Vinayak Mehta vinayak-mehta

## disease_outbreaks_camelot.ipynb

      
              1 file
            
          
              7 forks
            
          
              2 comments
            
          
              18 stars
            
          
                vinayak-mehta
                / disease_outbreaks_camelot.ipynb
            
            
              Last active
              November 5, 2023 18:54
            
              
                A jupyter notebook showing how Camelot can be used to extract tables from PDFs scraped from the IDSP website.
              
          
      Sorry, something went wrong. Reload?
      Sorry, we cannot display this file.
      Sorry, this file is invalid so it cannot be displayed.
      
          Viewer requires iframe.
      
    
## cutypr.py
import os
import sys
import json
import hmac
import hashlib
import traceback
from binascii import b2a_hex
from datetime import datetime, timezone

import zmq

## littlest-jupyter-console.py
import sys

from jupyter_client import KernelManager


try:
    manager = KernelManager()
    manager.start_kernel()

    client = manager.client()

## pdftopng-package-data-imports.txt
> Dependencies.exe -imports "C:\Users\Vinayak Mehta\pdftopng-package-data\pdftopng.cp38-win_amd64.pyd"
[-] Import listing for file : C:\Users\Vinayak Mehta\pdftopng-package-data\pdftopng.cp38-win_amd64.pyd
Import from module MSVCP140.dll :
	 Function ?_Random_device@std@@YAIXZ
	 Function ?__ExceptionPtrCreate@@YAXPEAX@Z
	 Function ?__ExceptionPtrCopy@@YAXPEAXPEBX@Z
	 Function ?_Xout_of_range@std@@YAXPEBD@Z
	 Function ?__ExceptionPtrAssign@@YAXPEAXPEBX@Z
	 Function ?_Xlength_error@std@@YAXPEBD@Z
	 Function ?__ExceptionPtrToBool@@YA_NPEBX@Z

## pdftopng-dll-mangling-imports.txt
> Dependencies.exe -imports "C:\Users\Vinayak Mehta\pdftopng-dll-mangling\pdftopng.cp38-win_amd64.pyd"
[-] Import listing for file : C:\Users\Vinayak Mehta\pdftopng-dll-mangling\pdftopng.cp38-win_amd64.pyd
Import from module MSVCP140.dll :
	 Function ?_Random_device@std@@YAIXZ
	 Function ?__ExceptionPtrCreate@@YAXPEAX@Z
	 Function ?__ExceptionPtrCopy@@YAXPEAXPEBX@Z
	 Function ?_Xout_of_range@std@@YAXPEBD@Z
	 Function ?__ExceptionPtrAssign@@YAXPEAXPEBX@Z
	 Function ?_Xlength_error@std@@YAXPEBD@Z
	 Function ?__ExceptionPtrToBool@@YA_NPEBX@Z

## pdftopng-package-data-imports.json
// Dependencies.exe -imports -json "C:\Users\Vinayak Mehta\pdftopng-package-data\pdftopng.cp38-win_amd64.pyd"
{
  "Imports": [
    {
      "Flags": 0,
      "Name": "MSVCP140.dll",
      "NumberOfEntries": 38,
      "ImportList": [
        {
          "Hint": 605,

## pdftopng-dll-mangling-imports.json
// Dependencies.exe -imports -json "C:\Users\Vinayak Mehta\pdftopng-dll-mangling\pdftopng.cp38-win_amd64.pyd"
{
  "Imports": [
    {
      "Flags": 0,
      "Name": "MSVCP140.dll",
      "NumberOfEntries": 38,
      "ImportList": [
        {
          "Hint": 605,

## pdf2png.txt
Python 3.7.8 (tags/v3.7.8:4b47a5b6ba, Jun 28 2020, 10:03:53) [MSC v.1916 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import os, time
>>> PDF_FILENAME = input("Path to PDF: ")
>>> OUT_FILE = os.path.abspath(input("Path to output PNG: "))
>>>
>>> import winrt.windows.data.pdf as PDF
>>> from winrt.windows.storage import StorageFile
>>> op = StorageFile.get_file_from_path_async(PDF_FILENAME)
>>> time.sleep(0.5) # should really await, but this is easier

## ttt.py
s = [" ", " ", " ", " ", " ", " ", " ", " ", " "]


def render_grid(s):
    grid = "\n-----------------\n".join(
        [
            f"  {s[0]}  |  {s[1]}  |  {s[2]}\n (1) | (2) | (3)",
            f"  {s[3]}  |  {s[4]}  |  {s[5]}\n (4) | (5) | (6)",
            f"  {s[6]}  |  {s[7]}  |  {s[8]}\n (7) | (8) | (9)"
        ]

## pdf_table_extract.py
#!/usr/bin/env python
"""
Usage: python pdf_table_extract.py <filename>
"""

import os
import sys

import pandas as pd
import pdftableextract as pdf
	import os
	import sys
	import json
	import hmac
	import hashlib
	import traceback
	from binascii import b2a_hex
	from datetime import datetime, timezone

	import zmq
	import sys

	from jupyter_client import KernelManager


	try:
	manager = KernelManager()
	manager.start_kernel()

	client = manager.client()
	> Dependencies.exe -imports "C:\Users\Vinayak Mehta\pdftopng-package-data\pdftopng.cp38-win_amd64.pyd"
	[-] Import listing for file : C:\Users\Vinayak Mehta\pdftopng-package-data\pdftopng.cp38-win_amd64.pyd
	Import from module MSVCP140.dll :
	Function ?_Random_device@std@@YAIXZ
	Function ?__ExceptionPtrCreate@@YAXPEAX@Z
	Function ?__ExceptionPtrCopy@@YAXPEAXPEBX@Z
	Function ?_Xout_of_range@std@@YAXPEBD@Z
	Function ?__ExceptionPtrAssign@@YAXPEAXPEBX@Z
	Function ?_Xlength_error@std@@YAXPEBD@Z
	Function ?__ExceptionPtrToBool@@YA_NPEBX@Z
	// Dependencies.exe -imports -json "C:\Users\Vinayak Mehta\pdftopng-package-data\pdftopng.cp38-win_amd64.pyd"
	{
	"Imports": [
	{
	"Flags": 0,
	"Name": "MSVCP140.dll",
	"NumberOfEntries": 38,
	"ImportList": [
	{
	"Hint": 605,
	Python 3.7.8 (tags/v3.7.8:4b47a5b6ba, Jun 28 2020, 10:03:53) [MSC v.1916 64 bit (AMD64)] on win32
	Type "help", "copyright", "credits" or "license" for more information.
	>>> import os, time
	>>> PDF_FILENAME = input("Path to PDF: ")
	>>> OUT_FILE = os.path.abspath(input("Path to output PNG: "))
	>>>
	>>> import winrt.windows.data.pdf as PDF
	>>> from winrt.windows.storage import StorageFile
	>>> op = StorageFile.get_file_from_path_async(PDF_FILENAME)
	>>> time.sleep(0.5) # should really await, but this is easier
	s = [" ", " ", " ", " ", " ", " ", " ", " ", " "]


	def render_grid(s):
	grid = "\n-----------------\n".join(
	[
	f" {s[0]} \| {s[1]} \| {s[2]}\n (1) \| (2) \| (3)",
	f" {s[3]} \| {s[4]} \| {s[5]}\n (4) \| (5) \| (6)",
	f" {s[6]} \| {s[7]} \| {s[8]}\n (7) \| (8) \| (9)"
	]
	#!/usr/bin/env python
	"""
	Usage: python pdf_table_extract.py <filename>
	"""

	import os
	import sys

	import pandas as pd
	import pdftableextract as pdf