crawler classes and image downloader

.gitignore

@@ -0,0 +1,6 @@
+env
+*.out
+images*
+*.tsv
+env_wikicrawler
+temp_images

README.md

@@ -0,0 +1,13 @@
+# Wikisource crawler and image downloader
+
+## Requirements:
+Python 3.8>
+ 
+## Install/setup:
+`pip install -r requirements.txt`
+
+## Usage crawler
+`python crawler.py --type {green or yellow or red} --output_file_name {output tsv file name} --start_file_name {name of file to start crawling from} --start_page_number {page of file to start crawling}`
+
+## Usage image downloader
+`python image_download.py --file_path {tsv file with data to download} --output_folder {folder to output images -> default images} --max_folder_size_mb {size in MB to stop, if not given will download all} --from_checkpoint {True to start from checkpoint if pickle available}`

crawler.py

@@ -6,54 +6,79 @@ from tqdm import tqdm
 import time
 import argparse
 
-def main():
+MAIN_URL = "https://pl.wikisource.org/"
-    MAIN_URL = "https://pl.wikisource.org/"
-    URL_YELLOW = "https://pl.wikisource.org/wiki/Kategoria:Skorygowana"
-    URL_GREEN = "https://pl.wikisource.org/wiki/Kategoria:Uwierzytelniona"
 
-    def get_page_data(page_element):
+def get_page_data(page_element):
         time.sleep(0.5)
         doc = requests.get(MAIN_URL + page_element['href'])
-        doc_soup = BeautifulSoup(doc.text, 'lxml', from_encoding="utf-8")
+        doc_soup = BeautifulSoup(doc.text, 'lxml')
-        text_elem = doc_soup.find("div", {"class": "pagetext"}).next_element
+        text = doc_soup.find("div", {"class": "pagetext"}).next_element
-        text = text_elem.text
+        image_url = doc_soup.find("div", {"class": "prp-page-image"}).find("img")['src']
-        image_url = doc_soup.find("div", {"class": "prp-page-image"}).next_element['src']
+        return {"title": page_element['title'], "href": MAIN_URL + page_element['href'], "image_url": image_url, "text": text.text}
-        return {"title": page_element['title'], "href": MAIN_URL + page_element['href'], "image_url": image_url, "text": text,}
 
-    r  = requests.get(URL_GREEN)
+def save_data(file_name, data, args):
+    if not args.testing:
+        df = pd.DataFrame(data)
+        df.to_csv(f"./{file_name}.tsv", sep="\t")
+
+def main(args):
+    category_dict = {"green": "Uwierzytelniona", "yellow": "Skorygowana", "red": "Przepisana"}
+    if args.start_file_name and args.start_page_number:
+        CATEGORY_URL = f"{MAIN_URL}/w/index.php?title=Kategoria:{category_dict[args.type]}&pagefrom={args.start_file_name}"
+    else:
+        CATEGORY_URL = f"{MAIN_URL}/wiki/Kategoria:{category_dict[args.type]}"
+
+    r  = requests.get(CATEGORY_URL)
     soup = BeautifulSoup(r.text, 'lxml')
-    page_number = 1
+    page_number = 1 if not args.start_page_number else args.start_page_number
+    data_number = 0 if not args.start_page_number else args.start_page_number * 200
     result_list = []
-    max_len = "".join(re.findall("\d", re.sub("\xa0",'', soup.find("div", {"id": "mw-pages"}).find("p").text))[3:])
+    max_len = int("".join(re.findall("\d", re.sub("\xa0",'', soup.find("div", {"id": "mw-pages"}).find("p").text))[3:]))
+    
     try:
-        while True:
+        with tqdm(total=max_len) as pbar:
-            time.sleep(5)
+                if args.start_page_number:
-            next_page = soup.find("a", {"href": re.compile(r"\/w\/index.php.*")}, string="następna strona").get('href', None)
+                    pbar.update(data_number)
-            if next_page and page_number != 1:
+                    pbar.refresh()
-                r = requests.get(MAIN_URL + next_page)
-            elif not next_page:
-                break
 
