s444415/wikisource-crawler
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
9ee285cf95
wikisource-crawler/image_download.ipynb

1043 lines
1.6 MiB
Plaintext
Raw Normal View History

image download script
2023-01-07 14:19:01 +01:00
{
"cells": [
{
"cell_type": "code",
"execution_count": 65,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import requests\n",
"from PIL import Image\n",
"import os"
]
},
{
"cell_type": "code",
"execution_count": 100,
"metadata": {},
"outputs": [],
"source": [
"session = requests.Session()\n",
"session.headers.update({'User-Agent': 'Mozilla/5.0'})"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"df = pd.read_csv(\"../wikisource-data/yellow.tsv\", sep=\"\\t\")"
]
},
{
"cell_type": "code",
"execution_count": 75,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Strona:A. Kuprin - Straszna chwila.djvu/8\n",
"Strona:Abramowski-braterstwo-solidarnosc-wspoldzialanie.pdf/17\n",
"Strona:Abramowski-braterstwo-solidarnosc-wspoldzialanie.pdf/18\n",
"Strona:Abramowski-braterstwo-solidarnosc-wspoldzialanie.pdf/19\n"
]
}
],
"source": [
"for i in df[6:10].iterrows():\n",
" print(i[1]['title'])"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"94945"
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(df)"
]
},
{
"cell_type": "code",
"execution_count": 40,
"metadata": {},
"outputs": [],
"source": [
"a = df.head()"
]
},
{
"cell_type": "code",
"execution_count": 25,
"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>94940</th>\n",
" <td>94940</td>\n",
" <td>Strona:Stanisław Vincenz - Na wysokiej połonin...</td>\n",
" <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
" <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
" <td>Wysoko po drugiej stronie drogi, sterczące z ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>94941</th>\n",
" <td>94941</td>\n",
" <td>Strona:Stanisław Vincenz - Na wysokiej połonin...</td>\n",
" <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
" <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
" <td>tępny także dla ludzi biednych, nieuczonych w ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>94942</th>\n",
" <td>94942</td>\n",
" <td>Strona:Stanisław Vincenz - Na wysokiej połonin...</td>\n",
" <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
" <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
" <td>brego, Baal Szem Tow. Tam mieszkał przez lata ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>94943</th>\n",
" <td>94943</td>\n",
" <td>Strona:Stanisław Vincenz - Na wysokiej połonin...</td>\n",
" <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
" <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
" <td>Z Sokolskiej góry widok rozległy. Jest jakby ...</td>\n",
" </tr>\n",
" <tr>\n",
" <th>94944</th>\n",
" <td>94944</td>\n",
" <td>Strona:Stanisław Vincenz - Na wysokiej połonin...</td>\n",
