Skip to content

Instantly share code, notes, and snippets.

View peci1's full-sized avatar

Martin Pecka peci1

81 followers · 3 following
  • Czech Technical University in Prague
  • Czech republic
View GitHub Profile
@peci1
peci1 / scipy_odr_test.py
Created September 17, 2018 14:18
Test of scipy.odr regressor
from __future__ import print_function
import numpy as np
import scipy.linalg
from scipy.odr import *
import matplotlib as mpl
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import pyplot as plt
import sys
import time
@peci1
peci1 / anonymize_pdf.sh
Last active January 4, 2021 18:07
Anonymize a PDF document with comments so that both document metadata and comment author data do not disclose who the author is.
qpdf --qdf --object-streams=disable $1 $1.tmp # uncompress /FlateDecode sections
perl -pe 's/(?<=\/T \()(.*?)(?=\))/ "x" x length($1) /e' $1.tmp > $1.tmp2 # remove /T commands containing author name in comments
qpdf --compress-streams=y $1.tmp2 $1.tmp3 # recompress streams
qpdf --empty --pages $1.tmp3 1-z -- $1.anonymous.pdf # remove document metadata (this also removes whole comments)
rm $1.tmp* # cleanup
@peci1
peci1 / pb_model_executor.py
Created June 1, 2018 20:25
Tensorflow executor of models from TF 1.4+ on TF 1.3- (python 3)
import tensorflow as tf
import numpy as np
class PbModelExecutor:
def __init__(self, model_path, input_shape, input_name='input', output_name='output:0'):
self.session = tf.Session()
self.input = tf.placeholder(tf.float32, shape=input_shape)
@peci1
peci1 / nas
Created May 31, 2018 00:01
OpenWRT: automount NTFS drive by UUID and before mountd comes into action and mounts it under /tmp/run/mountd
#!/bin/sh /etc/rc.common
# This is a procd initscript, so put it in /etc/init.d/
START=79 # Important: this number must be lower than the START number in /etc/init.d/mountd !!!
USE_PROCD=1
PROGNAME="/usr/bin/ntfs-3g"
# Fill these values as desired
@peci1
peci1 / print_bytes_in_all_encodings.py
Created December 25, 2017 23:05
Print the characters corresponding to the given byte sequence in all encodings known to the python interpreter.
#!/usr/bin/env python3
import encodings
import pkgutil
bytes_to_decode = b"\xc4\x9b"
codecs = []
for importer, modname, ispkg in pkgutil.iter_modules(encodings.__path__):
try:
@peci1
peci1 / torrc
Last active March 31, 2022 14:47
My torrc on Turris Omnia
SOCKSPort 9050
SOCKSPolicy accept 127.0.0.1/32
SOCKSPolicy accept 192.168.18.0/24
SOCKSPolicy accept6 FC00::/7
SOCKSPolicy reject *
Log notice syslog
RunAsDaemon 1
DataDirectory /srv/tor
AvoidDiskWrites 1
ORPort 9001
@peci1
peci1 / transparent-tor.diff
Created November 2, 2017 02:14
A trial to get transparent .onion translation on Omnia
diff -Nru @161/etc/config/resolver @/etc/config/resolver
--- @161/etc/config/resolver 2017-10-01 21:01:14.375423242 +0200
+++ @/etc/config/resolver 2017-11-02 01:13:47.862130135 +0100
@@ -21,6 +21,7 @@
option log_stderr '1'
option log_stdout '1'
option forks '1'
+ option include_config '/etc/kresd/config.conf'
config resolver 'unbound'
@peci1
peci1 / compile_darktable_on_windows.sh
Last active June 19, 2021 11:34
A script for compiling Darktable on Windows using the Windows Subsystem for Linux ("bash for Windows") introduced in Win 10.
#########################################
# DO NOT REPORT ANY BUGS FROM DARKTABLE INSTALLED THIS WAY.
#
# This is an unofficial compilation & installation guide.
# Darktable developers explicitly say that they don't support Windows ANY way and won't fix ANY bugs on Windows.
# Please, respect their decision and do not bother them with Windows bugreports.
#
# If you find a bug, please verify it on a regular Linux/Mac system or inside a virtual machine.
#
# There is a chance an official Windows build will appear. Just subscribe to https://github.com/darktable-org/darktable/pull/1327 to be informed.
@peci1
peci1 / Abbreviate Journal Names in Bibtex Database.py
Last active January 28, 2021 07:22 — forked from FilipDominec/Abbreviate Journal Names in Bibtex Database.py
Using the translation table from the Jabref program, finds and replaces all scientific journal names to their standardized abbreviated form. First argument is the file to be processed; outputs safely to 'abbreviated.bib'
#!/usr/bin/env python
#-*- coding: utf-8 -*-
import sys, os
import re
try: bibtexdb = open(sys.argv[1]).read()
except: print "Error: specify the file to be processed!"
if not os.path.isfile('journalList.txt'):
@peci1
peci1 / slow_topic_diag_update.h
Created February 22, 2016 18:05
// Author: Martin Pecka ( martin.pecka@cvut.cz )
// License: BSD
#ifndef DEM_GENERATION_SLOW_TOPIC_DIAG_UPDATE_H
#define DEM_GENERATION_SLOW_TOPIC_DIAG_UPDATE_H
#define private public
#include <diagnostic_updater/update_functions.h>
#undef private