-            if r.status_code != 200:
+                while data_number < max_len:
-                print(r.__dict__)
+                    pbar.set_description(f"Page number: {page_number}")
-                time.sleep(10)
+                    time.sleep(5)
-                r = requests.get(MAIN_URL + next_page)
+                    next_page = soup.find("a", {"href": re.compile(r"\/w\/index.php.*")}, string="następna strona").get('href', None)
-                if r.status_code != 200:
+                    if next_page and page_number != 1:
-                    break
+                        r = requests.get(MAIN_URL + next_page)
-            soup = BeautifulSoup(r.text, 'lxml')
+                    elif not next_page:
-            page_number += 1
+                        break
-            links = soup.find_all("a", {"href": re.compile(r"\/wiki\/Strona:.*")})
+
-            for link in tqdm(links):
+                    if r.status_code != 200:
-                result_list.append(get_page_data(link))
+                        print(r.__dict__)
-            print("Page number:", page_number)
+                        time.sleep(60)
-            print("Number of elements:", 200 * page_number, "/", max_len)
+                        r = requests.get(MAIN_URL + next_page)
+                        if r.status_code != 200:
+                            break
+                    soup = BeautifulSoup(r.text, 'lxml')
+                    links = soup.find_all("a", {"href": re.compile(r"\/wiki\/Strona:.*")})
+                    page_number += 1
+                    for link in links:
+                        result_list.append(get_page_data(link))
+                        data_number += 1
+                        pbar.update(1)
     except Exception as e:
-        print(e)
+        print("Error:", e)
-        df = pd.DataFrame(result_list)
+        save_data(f"./{args.output_file_name}-{args.type}", result_list, args)
-        df.to_csv("./green.tsv", sep="\t")
+    except KeyboardInterrupt:
+        save_data(f"./{args.output_file_name}-{args.type}", result_list, args)
+
+    save_data(f"./{args.output_file_name}-{args.type}", result_list, args)
 
-    df = pd.DataFrame(result_list)
-    df.to_csv("./yellow.tsv", sep="\t")
 
 if __name__ == "__main__":
-    main()
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--wiki_type", type=str, default='green', choices=["green", "yellow", "red"], required=True)
+    parser.add_argument("--output_file_name", type=str, required=True)
+    parser.add_argument("--start_file_name", type=str, required=False)
+    parser.add_argument("--start_page_number", type=int, required=False)
+    parser.add_argument("--testing", type=bool, required=False, default=False)
+    args, left_argv = parser.parse_known_args()
+    print(type(vars(args)))
+    # main(args)