" <td>https://pl.wikisource.org//wiki/Strona:Stanis%...</td>\n",
" <td>//upload.wikimedia.org/wikipedia/commons/thumb...</td>\n",
" <td>\\r\\n\\r\\nIII. PODRÓŻ BABCI\\r\\n W ciepły pogodny...</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Unnamed: 0 title \\\n",
"94940 94940 Strona:Stanisław Vincenz - Na wysokiej połonin... \n",
"94941 94941 Strona:Stanisław Vincenz - Na wysokiej połonin... \n",
"94942 94942 Strona:Stanisław Vincenz - Na wysokiej połonin... \n",
"94943 94943 Strona:Stanisław Vincenz - Na wysokiej połonin... \n",
"94944 94944 Strona:Stanisław Vincenz - Na wysokiej połonin... \n",
"\n",
" href \\\n",
"94940 https://pl.wikisource.org//wiki/Strona:Stanis%... \n",
"94941 https://pl.wikisource.org//wiki/Strona:Stanis%... \n",
"94942 https://pl.wikisource.org//wiki/Strona:Stanis%... \n",
"94943 https://pl.wikisource.org//wiki/Strona:Stanis%... \n",
"94944 https://pl.wikisource.org//wiki/Strona:Stanis%... \n",
"\n",
" image_url \\\n",
"94940 //upload.wikimedia.org/wikipedia/commons/thumb... \n",
"94941 //upload.wikimedia.org/wikipedia/commons/thumb... \n",
"94942 //upload.wikimedia.org/wikipedia/commons/thumb... \n",
"94943 //upload.wikimedia.org/wikipedia/commons/thumb... \n",
"94944 //upload.wikimedia.org/wikipedia/commons/thumb... \n",
"\n",
" text \n",
"94940  Wysoko po drugiej stronie drogi, sterczące z ... \n",
"94941 tępny także dla ludzi biednych, nieuczonych w ... \n",
"94942 brego, Baal Szem Tow. Tam mieszkał przez lata ... \n",
"94943  Z Sokolskiej góry widok rozległy. Jest jakby ... \n",
"94944 \\r\\n\\r\\nIII. PODRÓŻ BABCI\\r\\n W ciepły pogodny... "
]
},
"execution_count": 25,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a"
]
},
{
"cell_type": "code",
"execution_count": 99,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"'//upload.wikimedia.org/wikipedia/commons/thumb/1/17/A._Kuprin_-_Straszna_chwila.djvu/page110-1024px-A._Kuprin_-_Straszna_chwila.djvu.jpg'"
]
},
"execution_count": 99,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a['image_url'][4]"
]
},
{
"cell_type": "code",
"execution_count": 104,
"metadata": {},
"outputs": [],
"source": [
"r = requests.get(f\"https:{a['image_url'][3]}\")"
]
},
{
"cell_type": "code",
"execution_count": 105,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'_content': b'<!DOCTYPE html>\\n<html lang=\"en\">\\n<meta charset=\"utf-8\">\\n<title>Wikimedia Error</title>\\n<style>\\n* { margin: 0; padding: 0; }\\nbody { background: #fff; font: 15px/1.6 sans-serif; color: #333; }\\n.content { margin: 7% auto 0; padding: 2em 1em 1em; max-width: 640px; }\\n.footer { clear: both; margin-top: 14%; border-top: 1px solid #e5e5e5; background: #f9f9f9; padding: 2em 0; font-size: 0.8em; text-align: center; }\\nimg { float: left; margin: 0 2em 2em 0; }\\na img { border: 0; }\\nh1 { margin-top: 1em; font-size: 1.2em; }\\n.content-text { overflow: hidden; overflow-wrap: break-word; word-wrap: break-word; -webkit-hyphens: auto; -moz-hyphens: auto; -ms-hyphens: auto; hyphens: auto; }\\np { margin: 0.7em 0 1em 0; }\\na { color: #0645ad; text-decoration: none; }\\na:hover { text-decoration: underline; }\\ncode { font-family: sans-serif; }\\n.text-muted { color: #777; }\\n</style>\\n<div class=\"content\" role=\"main\">\\n<a href=\"https://www.wikimedia.org\"><img src=\"https://www.wikimedia.org/static/images/wmf-logo.png\" srcset=\"https://www.wikimedia.org/static/images/wmf-logo-2x.png 2x\" alt=\"Wikimedia\" width=\"135\" height=\"101\">\\n</a>\\n<h1>Error</h1>\\n<div class=\"content-text\">\\n<p>Our servers are currently under maintenance or experiencing a technical problem.\\n\\nPlease <a href=\"\" title=\"Reload this page\" onclick=\"window.location.reload(false); return false\">try again</a> in a few&nbsp;minutes.</p>\\n\\n<p>See the error message at the bottom of this page for more&nbsp;information.</p>\\n</div>\\n</div>\\n<div class=\"footer\"><p>If you report this error to the Wikimedia System Administrators, please include the details below.</p><p class=\"text-muted\"><code>Request from 83.8.149.228 via cp3063 cp3063, Varnish XID 214100233<br>Upstream caches: cp3063 int<br>Error: 403, Forbidden. Please comply with the User-Agent policy: https://meta.wikimedia.org/wiki/User-Agent_policy at Sat, 07 Jan 2023 13:09:42 GMT</code></p>\\n</div>\\n</html>\\n',\n",
" '_content_consumed': True,\n",
" '_next': None,\n",
" 'status_code': 403,\n",
" 'headers': {'date': 'Sat, 07 Jan 2023 13:09:42 GMT', 'server': 'Varnish', 'x-cache': 'cp3063 int', 'x-cache-status': 'int-front', 'server-timing': 'cache;desc=\"int-front\", host;desc=\"cp3063\"', 'strict-transport-security': 'max-age=106384710; includeSubDomains; preload', 'report-to': '{ \"group\": \"wm_nel\", \"max_age\": 86400, \"endpoints\": [{ \"url\": \"https://intake-logging.wikimedia.org/v1/events?stream=w3c.reportingapi.network_error&schema_uri=/w3c/reportingapi/network_error/1.0.0\" }] }', 'nel': '{ \"report_to\": \"wm_nel\", \"max_age\": 86400, \"failure_fraction\": 0.05, \"success_fraction\": 0.0}', 'accept-ch': 'Sec-CH-UA-Arch,Sec-CH-UA-Bitness,Sec-CH-UA-Full-Version-List,Sec-CH-UA-Model,Sec-CH-UA-Platform-Version', 'permissions-policy': 'interest-cohort=(),ch-ua-arch=(self \"intake-analytics.wikimedia.org\"),ch-ua-bitness=(self \"intake-analytics.wikimedia.org\"),ch-ua-full-version-list=(self \"intake-analytics.wikimedia.org\"),ch-ua-model=(self \"intake-analytics.wikimedia.org\"),ch-ua-platform-version=(self \"intake-analytics.wikimedia.org\")', 'x-client-ip': '83.8.149.228', 'content-type': 'text/html; charset=utf-8', 'content-length': '1905'},\n",
" 'raw': <urllib3.response.HTTPResponse at 0x21da81ca910>,\n",
