
Perform file existence checks.
Control characters in djvu text-layer are converted for convenience
(see http://djvu.sourceforge.net/doc/man/djvused.html for control chars
details).
"""
def __init__(self, file: str) -> None:
"""Initializer.
:param file: filename (including path) to djvu file
"""
self._filename = file
filename = os.path.expanduser(file)
filename = os.path.abspath(filename)
# Check file exists and has read permissions.
with open(filename):
self.file = filename
self.dirname = os.path.dirname(filename)
# pattern for parsing of djvudump output.
self._pat_form = re.compile(
r' *?FORM:DJVU *?\[\d+\] *?(?P<id>{[^\}]*?})? *?\[P(?P<n>\d+)\]')
self._pat_info = re.compile(
r'DjVu.*?(?P<size>\d+x\d+).*?(?P<dpi>\d+) dpi')
def __repr__(self) -> str:
"""Return a more complete string representation."""
return f"{self.__module__}.{type(self).__name__}('{self._filename}')"
def __str__(self) -> str:
"""Return a string representation."""
return f"{self.__class__.__name__}('{self._filename}')"
[docs]
def check_cache(fn):
"""Decorator to check if cache shall be cleared."""
cache = ['_page_count', '_has_text', '_page_info']
def wrapper(obj, *args, **kwargs):
force = kwargs.get('force', False)
if force:
for el in cache:
obj.__dict__.pop(el, None)
return fn(obj, *args, **kwargs)
return wrapper
[docs]
def check_page_number(fn):
"""Decorator to check if page number is valid.
:raises ValueError
"""
def wrapper(obj, *args, **kwargs):
n = args[0]
force = kwargs.get('force', False)
if not 1 <= n <= obj.number_of_images(force=force):
raise ValueError(f'Page {int(n)} not in file {obj.file} '
f'[{int(n)}-{int(obj.number_of_images())}]')
return fn(obj, *args, **kwargs)
return wrapper
[docs]
@check_cache
def number_of_images(self, force: bool = False):
"""Return the number of images in the djvu file.
:param force: if True, refresh the cached data
"""
if not hasattr(self, '_page_count'):
res, stdoutdata = _call_cmd(['djvused', '-e', 'n', self.file])
if not res:
return False
self._page_count = int(stdoutdata)
return self._page_count
[docs]
@check_page_number
def page_info(self, n: int, force: bool = False):
"""Return a tuple (id, (size, dpi)) for page n of djvu file.
:param n: page n of djvu file
:param force: if True, refresh the cached data
"""
if not hasattr(self, '_page_info') or force:
self._get_page_info(force=force)
return self._page_info[n]
@check_cache
def _get_page_info(self, force: bool = False):
"""Return a dict of tuples for all pages of djvu file.
The tuples consist of (id, (size, dpi)).
:param force: if True, refresh the cached data
"""
if not hasattr(self, '_page_info'):
self._page_info = {}
res, stdoutdata = _call_cmd(['djvudump', self.file])
if not res:
return False
has_text = False
for line in stdoutdata.decode('utf-8').split('\n'):
if 'TXTz' in line:
has_text = True
if 'FORM:DJVU' in line:
m = self._pat_form.search(line)
if m:
key, id = int(m['n']), m['id']
else:
# If djvu doc has only one page,
# FORM:DJVU line in djvudump has no id
key, id = 1, ''
if 'INFO' in line:
m = self._pat_info.search(line)
if m:
size, dpi = m['size'], int(m['dpi'])
else:
size, dpi = None, None
else:
continue
self._page_info[key] = (id, (size, dpi))
self._has_text = has_text
return self._page_info
[docs]
def get_most_common_info(self):
"""Return most common size and dpi for pages in djvu file."""
cnt = Counter(s_d for _, s_d in self._get_page_info().values())
(size, dpi), _ = cnt.most_common()[0]
return size, dpi
[docs]
@check_cache
def has_text(self, force: bool = False):
"""Test if the djvu file has a text-layer.