crawler_class.py

@@ -0,0 +1,115 @@
+from bs4 import BeautifulSoup
+import requests
+import time
+import os
+from tqdm import tqdm
+import csv
+from collections import deque
+import argparse
+import re
+import pandas as pd
+
+MAIN_URL = "https://pl.wikisource.org/"
+
+
+class WikiCrawler:
+    def __init__(self, wiki_type: str, output_file_name: str):
+        self.wiki_type = wiki_type
+        self.output_file_name = output_file_name
+        self.page_number = 1
+        self.index = 1
+        self.load_last_checkpoint()
+    
+    def load_last_checkpoint(self):
+        self.start_file_name = None
+        if os.path.exists(self.output_file_name):
+            df = pd.read_csv(self.output_file_name, encoding='utf-8', sep='\t')
+            last_line = df.tail(1).iloc()[0]
+            self.start_file_name = last_line[0]
+            self.page_number = int(last_line[-1])
+            self.index = int(last_line[-2])
+            del df
+            print(f"Starting from index: {self.index}, page: {self.page_number}")
+
+    def _init_crawl(self):
+        category_dict = {"green": "Uwierzytelniona", "yellow": "Skorygowana", "red": "Przepisana"}
+        CATEGORY_URL = f"{MAIN_URL}/wiki/Kategoria:{category_dict[self.wiki_type]}"
+        # if should start from other step
+        if self.start_file_name:
+            CATEGORY_URL = f"{MAIN_URL}/w/index.php?title=Kategoria:{category_dict[self.wiki_type]}&pagefrom={self.start_file_name}"
+        request = requests.get(CATEGORY_URL)
+        assert request.status_code == 200, f"Status diffrent on main request, status: {request.status_code}"
+
+        soup = BeautifulSoup(request.text, 'lxml')
+        self.max_len = int("".join(re.findall("\d", re.sub("\xa0",'', soup.find("div", {"id": "mw-pages"}).find("p").text))[3:]))
+        self.pbar = tqdm(total=self.max_len)
+
+        if self.start_file_name:
+            self.pbar.update(self.index)
+            self.pbar.refresh()
+
+        return soup, request
+
+    def save_page_data(self, page_element):
+        time.sleep(0.3)     
+        doc_request = requests.get(MAIN_URL + page_element['href'])
+        assert doc_request.status_code == 200, f"Wrong status on requesting doc link: {MAIN_URL + page_element['href']}"
+        doc_soup = BeautifulSoup(doc_request.text, 'lxml')
+        text = doc_soup.find("div", {"class": "pagetext"}).next_element
+        image_url = doc_soup.find("div", {"class": "prp-page-image"}).find("img")['src']       
+
+        with open(self.output_file_name, 'a', newline='') as output_csv:
+            row_dict = {
+                "title": page_element['title'],
+                "href": MAIN_URL + page_element['href'],
+                "image_url": image_url, 
+                "text": text.text,
+                "index": self.index,
+                "page_number": self.page_number
+            }
+            fields = list(row_dict.keys())
+            writer = csv.DictWriter(output_csv, fieldnames=list(row_dict.keys()), delimiter='\t')
+            writer.writerow(row_dict)
+
+    def crawl(self):
+        soup, r = self._init_crawl()
+        first_search = True
+        while self.index < self.max_len:
+            time.sleep(0.3)
+            self.pbar.set_description(f"Page number: {self.page_number}")
+            next_page = soup.find("a", {"href": re.compile(r"\/w\/index.php.*")}, string="następna strona")
+            if next_page:
+                next_page = next_page.get('href', None)
+            if next_page and not first_search:
+                r = requests.get(MAIN_URL + next_page)
+            elif not next_page:
+                print(soup)
+                print("\n\n\n", soup.text)
+                print("End of pages, or next page not found")
+                break
+            
+            # handle wrong request
+            if r.status_code != 200:
+                print('Retry of request, request data: ', r.__dict__)
+                time.sleep(60)
+                r = requests.get(MAIN_URL + next_page)
+                assert r.status_code == 200, f"Retry failed, request status: {r.status_code}, full_info: {r.__dict__}"
+            
+            soup = BeautifulSoup(r.text, 'lxml')
+            links = soup.find_all("a", {"href": re.compile(r"\/wiki\/Strona:.*")})
+            for link in links:
+                self.save_page_data(link)
+                self.index += 1
+                self.pbar.update(1)
+            self.page_number += 1
+            first_search = False
+        print("Finished")
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--wiki_type", type=str, choices=["green", "yellow", "red"], required=True)
+    parser.add_argument("--output_file_name", type=str, required=True)
+    args, left_argv = parser.parse_known_args()
+
+    crawler = WikiCrawler(**vars(args))
+    crawler.crawl()