" 'url': 'https://upload.wikimedia.org/wikipedia/commons/thumb/3/33/A._Kuprin_-_Mi%C5%82o%C5%9B%C4%87_Sulamity.djvu/page9-1024px-A._Kuprin_-_Mi%C5%82o%C5%9B%C4%87_Sulamity.djvu.jpg',\n",
" 'encoding': 'utf-8',\n",
" 'history': [],\n",
" 'reason': 'Forbidden. Please comply with the User-Agent policy: https://meta.wikimedia.org/wiki/User-Agent_policy',\n",
" 'cookies': <RequestsCookieJar[]>,\n",
" 'elapsed': datetime.timedelta(microseconds=636314),\n",
" 'request': <PreparedRequest [GET]>,\n",
" 'connection': <requests.adapters.HTTPAdapter at 0x21da81ca040>}"
]
},
"execution_count": 105,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"r.__dict__"
]
},
{
"cell_type": "code",
"execution_count": 95,
"metadata": {},
"outputs": [
{
"ename": "UnidentifiedImageError",
"evalue": "cannot identify image file <_io.BytesIO object at 0x0000021DA1EA14F0>",
"output_type": "error",
"traceback": [
"\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[1;31mUnidentifiedImageError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[1;32mIn[95], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m b \u001b[39m=\u001b[39m Image\u001b[39m.\u001b[39;49mopen(requests\u001b[39m.\u001b[39;49mget(\u001b[39mf\u001b[39;49m\u001b[39m\"\u001b[39;49m\u001b[39mhttps:\u001b[39;49m\u001b[39m{\u001b[39;49;00ma[\u001b[39m'\u001b[39;49m\u001b[39mimage_url\u001b[39;49m\u001b[39m'\u001b[39;49m][\u001b[39m4\u001b[39;49m]\u001b[39m}\u001b[39;49;00m\u001b[39m\"\u001b[39;49m, stream\u001b[39m=\u001b[39;49m\u001b[39mTrue\u001b[39;49;00m)\u001b[39m.\u001b[39;49mraw)\n",
"File \u001b[1;32mc:\\Users\\PC\\anaconda3\\envs\\um\\lib\\site-packages\\PIL\\Image.py:3147\u001b[0m, in \u001b[0;36mopen\u001b[1;34m(fp, mode, formats)\u001b[0m\n\u001b[0;32m 3145\u001b[0m \u001b[39mfor\u001b[39;00m message \u001b[39min\u001b[39;00m accept_warnings:\n\u001b[0;32m 3146\u001b[0m warnings\u001b[39m.\u001b[39mwarn(message)\n\u001b[1;32m-> 3147\u001b[0m \u001b[39mraise\u001b[39;00m UnidentifiedImageError(\n\u001b[0;32m 3148\u001b[0m \u001b[39m\"\u001b[39m\u001b[39mcannot identify image file \u001b[39m\u001b[39m%r\u001b[39;00m\u001b[39m\"\u001b[39m \u001b[39m%\u001b[39m (filename \u001b[39mif\u001b[39;00m filename \u001b[39melse\u001b[39;00m fp)\n\u001b[0;32m 3149\u001b[0m )\n",
"\u001b[1;31mUnidentifiedImageError\u001b[0m: cannot identify image file <_io.BytesIO object at 0x0000021DA1EA14F0>"
]
}
],
"source": [
"b = Image.open(requests.get(f\"https:{a['image_url'][4]}\", stream=True).raw)"
]
},
{
"cell_type": "code",
"execution_count": 83,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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
"text/plain": [
"<PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=1024x1478>"
]
},
"execution_count": 83,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"b"
]
},
{
"cell_type": "code",
"execution_count": 59,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"'A. Baranowski - O wzorach.pdf-7'"
