Phpwcms Backend File Downloader

get-links.js

/**
 * Execute this on the phpwcms files page when all wanted folders are opened
 * (all visible download buttons will be added to the array)
 **/
var save_buttons = $('img[src="img/button/download_disc.gif"]');

function isFolder($tr) {
  return $tr.attr('bgcolor') == '#EBF2F4';
}

function getIndent($tr) {
  if(isFolder($tr)) {
    var spacer = $tr.find('td.msglist:eq(0) > img:eq(1)');
    var ind = spacer.attr('width');
  } else {
    var spacer = $tr.find('td.msglist');
    var ind = spacer.attr('width') - 15;
  }
  return ind ? ind : 999;
}

function getName($tr) {
  return $tr.find('strong').text();
}

function getLink($e) {
  return $e.closest('a').attr('href');
}

function getFileName($e) {
  return $e.closest('tr').find('td:eq(2) > a').text();
}

function getParentDir($e) {
  // get subtable
  /*while (!(closest = $e.closest('table')).is('#files')) {
    $e = closest;
  }*/
  if (!$e.is('tr')) {
   var $row = $e.closest('table').closest('tr');
  } else {
   var $row = $e;
  }
  var ind = getIndent($row);
  
  while($row.length && getIndent($row) >= ind) // skip boring tr's
    $row = $row.prev();
  return $row;
}

function getPath($e) {
  var path = "";
  while (($parent = getParentDir($e)).length) {
    path = getName($parent) + "/" + path;
    $e = $parent;
  }
  return path;
}



var data = {};
for(var save of saves) {
  var $save = $(save);
  data[getLink($save)] = [getFileName($save), getPath($save)];
}

// print result to console
console.log(JSON.stringify(data));

wcms-downloader.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This scripts downloads files from the phpwcms backend\n",
    "\n",
    "### Required data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# ENTER DATA HERE:\n",
    "backend_links_file = 'wcms-files.json'  # contains download links+paths of wcms backend, generated by a JS script\n",
    "target_dir = 'wcms-files/' # target directory for downloads\n",
    "\n",
    "wcms_baseurl = 'http://www.your-site.org/'\n",
    "cookies = {'PHPSESSID': 'XYZ'}  # for Phpwcms backend"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Initialize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import json\n",
    "import codecs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# small helper function to peek into variables\n",
    "def glance(collection, n=3):\n",
    "    print(\"%d entries\" % len(collection))\n",
    "    if type(collection) == dict:\n",
    "        return list(collection.items())[:n]\n",
    "    elif type(collection) == list:\n",
    "        return collection[:n]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "332 entries\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('include/inc_act/act_download.php?pl=1&dl=455',\n",
       "  ['Example file 1.pdf', 'PDF/']),\n",
       " ('include/inc_act/act_download.php?pl=1&dl=48',\n",
       "  ['Example file 2.jpg', 'Images/Other/']),\n",
       " ('include/inc_act/act_download.php?pl=1&dl=142',\n",
       "  ['DSC_0047_2.jpg', 'Images'])]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# generated by JS\n",
    "with codecs.open(backend_links_file, encoding='utf-8') as f:\n",
    "    wcms_links = json.load(f)\n",
    "    \n",
    "glance(wcms_links) # quick glance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Download"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import requests\n",
    "\n",
    "def dl(link, path, filename):\n",
    "    r = requests.get(wcms_baseurl + link, cookies=cookies, stream=True)\n",
    "    r.raise_for_status()\n",
    "    \n",
    "    os.makedirs(target_dir + path, exist_ok=True) # create directory if not existent\n",
    "    with open(target_dir + path + filename, 'wb') as fd:\n",
    "        for chunk in r.iter_content(1024):\n",
    "            fd.write(chunk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
   ],
   "source": [
    "done_count = 0\n",
    "total_count = len(wcms_links)\n",
    "for link, loc in wcms_links.items():\n",
    "    print('Downloading [%d / %d]: %s -> %s' % (done_count+1, total_count, link, loc[1] + loc[0]))\n",
    "    dl(link, loc[1], loc[0])\n",
    "    done_count += 1"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [Root]",
   "language": "python",
   "name": "Python [Root]"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}

wcms-downloader.py

# coding: utf-8

# This scripts downloads files from the phpwcms backend
# 
# ### Required data:

# In[19]:

# ENTER DATA HERE:
backend_links_file = 'wcms-files.json'  # contains download links+paths of wcms backend, generated by a JS script
target_dir = 'wcms-files/' # target directory for downloads

wcms_baseurl = 'http://www.your-site.org/'
cookies = {'PHPSESSID': 'XYZ'}  # for Phpwcms backend


# # Initialize

# In[5]:

import json
import codecs


# In[3]:

# small helper function to peek into variables
def glance(collection, n=3):
    print("%d entries" % len(collection))
    if type(collection) == dict:
        return list(collection.items())[:n]
    elif type(collection) == list:
        return collection[:n]


# # Load Data

# In[10]:

# generated by JS
with codecs.open(backend_links_file, encoding='utf-8') as f:
    wcms_links = json.load(f)
    
glance(wcms_links) # quick glance


# # Download

# In[15]:

import os
import requests

def dl(link, path, filename):
    r = requests.get(wcms_baseurl + link, cookies=cookies, stream=True)
    r.raise_for_status()
    
    os.makedirs(target_dir + path, exist_ok=True) # create directory if not existent
    with open(target_dir + path + filename, 'wb') as fd:
        for chunk in r.iter_content(1024):
            fd.write(chunk)


# In[18]:

done_count = 0
total_count = len(wcms_links)
for link, loc in wcms_links.items():
    print('Downloading [%d / %d]: %s -> %s' % (done_count+1, total_count, link, loc[1] + loc[0]))
    dl(link, loc[1], loc[0])
    done_count += 1