image_class.py

@@ -0,0 +1,91 @@
+import os
+import argparse
+import pandas as pd
+import requests
+from PIL import Image
+from tqdm import tqdm
+import pickle
+import time
+from pprint import pprint
+import json
+from datasets import load_dataset
+from huggingface_hub import login
+import shutil
+
+headers = {'User-Agent': 'ImageDownloadOcrBot/1.0 (no.rp.mk.info@gmail.com) requests/2.28.1'}
+
+class WikiImage:
+
+    def __init__(self, input_file_path: str, dataset_name: str,  output_folder: str = 'temp_images', split_number: int = 1):
+        self.input_file_path = input_file_path
+        self.split_number = split_number
+        self.max_dataset_len = 10000
+        self.output_folder = output_folder
+        self.dataset_name = dataset_name
+        print("Loading input file")
+        self.dataframe = pd.read_csv(self.input_file_path, sep='\t')[(self.split_number - 1) * self.max_dataset_len:]
+        if os.path.exists(self.output_folder):
+            print("Removing old dear")
+            if os.path.exists('/home/zombely/.cache/huggingface/datasets'):
+                shutil.rmtree('/home/zombely/.cache/huggingface/datasets')
+            shutil.rmtree(self.output_folder)
+        os.mkdir(self.output_folder)
+        self.pbar = tqdm(self.dataframe.iterrows(), total=len(self.dataframe), desc=f"Split: {self.split_number}")
+
+        login(os.environ.get("HUG_TOKEN"), True)
+
+    def image_save(self, row):
+        time.sleep(0.3)
+        image_request = requests.get(f"https:{row[1]['image_url']}", stream=True, headers=headers)
+        if image_request.status_code in [500, 404]:
+            print(f"Image {row[1]['title']} is not reacheable")
+            return
+        if image_request.status_code != 200:
+            time.sleep(80)
+            image_request = requests.get(f"https:{row[1]['image_url']}", stream=True, headers=headers)
+            assert image_request.status_code == 200, f"Response status is diffrent, status_code: {image_request.status_code}, full info: {image_request.__dict__}" 
+        
+        image = Image.open(image_request.raw)
+        if image.mode != "RGB":
+            image = image.convert("RGB")
+        title = row[1]['title'].replace("Strona:", "").replace("/", "-")
+        image.save(f"{self.output_folder}/{title}.png")
+
+        with open(f"{self.output_folder}/metadata.jsonl", mode='a', encoding='utf-8') as f:
+            # f.write(str({"file_name": f"{title}.png", "ground_truth": json.dumps({"gt_parse": {"text_sequance": row[1]['text'].replace('"', "'")}}, ensure_ascii=False)})+"\n")
+            json.dump({"file_name": f"{title}.png", "ground_truth": json.dumps({"gt_parse": {"text_sequance": row[1]['text'].replace('"', "'")}}, ensure_ascii=False)}, f, ensure_ascii=False)
+            f.write("\n")
+
+    def push_dataset(self, split_name: str):
+        print(f"Pushing split: {split_name}")
+        dataset = load_dataset(self.output_folder)
+        dataset[split_name] = dataset.pop('train')
+        dataset.push_to_hub(f'Zombely/{self.dataset_name}')
+        shutil.rmtree(self.output_folder)
+        shutil.rmtree('/home/zombely/.cache/huggingface/datasets')
+        os.mkdir(self.output_folder)
+        del dataset
+        print("Upload finished")
+
+    def crawl(self):
+        print("Start download")
+        for index, row in enumerate(self.pbar):
+            self.image_save(row)
+            if (index + 1) % self.max_dataset_len == 0:
+                self.push_dataset(f'train_{self.split_number}')
+                self.split_number += 1
+                self.pbar.set_description(f'Split: {self.split_number}')
+
+        self.push_dataset('validation')