]
},
"execution_count": 59,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"a[\"title\"][0].replace(\"Strona:\", \"\").replace(\"/\", \"-\")"
]
},
{
"cell_type": "code",
"execution_count": 67,
"metadata": {},
"outputs": [],
"source": [
"b.save(f'./images/{a[\"title\"][0].replace(\"Strona:\", \"\").replace(\"/\", \"-\")}2.jpg')"
]
},
{
"cell_type": "code",
"execution_count": 68,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.18"
]
},
"execution_count": 68,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"round(sum(os.path.getsize(f\"./images/{file}\") for file in os.listdir(\"./images\")) * 0.000001, 2)"
]
},
{
"cell_type": "code",
"execution_count": 39,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"//upload.wikimedia.org/wikipedia/commons/thumb/7/7e/A._Baranowski_-_O_wzorach.pdf/page7-1024px-A._Baranowski_-_O_wzorach.pdf.jpg\n",
"//upload.wikimedia.org/wikipedia/commons/thumb/7/7e/A._Baranowski_-_O_wzorach.pdf/page8-1024px-A._Baranowski_-_O_wzorach.pdf.jpg\n",
"//upload.wikimedia.org/wikipedia/commons/thumb/3/33/A._Kuprin_-_Mi%C5%82o%C5%9B%C4%87_Sulamity.djvu/page119-1024px-A._Kuprin_-_Mi%C5%82o%C5%9B%C4%87_Sulamity.djvu.jpg\n",
"//upload.wikimedia.org/wikipedia/commons/thumb/3/33/A._Kuprin_-_Mi%C5%82o%C5%9B%C4%87_Sulamity.djvu/page9-1024px-A._Kuprin_-_Mi%C5%82o%C5%9B%C4%87_Sulamity.djvu.jpg\n",
"//upload.wikimedia.org/wikipedia/commons/thumb/1/17/A._Kuprin_-_Straszna_chwila.djvu/page110-1024px-A._Kuprin_-_Straszna_chwila.djvu.jpg\n"
]
}
],
"source": [
"for row in list(df.head().iterrows()):\n",
" print(row[1]['image_url'])"
]
},
{
"cell_type": "code",
"execution_count": 79,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"^C\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\n",
"0it [00:00, ?it/s]\n",
"1it [00:01, 1.26s/it]\n",
"2it [00:02, 1.00it/s]\n",
"3it [00:02, 1.24it/s]\n",
"4it [00:03, 1.40it/s]\n",
"5it [00:03, 1.51it/s]\n",
"6it [00:04, 1.58it/s]\n",
"7it [00:04, 1.63it/s]\n",
"8it [00:05, 1.64it/s]\n",
"9it [00:06, 1.67it/s]\n",
"10it [00:06, 1.69it/s]\n",
"11it [00:07, 1.71it/s]\n",
"12it [00:07, 1.72it/s]\n",
"13it [00:08, 1.73it/s]\n",
"14it [00:08, 1.74it/s]\n",
"15it [00:09, 1.74it/s]\n",
"16it [00:10, 1.75it/s]\n",
"17it [00:10, 1.75it/s]\n",
"18it [00:11, 1.71it/s]\n",
"19it [00:11, 1.67it/s]\n",
"20it [00:12, 1.65it/s]\n",
"21it [00:13, 1.66it/s]\n",
"22it [00:13, 1.60it/s]\n",
"23it [00:14, 1.57it/s]\n",
"24it [00:15, 1.59it/s]\n",
"25it [00:15, 1.60it/s]\n",
"26it [00:16, 1.65it/s]\n",
"27it [00:16, 1.69it/s]\n",
"28it [00:17, 1.70it/s]\n",
"29it [00:18, 1.70it/s]\n",
"30it [00:18, 1.70it/s]\n",
"31it [00:19, 1.68it/s]\n",
"32it [00:19, 1.65it/s]\n",
"33it [00:20, 1.68it/s]\n",
"34it [00:21, 1.67it/s]\n",
"35it [00:21, 1.67it/s]\n",
"36it [00:22, 1.68it/s]\n",
"37it [00:22, 1.67it/s]\n",
"38it [00:23, 1.68it/s]\n",
