├── DatasetScraper └── scraper.py ├── LICENSE ├── README.md ├── build └── lib │ └── datasetscraper │ ├── Scraper.py │ ├── __init__.py │ ├── downloader.py │ └── pageagent.py ├── datasetscraper.egg-info ├── PKG-INFO ├── SOURCES.txt ├── dependency_links.txt ├── requires.txt └── top_level.txt ├── datasetscraper ├── Scraper.py ├── __init__.py ├── __pycache__ │ ├── __init__.cpython-36.pyc │ ├── downloader.cpython-36.pyc │ ├── pageagent.cpython-36.pyc │ └── scraper.cpython-36.pyc ├── downloader.py └── pageagent.py ├── dist ├── datasetscraper-0.0.4-py3-none-any.whl └── datasetscraper-0.0.4.tar.gz └── setup.py /DatasetScraper/scraper.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | from .downloader import Downloader 12 | from .pageagent import PageAgent 13 | 14 | class Scraper(): 15 | def __init__(self, logLevel = logging.WARNING, headless=True): 16 | self.configure_logging(logLevel) 17 | self.patch_pyppeteer() 18 | self.headless = headless 19 | 20 | def patch_pyppeteer(self): 21 | import pyppeteer.connection 22 | original_method = pyppeteer.connection.websockets.client.connect 23 | def new_method(*args, **kwargs): 24 | kwargs['ping_interval'] = None 25 | kwargs['ping_timeout'] = None 26 | return original_method(*args, **kwargs) 27 | 28 | pyppeteer.connection.websockets.client.connect = new_method 29 | 30 | def configure_logging(self, logLevel): 31 | logger = logging.getLogger() 32 | pyppeteer_logger = logging.getLogger('pyppeteer') 33 | logger.setLevel(logLevel) 34 | handler = logging.StreamHandler() 35 | handler.setFormatter( 36 | logging.Formatter('[%(asctime)s %(levelname)s \ 37 | %(module)s]: %(message)s')) 38 | logger.addHandler(handler) 39 | return logger 40 | 41 | def fetch_urls(self, query, engine='google', maxlist=[200], format='jpg'): 42 | logger = logging.getLogger() 43 | print("Fetching URLs..") 44 | if type(engine) is not list: engine = [ engine ] 45 | if type(format) is not list: format = [ format ] 46 | if type(maxlist) is not list: maxlist = [ maxlist for e in engine ] 47 | assert len(maxlist) == len(engine), f"Length of max ({len(maxlist)}) not same as engine ({len(engine)})" 48 | self.engine, self.format = engine, format 49 | urlDict, totalLen = {}, 0 50 | for e, max in zip(engine, maxlist): 51 | logger.info(f"Fetching URLs from {e}") 52 | urlDict[e] = asyncio.get_event_loop().run_until_complete( 53 | self.launch_engine(e, query, max)) 54 | totalLen += len(urlDict[e]) 55 | logger.info(f"Fetched {len(urlDict[e])} URLs from {e}") 56 | print(f"Fetched {len(urlDict[e])} URLs from {e}") 57 | logger = logging.getLogger() 58 | logger.info(f"Total Number of URLs fetched: {totalLen}") 59 | urlSaved = self.mixUrls(urlDict, maxlist) 60 | logger.info(f"Number of URLs saving: {len(urlSaved)}") 61 | print(f"{len(urlSaved)} URLs fetched") 62 | return urlSaved 63 | 64 | def mixUrls(self, urlDict, maxlist): 65 | urls = [] 66 | for key, max in zip(urlDict, maxlist): 67 | urlDict[key] = urlDict[key][:max] 68 | for url in urlDict[key]: 69 | if url not in urls: 70 | urls.append(url) 71 | return urls 72 | 73 | def download(self, urls, directory='images/', 74 | formats=['jpg', 'png', 'jpeg'], default='jpg', 75 | nworkers=10, timeout=30): 76 | downloader = Downloader(directory, formats, default, 77 | nworkers, timeout) 78 | downloader.download(urls) 79 | 80 | async def launch_engine(self, engine, query, max=200): 81 | agent = PageAgent(engine, query, max, self.headless) 82 | return await agent.get_list() 83 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU GENERAL PUBLIC LICENSE 2 | Version 2, June 1991 3 | 4 | Copyright (C) 1989, 1991 Free Software Foundation, Inc., 5 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 6 | Everyone is permitted to copy and distribute verbatim copies 7 | of this license document, but changing it is not allowed. 