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--input_file_path", type=str, required=True)
+    parser.add_argument("--dataset_name", type=str, required=True)
+    parser.add_argument("--output_folder", type=str, required=False, default='temp_images')
+    parser.add_argument("--split_number", type=int, required=False, default=1)
+    args, left_argv = parser.parse_known_args()
+    crawler = WikiImage(**vars(args))
+    crawler.crawl()
+

image_download.py

@@ -0,0 +1,79 @@
+import os
+import argparse
+import pandas as pd
+import requests
+from PIL import Image
+from tqdm import tqdm
+import pickle
+import time
+from pprint import pprint
+import json
+
+headers = {'User-Agent': 'ImageDownloadOcrBot/1.0 (micha9@op.pl) requests/2.28.1'}
+
+def save_state(index, offset):
+    with open("./state.pickle", "wb") as state_file:
+        pickle.dump({"row_index": index+offset}, state_file, protocol=pickle.HIGHEST_PROTOCOL)
+        print("Saving state, index: ", index+offset)
+
+def main(args):
+
+    df = pd.read_csv(args.file_path, sep="\t")
+    offset = 0 
+    if not os.path.exists(args.output_folder):
+        print(f"Creating missing folder: {args.output_folder}")
+        os.mkdir(args.output_folder)
+    
+    if args.from_checkpoint and os.path.exists("./state.pickle"):
+        with open("state.pickle", "rb") as state:
+            state_dict = pickle.load(state)
+            offset = state_dict["row_index"]
+            print("Starting from checkpoint, index: ", offset)
+            df = df[offset:]
+
+    pbar = tqdm(df.iterrows(), total=len(df), desc=f"0/{args.max_folder_size_mb if args.max_folder_size_mb else 'No limit given'} MB")
+    for n, row in enumerate(pbar):
+        try:
+            time.sleep(0.2)
+            r = requests.get(f"https:{row[1]['image_url']}", stream=True, headers=headers)
+            if r.status_code != 200:
+                time.sleep(80)
+                r = requests.get(f"https:{row[1]['image_url']}", stream=True, headers=headers)
+                if r.status_code != 200:
+                    pprint(r.__dict__)
+                    save_state(n, offset)
+                    return
+            image = Image.open(r.raw)
+            if image.mode != "RGB":
+                image = image.convert("RGB")
+            title = row[1]['title'].replace("Strona:", "").replace("/", "-")
+            image.save(f"{args.output_folder}/{title}.png")
+
+            with open(f"{args.output_folder}/metadata.jsonl", mode='a', encoding='utf-8') as f:
+                f.write(str({"file_name": f"{title}.png", "ground_truth": json.dumps({"gt_parse": {"text_sequance": row[1]['text']}}, ensure_ascii=False)}) + "\n")
+
+            dir_size = round(sum(os.path.getsize(f"./{args.output_folder}/{file}") for file in os.listdir(f"./{args.output_folder}")) * 0.000001, 2)
+
+            pbar.set_description(f"{dir_size}/{args.max_folder_size_mb if args.max_folder_size_mb else 'No limit given'} MB")
+
+            if args.max_folder_size_mb and dir_size > args.max_folder_size_mb:
+                print(f"Limit size of: {args.max_folder_size_mb}, exceeded")
+                save_state(n, offset) 
+                return
+
+        except (Exception, KeyboardInterrupt) as e:
+            print(f"Error: {str(e)} \n")
+            print(f"Row: {row}")
+            save_state(n, offset)
+            return
+
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--file_path", type=str, required=True)
+    parser.add_argument("--output_folder", type=str, default="./images")
+    parser.add_argument("--max_folder_size_mb", type=float, required=False)
+    parser.add_argument("--from_checkpoint", type=bool, required=False, default=False)
+    args, left_argv = parser.parse_known_args()
+    main(args)