"39it [00:24, 1.57it/s]\n",
"40it [00:24, 1.49it/s]\n",
"41it [00:25, 1.45it/s]\n",
"42it [00:26, 1.52it/s]\n",
"43it [00:26, 1.44it/s]\n",
"44it [00:27, 1.46it/s]\n",
"45it [00:28, 1.51it/s]\n",
"46it [00:28, 1.55it/s]\n",
"47it [00:29, 1.46it/s]\n",
"48it [00:30, 1.50it/s]\n",
"49it [00:30, 1.54it/s]\n",
"50it [00:31, 1.60it/s]\n",
"51it [00:32, 1.59it/s]\n",
"52it [00:32, 1.54it/s]\n",
"53it [00:33, 1.57it/s]\n",
"54it [00:33, 1.61it/s]\n",
"55it [00:34, 1.62it/s]\n",
"56it [00:35, 1.65it/s]\n",
"57it [00:35, 1.69it/s]\n",
"58it [00:36, 1.71it/s]\n",
"59it [00:36, 1.72it/s]\n",
"60it [00:37, 1.73it/s]\n",
"61it [00:38, 1.72it/s]\n",
"62it [00:38, 1.72it/s]\n",
"63it [00:39, 1.71it/s]\n",
"64it [00:39, 1.70it/s]\n",
"65it [00:40, 1.68it/s]\n",
"66it [00:40, 1.70it/s]\n",
"67it [00:41, 1.68it/s]\n",
"68it [00:42, 1.64it/s]\n",
"69it [00:42, 1.62it/s]\n",
"70it [00:43, 1.62it/s]\n",
"71it [00:44, 1.67it/s]\n",
"72it [00:44, 1.57it/s]\n",
"73it [00:45, 1.47it/s]\n",
"74it [00:46, 1.54it/s]\n",
"75it [00:46, 1.61it/s]\n",
"76it [00:47, 1.67it/s]\n",
"77it [00:47, 1.65it/s]\n",
"78it [00:48, 1.66it/s]\n",
"79it [00:49, 1.60it/s]\n",
"80it [00:49, 1.65it/s]\n",
"81it [00:50, 1.67it/s]\n",
"82it [00:50, 1.70it/s]\n",
"83it [00:51, 1.74it/s]\n",
"84it [00:51, 1.73it/s]\n",
"85it [00:52, 1.70it/s]\n",
"86it [00:53, 1.73it/s]\n",
"87it [00:53, 1.75it/s]\n",
"88it [00:54, 1.75it/s]\n",
"89it [00:54, 1.77it/s]\n",
"90it [00:55, 1.78it/s]\n",
"91it [00:55, 1.79it/s]\n",
"92it [00:56, 1.80it/s]\n",
"93it [00:56, 1.81it/s]\n",
"94it [00:57, 1.81it/s]\n",
"95it [00:58, 1.78it/s]\n",
"96it [00:58, 1.79it/s]\n",
"97it [00:59, 1.78it/s]\n",
"98it [00:59, 1.77it/s]\n",
"99it [01:00, 1.78it/s]\n",
"100it [01:00, 1.79it/s]\n",
"101it [01:01, 1.80it/s]\n",
"102it [01:02, 1.81it/s]\n",
"103it [01:02, 1.81it/s]\n",
"104it [01:03, 1.82it/s]\n",
"105it [01:03, 1.80it/s]\n",
"106it [01:04, 1.80it/s]\n",
"107it [01:04, 1.77it/s]\n",
"108it [01:05, 1.79it/s]\n",
"109it [01:05, 1.78it/s]\n",
"110it [01:06, 1.79it/s]\n",
"111it [01:07, 1.80it/s]\n",
"112it [01:07, 1.81it/s]\n",
"113it [01:08, 1.82it/s]\n",
"114it [01:08, 1.79it/s]\n",
"115it [01:09, 1.80it/s]\n",
"116it [01:09, 1.81it/s]\n",
"117it [01:10, 1.82it/s]\n",
"118it [01:10, 1.79it/s]\n",
"119it [01:11, 1.78it/s]\n",
"120it [01:12, 1.79it/s]\n",
"121it [01:12, 1.80it/s]\n",
"122it [01:13, 1.81it/s]\n",
"123it [01:13, 1.81it/s]\n",
"124it [01:14, 1.81it/s]\n",
"125it [01:14, 1.81it/s]\n",
"126it [01:15, 1.81it/s]\n",
"127it [01:15, 1.81it/s]\n",
"128it [01:16, 1.82it/s]\n",
"129it [01:16, 1.82it/s]\n",
"130it [01:17, 1.82it/s]\n",
"131it [01:18, 1.82it/s]\n",
"132it [01:18, 1.82it/s]\n",
"133it [01:19, 1.79it/s]\n",
"134it [01:19, 1.80it/s]\n",
"135it [01:20, 1.78it/s]\n",
"136it [01:20, 1.76it/s]\n",
"137it [01:21, 1.78it/s]\n",
"138it [01:22, 1.79it/s]\n",