:param force: if True, refresh the cached data
"""
if not hasattr(self, '_has_text'):
self._get_page_info(force=force)
return self._has_text
@staticmethod
def _remove_control_chars(data):
"""Remove djvu format control characters.
See http://djvu.sourceforge.net/doc/man/djvused.html for control chars.
:param data: the data checked for djvu format control characters
"""
txt = data.decode('utf-8')
# vertical tab (\013=\x0b): remove
txt = txt.replace('\x0b', '')
# group (\035=\x1d) separator: replace with \n
txt = txt.replace('\x1d', '\n')
# unit separator (\037=\x1f): replace with \n
txt = txt.replace('\x1f', '\n')
# feed char (\f=\x0c), \n and trailing spaces: strip
txt = txt.strip('\x0c\n ')
return txt
[docs]
@check_page_number
@check_cache
def get_page(self, n: int, force: bool = False):
"""Get page n for djvu file.
:param n: page n of djvu file
:param force: if True, refresh the cached data
"""
if not self.has_text(force=force):
raise ValueError(f'Djvu file {self.file} has no text layer.')
res, stdoutdata = _call_cmd(['djvutxt', f'--page={n}',
self.file])
if not res:
return False
return self._remove_control_chars(stdoutdata)
[docs]
@check_page_number
def whiten_page(self, n) -> bool:
"""Replace page 'n' of djvu file with a blank page.
:param n: page n of djvu file
:type n: int
"""
# tmp files for creation/insertion of a white page.
white_ppm = os.path.join(self.dirname, 'white_page.ppm')
white_djvu = os.path.join(self.dirname, 'white_page.djvu')
n_tot = self.number_of_images()
# Check n is in valid range and set ref_page number for final checks.
ref_page = 2 if n == 1 else n - 1
size, dpi = self.get_most_common_info()
# Generate white_page.
res, _ = _call_cmd(['convert', '-size', size, 'xc:white', white_ppm],
lib='ImageMagik')
if not res:
return False
# Convert white_page to djvu.
res, _ = _call_cmd(['c44', white_ppm, '-dpi', str(dpi)])
os.unlink(white_ppm) # rm white_page.ppm before returning.
if not res:
return False
# Delete page n.
# Get ref page info for later checks.
info_ref_page = self.page_info(ref_page)
res, _ = _call_cmd(['djvm', '-d', self.file, str(n)])
if not res:
return False
# Insert new page
res, _ = _call_cmd(['djvm', '-i', self.file, white_djvu, str(n)])
os.unlink(white_djvu) # rm white_page.djvu before returning.
if not res:
return False
# Check if page processing is as expected.
expected_id = '{%s}' % os.path.basename(white_djvu)
assert self.number_of_images(force=True) == n_tot
assert self.page_info(n) == (expected_id, (size, dpi)) # white page id
assert self.page_info(ref_page) == info_ref_page # ref page info.
return True
[docs]
@check_page_number
def delete_page(self, n) -> bool:
"""Delete page 'n' of djvu file.
:param n: page n of djvu file
:type n: int
"""
n_tot = self.number_of_images()
# Check n is in valid range and set ref_page number for final checks.
ref_page = n - 1 if n == n_tot else n + 1
new_ref_page = n - 1 if n == n_tot else n
# Delete page n.
# Get ref page info for later checks.
info_ref_page = self.page_info(ref_page)
res, _ = _call_cmd(['djvm', '-d', self.file, str(n)])
if not res:
return False
# Check if page processing is as expected.
# ref page info.
if n_tot > 2:
assert self.number_of_images(force=True) == n_tot - 1
# cache cleared above
assert self.page_info(new_ref_page) == info_ref_page
else:
# If djvu has only one page, FORM:DJVU line in djvudump has no id
_id, (sz, dpi) = info_ref_page
assert self.page_info(new_ref_page, force=True) == ('', (sz, dpi))
return True
# This is to be used only if this class is subclassed and the decorators
# needs to be used by the child.
check_page_number = staticmethod(check_page_number)
check_cache = staticmethod(check_cache)