mail_test.py

@@ -0,0 +1,8 @@
+import smtplib
+
+def main():
+    smtp = smtplib.SMTP("0.0.0.0", 25, 'mail')
+    smtp.sendmail('info@zbhome.com', ['michalkozlowski936@gmail.com'], "Hello from zbhome")
+
+if __name__ == "__main__":
+    main()

notebooks/crawler.ipynb

@@ -343,7 +343,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.15"
+   "version": "3.9.15 | packaged by conda-forge | (main, Nov 22 2022, 08:41:22) [MSC v.1929 64 bit (AMD64)]"
   },
   "orig_nbformat": 4,
   "vscode": {

notebooks/image_download.ipynb

@@ -0,0 +1,238 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "a = pd.read_csv(\"../../wikisource-data/yellow-continue-yellow.tsv.tsv\", sep=\"\\t\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>Unnamed: 0</th>\n",
+       "      <th>title</th>\n",
+       "      <th>href</th>\n",
+       "      <th>image_url</th>\n",
+       "      <th>text</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0</td>\n",
+       "      <td>Strona:Stanisław Antoni Wotowski - George Sand...</td>\n",
+       "      <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
+       "      <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
+       "      <td>zmieniła się; piękne oczy są tak samo błyszczą...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1</td>\n",
+       "      <td>Strona:Stanisław Antoni Wotowski - George Sand...</td>\n",
+       "      <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
+       "      <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
+       "      <td>najświetniejszej chociażby sławy... i po piętn...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>2</td>\n",
+       "      <td>Strona:Stanisław Antoni Wotowski - George Sand...</td>\n",
+       "      <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
+       "      <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
+       "      <td>Chopin gra. Ledwie dostrzegalnie muskają smuk...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>3</td>\n",
+       "      <td>Strona:Stanisław Antoni Wotowski - George Sand...</td>\n",
+       "      <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
+       "      <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
+       "      <td>\\nDZIWACZNE MAŁŻEŃSTWO.\\n\\n Był grudzień 1830 ...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>4</td>\n",
+       "      <td>Strona:Stanisław Antoni Wotowski - George Sand...</td>\n",
+       "      <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
+       "      <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
+       "      <td>Ale bliższego związku z panią Sand jakby się ...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   Unnamed: 0                                              title  \\\n",
+       "0           0  Strona:Stanisław Antoni Wotowski - George Sand...   \n",
+       "1           1  Strona:Stanisław Antoni Wotowski - George Sand...   \n",
+       "2           2  Strona:Stanisław Antoni Wotowski - George Sand...   \n",
+       "3           3  Strona:Stanisław Antoni Wotowski - George Sand...   \n",
+       "4           4  Strona:Stanisław Antoni Wotowski - George Sand...   \n",
+       "\n",
+       "                                                href  \\\n",
+       "0  https://pl.wikisource.org//wiki/Strona:Stanis%...   \n",
+       "1  https://pl.wikisource.org//wiki/Strona:Stanis%...   \n",
+       "2  https://pl.wikisource.org//wiki/Strona:Stanis%...   \n",
+       "3  https://pl.wikisource.org//wiki/Strona:Stanis%...   \n",
+       "4  https://pl.wikisource.org//wiki/Strona:Stanis%...   \n",
+       "\n",
+       "                                           image_url  \\\n",
+       "0  //upload.wikimedia.org/wikipedia/commons/thumb...   \n",
+       "1  //upload.wikimedia.org/wikipedia/commons/thumb...   \n",
+       "2  //upload.wikimedia.org/wikipedia/commons/thumb...   \n",
+       "3  //upload.wikimedia.org/wikipedia/commons/thumb...   \n",
+       "4  //upload.wikimedia.org/wikipedia/commons/thumb...   \n",
+       "\n",
+       "                                                text  \n",
+       "0  zmieniła się; piękne oczy są tak samo błyszczą...  \n",
+       "1  najświetniejszej chociażby sławy... i po piętn...  \n",
+       "2   Chopin gra. Ledwie dostrzegalnie muskają smuk...  \n",
+       "3  \\nDZIWACZNE MAŁŻEŃSTWO.\\n\\n Był grudzień 1830 ...  \n",
+       "4   Ale bliższego związku z panią Sand jakby się ...  "
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "a.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset\n",
+    "from huggingface_hub import login"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Resolving data files: 100%|██████████| 29/29 [00:00<?, ?it/s]\n",