"139it [01:22, 1.78it/s]\n",
"140it [01:23, 1.76it/s]\n",
"141it [01:23, 1.78it/s]\n",
"142it [01:24, 1.80it/s]\n",
"143it [01:24, 1.80it/s]\n",
"144it [01:25, 1.78it/s]\n",
"145it [01:25, 1.80it/s]\n",
"146it [01:26, 1.80it/s]\n",
"147it [01:27, 1.81it/s]\n",
"148it [01:27, 1.78it/s]\n",
"149it [01:28, 1.80it/s]\n",
"150it [01:28, 1.80it/s]\n",
"151it [01:29, 1.79it/s]\n",
"152it [01:29, 1.78it/s]\n",
"153it [01:30, 1.79it/s]\n",
"154it [01:30, 1.80it/s]\n",
"155it [01:31, 1.80it/s]\n",
"156it [01:32, 1.81it/s]\n",
"157it [01:32, 1.81it/s]\n",
"158it [01:33, 1.79it/s]\n",
"159it [01:33, 1.76it/s]\n",
"160it [01:34, 1.78it/s]\n",
"161it [01:34, 1.79it/s]\n",
"162it [01:35, 1.80it/s]\n",
"163it [01:35, 1.81it/s]\n",
"164it [01:36, 1.82it/s]\n",
"165it [01:37, 1.82it/s]\n",
"166it [01:37, 1.79it/s]\n",
"167it [01:38, 1.78it/s]\n",
"168it [01:38, 1.79it/s]\n",
"169it [01:39, 1.80it/s]\n",
"170it [01:39, 1.81it/s]\n",
"171it [01:40, 1.77it/s]\n",
"172it [01:40, 1.79it/s]\n",
"173it [01:41, 1.80it/s]\n",
"174it [01:42, 1.81it/s]\n",
"175it [01:42, 1.81it/s]\n",
"176it [01:43, 1.82it/s]\n",
"177it [01:43, 1.82it/s]\n",
"178it [01:44, 1.81it/s]\n",
"179it [01:44, 1.81it/s]\n",
"180it [01:45, 1.81it/s]\n",
"181it [01:45, 1.82it/s]\n",
"182it [01:46, 1.78it/s]\n",
"183it [01:47, 1.80it/s]\n",
"184it [01:47, 1.81it/s]\n",
"185it [01:48, 1.81it/s]\n",
"186it [01:48, 1.82it/s]\n",
"187it [01:49, 1.82it/s]\n",
"188it [01:49, 1.79it/s]\n",
"189it [01:50, 1.80it/s]\n",
"190it [01:50, 1.79it/s]\n",
"191it [01:51, 1.80it/s]\n",
"192it [01:52, 1.80it/s]\n",
"193it [01:52, 1.81it/s]\n",
"194it [01:53, 1.80it/s]\n",
"195it [01:53, 1.80it/s]\n",
"196it [01:54, 1.81it/s]\n",
"197it [01:54, 1.81it/s]\n",
"198it [01:55, 1.81it/s]\n",
"199it [01:55, 1.82it/s]\n",
"200it [01:56, 1.81it/s]\n",
"201it [01:57, 1.82it/s]\n",
"202it [01:57, 1.80it/s]\n",
"203it [01:58, 1.78it/s]\n",
"204it [01:58, 1.78it/s]\n",
"205it [01:59, 1.79it/s]\n",
"206it [01:59, 1.80it/s]\n",
"207it [02:00, 1.81it/s]\n",
"208it [02:00, 1.79it/s]\n",
"209it [02:01, 1.80it/s]\n",
"210it [02:02, 1.81it/s]\n",
"211it [02:02, 1.81it/s]\n",
"212it [02:03, 1.80it/s]\n",
"213it [02:03, 1.81it/s]\n",
"214it [02:04, 1.81it/s]\n",
"215it [02:04, 1.81it/s]\n",
"216it [02:05, 1.79it/s]\n",
"217it [02:05, 1.78it/s]\n",
"218it [02:06, 1.76it/s]\n",
"219it [02:07, 1.75it/s]\n",
"220it [02:07, 1.75it/s]\n",
"221it [02:08, 1.73it/s]\n",
"222it [02:08, 1.76it/s]\n",
"223it [02:09, 1.78it/s]\n",
"224it [02:09, 1.77it/s]\n",
"225it [02:10, 1.71it/s]\n",
"226it [02:11, 1.73it/s]\n",
"227it [02:11, 1.76it/s]\n",
"228it [02:12, 1.77it/s]\n",
"229it [02:12, 1.78it/s]\n",
"230it [02:13, 1.76it/s]\n",
"231it [02:13, 1.77it/s]\n",
"232it [02:14, 1.79it/s]\n",
"233it [02:15, 1.78it/s]\n",
"234it [02:15, 1.78it/s]\n",
"235it [02:16, 1.78it/s]\n",
"236it [02:16, 1.77it/s]\n",