+      "Using custom data configuration images-8b1ad802b6988161\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Downloading and preparing dataset imagefolder/images to C:/Users/PC/.cache/huggingface/datasets/imagefolder/images-8b1ad802b6988161/0.0.0/37fbb85cc714a338bea574ac6c7d0b5be5aff46c1862c1989b20e0771199e93f...\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Downloading data files: 0it [00:00, ?it/s]\n",
+      "Extracting data files: 0it [00:00, ?it/s]\n"
+     ]
+    },
+    {
+     "ename": "ArrowInvalid",
+     "evalue": "JSON parse error: Missing a name for object member. in row 0",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mArrowInvalid\u001b[0m                              Traceback (most recent call last)",
+      "Cell \u001b[1;32mIn[6], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m dataset \u001b[39m=\u001b[39m load_dataset(\u001b[39m\"\u001b[39;49m\u001b[39m../images\u001b[39;49m\u001b[39m\"\u001b[39;49m)\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\datasets\\load.py:1741\u001b[0m, in \u001b[0;36mload_dataset\u001b[1;34m(path, name, data_dir, data_files, split, cache_dir, features, download_config, download_mode, ignore_verifications, keep_in_memory, save_infos, revision, use_auth_token, task, streaming, num_proc, **config_kwargs)\u001b[0m\n\u001b[0;32m   1738\u001b[0m try_from_hf_gcs \u001b[39m=\u001b[39m path \u001b[39mnot\u001b[39;00m \u001b[39min\u001b[39;00m _PACKAGED_DATASETS_MODULES\n\u001b[0;32m   1740\u001b[0m \u001b[39m# Download and prepare data\u001b[39;00m\n\u001b[1;32m-> 1741\u001b[0m builder_instance\u001b[39m.\u001b[39;49mdownload_and_prepare(\n\u001b[0;32m   1742\u001b[0m     download_config\u001b[39m=\u001b[39;49mdownload_config,\n\u001b[0;32m   1743\u001b[0m     download_mode\u001b[39m=\u001b[39;49mdownload_mode,\n\u001b[0;32m   1744\u001b[0m     ignore_verifications\u001b[39m=\u001b[39;49mignore_verifications,\n\u001b[0;32m   1745\u001b[0m     try_from_hf_gcs\u001b[39m=\u001b[39;49mtry_from_hf_gcs,\n\u001b[0;32m   1746\u001b[0m     use_auth_token\u001b[39m=\u001b[39;49muse_auth_token,\n\u001b[0;32m   1747\u001b[0m     num_proc\u001b[39m=\u001b[39;49mnum_proc,\n\u001b[0;32m   1748\u001b[0m )\n\u001b[0;32m   1750\u001b[0m \u001b[39m# Build dataset for splits\u001b[39;00m\n\u001b[0;32m   1751\u001b[0m keep_in_memory \u001b[39m=\u001b[39m (\n\u001b[0;32m   1752\u001b[0m     keep_in_memory \u001b[39mif\u001b[39;00m keep_in_memory \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m \u001b[39melse\u001b[39;00m is_small_dataset(builder_instance\u001b[39m.\u001b[39minfo\u001b[39m.\u001b[39mdataset_size)\n\u001b[0;32m   1753\u001b[0m )\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\datasets\\builder.py:822\u001b[0m, in \u001b[0;36mDatasetBuilder.download_and_prepare\u001b[1;34m(self, output_dir, download_config, download_mode, ignore_verifications, try_from_hf_gcs, dl_manager, base_path, use_auth_token, file_format, max_shard_size, num_proc, storage_options, **download_and_prepare_kwargs)\u001b[0m\n\u001b[0;32m    820\u001b[0m     \u001b[39mif\u001b[39;00m num_proc \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m:\n\u001b[0;32m    821\u001b[0m         prepare_split_kwargs[\u001b[39m\"\u001b[39m\u001b[39mnum_proc\u001b[39m\u001b[39m\"\u001b[39m] \u001b[39m=\u001b[39m num_proc\n\u001b[1;32m--> 822\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_download_and_prepare(\n\u001b[0;32m    823\u001b[0m         dl_manager\u001b[39m=\u001b[39mdl_manager,\n\u001b[0;32m    824\u001b[0m         verify_infos\u001b[39m=\u001b[39mverify_infos,\n\u001b[0;32m    825\u001b[0m         \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mprepare_split_kwargs,\n\u001b[0;32m    826\u001b[0m         \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mdownload_and_prepare_kwargs,\n\u001b[0;32m    827\u001b[0m     )\n\u001b[0;32m    828\u001b[0m \u001b[39m# Sync info\u001b[39;00m\n\u001b[0;32m    829\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39minfo\u001b[39m.\u001b[39mdataset_size \u001b[39m=\u001b[39m \u001b[39msum\u001b[39m(split\u001b[39m.\u001b[39mnum_bytes \u001b[39mfor\u001b[39;00m split \u001b[39min\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39minfo\u001b[39m.\u001b[39msplits\u001b[39m.\u001b[39mvalues())\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\datasets\\builder.py:1555\u001b[0m, in \u001b[0;36mGeneratorBasedBuilder._download_and_prepare\u001b[1;34m(self, dl_manager, verify_infos, **prepare_splits_kwargs)\u001b[0m\n\u001b[0;32m   1554\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m_download_and_prepare\u001b[39m(\u001b[39mself\u001b[39m, dl_manager, verify_infos, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mprepare_splits_kwargs):\n\u001b[1;32m-> 1555\u001b[0m     \u001b[39msuper\u001b[39m()\u001b[39m.\u001b[39m_download_and_prepare(\n\u001b[0;32m   1556\u001b[0m         dl_manager, verify_infos, check_duplicate_keys\u001b[39m=\u001b[39mverify_infos, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mprepare_splits_kwargs\n\u001b[0;32m   1557\u001b[0m     )\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\datasets\\builder.py:891\u001b[0m, in \u001b[0;36mDatasetBuilder._download_and_prepare\u001b[1;34m(self, dl_manager, verify_infos, **prepare_split_kwargs)\u001b[0m\n\u001b[0;32m    889\u001b[0m split_dict \u001b[39m=\u001b[39m SplitDict(dataset_name\u001b[39m=\u001b[39m\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mname)\n\u001b[0;32m    890\u001b[0m split_generators_kwargs \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_make_split_generators_kwargs(prepare_split_kwargs)\n\u001b[1;32m--> 891\u001b[0m split_generators \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_split_generators(dl_manager, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39msplit_generators_kwargs)\n\u001b[0;32m    893\u001b[0m \u001b[39m# Checksums verification\u001b[39;00m\n\u001b[0;32m    894\u001b[0m \u001b[39mif\u001b[39;00m verify_infos \u001b[39mand\u001b[39;00m dl_manager\u001b[39m.\u001b[39mrecord_checksums:\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\datasets\\packaged_modules\\folder_based_builder\\folder_based_builder.py:189\u001b[0m, in \u001b[0;36mFolderBasedBuilder._split_generators\u001b[1;34m(self, dl_manager)\u001b[0m\n\u001b[0;32m    186\u001b[0m metadata_ext \u001b[39m=\u001b[39m metadata_ext\u001b[39m.\u001b[39mpop()\n\u001b[0;32m    188\u001b[0m \u001b[39mfor\u001b[39;00m _, downloaded_metadata_file \u001b[39min\u001b[39;00m itertools\u001b[39m.\u001b[39mchain\u001b[39m.\u001b[39mfrom_iterable(metadata_files\u001b[39m.\u001b[39mvalues()):\n\u001b[1;32m--> 189\u001b[0m     pa_metadata_table \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_read_metadata(downloaded_metadata_file)\n\u001b[0;32m    190\u001b[0m     features_per_metadata_file\u001b[39m.\u001b[39mappend(\n\u001b[0;32m    191\u001b[0m         (downloaded_metadata_file, datasets\u001b[39m.\u001b[39mFeatures\u001b[39m.\u001b[39mfrom_arrow_schema(pa_metadata_table\u001b[39m.\u001b[39mschema))\n\u001b[0;32m    192\u001b[0m     )\n\u001b[0;32m    193\u001b[0m \u001b[39mfor\u001b[39;00m downloaded_metadata_file, metadata_features \u001b[39min\u001b[39;00m features_per_metadata_file:\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\datasets\\packaged_modules\\folder_based_builder\\folder_based_builder.py:260\u001b[0m, in \u001b[0;36mFolderBasedBuilder._read_metadata\u001b[1;34m(self, metadata_file)\u001b[0m\n\u001b[0;32m    258\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[0;32m    259\u001b[0m     \u001b[39mwith\u001b[39;00m \u001b[39mopen\u001b[39m(metadata_file, \u001b[39m\"\u001b[39m\u001b[39mrb\u001b[39m\u001b[39m\"\u001b[39m) \u001b[39mas\u001b[39;00m f:\n\u001b[1;32m--> 260\u001b[0m         \u001b[39mreturn\u001b[39;00m paj\u001b[39m.\u001b[39;49mread_json(f)\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\pyarrow\\_json.pyx:259\u001b[0m, in \u001b[0;36mpyarrow._json.read_json\u001b[1;34m()\u001b[0m\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\pyarrow\\error.pxi:144\u001b[0m, in \u001b[0;36mpyarrow.lib.pyarrow_internal_check_status\u001b[1;34m()\u001b[0m\n",
+      "File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\pyarrow\\error.pxi:100\u001b[0m, in \u001b[0;36mpyarrow.lib.check_status\u001b[1;34m()\u001b[0m\n",
+      "\u001b[1;31mArrowInvalid\u001b[0m: JSON parse error: Missing a name for object member. in row 0"
+     ]
+    }
+   ],
+   "source": [
+    "dataset = load_dataset(\"../images\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "login('',True)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "um",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.15"
+  },
+  "orig_nbformat": 4,
+  "vscode": {
+   "interpreter": {
+    "hash": "876e189cbbe99a9a838ece62aae1013186c4bb7e0254a10cfa2f9b2381853efb"
+   }
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

notebooks/join.ipynb

@@ -0,0 +1,95 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "green = pd.read_csv(\"../../wikisource-data/green.tsv\", sep=\"\\t\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "green.tail()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "green = pd.read_csv(\"../green-full.tsv\", sep=\"\\t\")\n",
+    "yellow = pd.read_csv(\"../yellow-full.tsv\", sep=\"\\t\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "whole = pd.concat([green, yellow], axis=0)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "len(whole)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "whole.to_csv(\"./wikisource-full.tsv\", sep=\"\\t\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "um",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.9.15"
+  },
+  "orig_nbformat": 4,
+  "vscode": {
+   "interpreter": {
+    "hash": "876e189cbbe99a9a838ece62aae1013186c4bb7e0254a10cfa2f9b2381853efb"
+   }
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

requirements.txt

@@ -5,6 +5,7 @@ idna==3.4
 lxml==4.9.2
 numpy==1.24.1
 pandas==1.5.2
+Pillow==9.4.0
 python-dateutil==2.8.2
 pytz==2022.7
 requests==2.28.1