"237it [02:17, 1.77it/s]\n",
"238it [02:17, 1.77it/s]\n",
"239it [02:18, 1.73it/s]\n",
"240it [02:19, 1.72it/s]\n",
"241it [02:19, 1.66it/s]\n",
"242it [02:20, 1.69it/s]\n",
"243it [02:20, 1.67it/s]\n",
"244it [02:21, 1.70it/s]\n",
"245it [02:22, 1.73it/s]\n",
"246it [02:22, 1.73it/s]\n",
"247it [02:23, 1.72it/s]\n",
"248it [02:23, 1.75it/s]\n",
"249it [02:24, 1.76it/s]\n",
"250it [02:24, 1.77it/s]\n",
"251it [02:25, 1.78it/s]\n",
"252it [02:25, 1.78it/s]\n",
"253it [02:26, 1.78it/s]\n",
"254it [02:27, 1.79it/s]\n",
"255it [02:27, 1.73it/s]\n",
"256it [02:28, 1.75it/s]\n",
"257it [02:28, 1.71it/s]\n",
"258it [02:29, 1.71it/s]\n",
"259it [02:30, 1.73it/s]\n",
"260it [02:30, 1.74it/s]\n",
"261it [02:31, 1.75it/s]\n",
"262it [02:31, 1.76it/s]\n",
"263it [02:32, 1.77it/s]\n",
"264it [02:32, 1.76it/s]\n",
"265it [02:33, 1.71it/s]\n",
"266it [02:34, 1.73it/s]\n",
"267it [02:34, 1.69it/s]\n",
"268it [02:35, 1.67it/s]\n",
"269it [02:35, 1.64it/s]\n",
"270it [02:36, 1.62it/s]\n",
"271it [02:37, 1.61it/s]\n",
"272it [02:37, 1.60it/s]\n",
"273it [02:38, 1.59it/s]\n",
"274it [02:39, 1.59it/s]\n",
"275it [02:39, 1.58it/s]\n",
"276it [02:40, 1.58it/s]\n",
"277it [02:40, 1.57it/s]\n",
"278it [02:41, 1.57it/s]\n",
"279it [02:42, 1.57it/s]\n",
"280it [02:42, 1.54it/s]\n",
"281it [02:43, 1.52it/s]\n",
"282it [02:44, 1.58it/s]\n",
"283it [02:44, 1.57it/s]\n",
"284it [02:45, 1.58it/s]\n",
"285it [02:46, 1.64it/s]\n",
"286it [02:46, 1.69it/s]\n",
"287it [02:47, 1.70it/s]\n",
"288it [02:47, 1.65it/s]\n",
"289it [02:48, 1.63it/s]\n",
"290it [02:49, 1.61it/s]\n",
"291it [02:49, 1.46it/s]\n",
"292it [02:50, 1.46it/s]\n",
"293it [02:51, 1.49it/s]\n",
"294it [02:51, 1.46it/s]\n",
"295it [02:52, 1.46it/s]\n",
"296it [02:53, 1.45it/s]\n",
"297it [02:54, 1.38it/s]\n",
"298it [02:54, 1.39it/s]\n",
"299it [02:55, 1.40it/s]\n",
"300it [02:56, 1.42it/s]\n",
"301it [02:57, 1.24it/s]\n",
"302it [02:57, 1.30it/s]\n",
"303it [02:58, 1.35it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"'Namespace' object has no attribute 'max_folder'\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA073B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA53090>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA073B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43EA0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43810>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA435E0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA438B0>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BA43540>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090B982220>\n",
"cannot identify image file <_io.BytesIO object at 0x000002090BAD3E50>\n"
]
}
],
"source": [
"!python image_download.py --file_path D:\\Michał\\studia\\wikisource-data\\yellow.tsv"
]
}
],
"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
}
Reference in New Issue Copy Permalink