8 | 9 | Preamble 10 | 11 | The licenses for most software are designed to take away your 12 | freedom to share and change it. By contrast, the GNU General Public 13 | License is intended to guarantee your freedom to share and change free 14 | software--to make sure the software is free for all its users. This 15 | General Public License applies to most of the Free Software 16 | Foundation's software and to any other program whose authors commit to 17 | using it. (Some other Free Software Foundation software is covered by 18 | the GNU Lesser General Public License instead.) You can apply it to 19 | your programs, too. 20 | 21 | When we speak of free software, we are referring to freedom, not 22 | price. Our General Public Licenses are designed to make sure that you 23 | have the freedom to distribute copies of free software (and charge for 24 | this service if you wish), that you receive source code or can get it 25 | if you want it, that you can change the software or use pieces of it 26 | in new free programs; and that you know you can do these things. 27 | 28 | To protect your rights, we need to make restrictions that forbid 29 | anyone to deny you these rights or to ask you to surrender the rights. 30 | These restrictions translate to certain responsibilities for you if you 31 | distribute copies of the software, or if you modify it. 32 | 33 | For example, if you distribute copies of such a program, whether 34 | gratis or for a fee, you must give the recipients all the rights that 35 | you have. You must make sure that they, too, receive or can get the 36 | source code. And you must show them these terms so they know their 37 | rights. 38 | 39 | We protect your rights with two steps: (1) copyright the software, and 40 | (2) offer you this license which gives you legal permission to copy, 41 | distribute and/or modify the software. 42 | 43 | Also, for each author's protection and ours, we want to make certain 44 | that everyone understands that there is no warranty for this free 45 | software. If the software is modified by someone else and passed on, we 46 | want its recipients to know that what they have is not the original, so 47 | that any problems introduced by others will not reflect on the original 48 | authors' reputations. 49 | 50 | Finally, any free program is threatened constantly by software 51 | patents. We wish to avoid the danger that redistributors of a free 52 | program will individually obtain patent licenses, in effect making the 53 | program proprietary. To prevent this, we have made it clear that any 54 | patent must be licensed for everyone's free use or not licensed at all. 55 | 56 | The precise terms and conditions for copying, distribution and 57 | modification follow. 58 | 59 | GNU GENERAL PUBLIC LICENSE 60 | TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 61 | 62 | 0. This License applies to any program or other work which contains 63 | a notice placed by the copyright holder saying it may be distributed 64 | under the terms of this General Public License. The "Program", below, 65 | refers to any such program or work, and a "work based on the Program" 66 | means either the Program or any derivative work under copyright law: 67 | that is to say, a work containing the Program or a portion of it, 68 | either verbatim or with modifications and/or translated into another 69 | language. (Hereinafter, translation is included without limitation in 70 | the term "modification".) Each licensee is addressed as "you". 71 | 72 | Activities other than copying, distribution and modification are not 73 | covered by this License; they are outside its scope. The act of 74 | running the Program is not restricted, and the output from the Program 75 | is covered only if its contents constitute a work based on the 76 | Program (independent of having been made by running the Program). 77 | Whether that is true depends on what the Program does. 78 | 79 | 1. You may copy and distribute verbatim copies of the Program's 80 | source code as you receive it, in any medium, provided that you 81 | conspicuously and appropriately publish on each copy an appropriate 82 | copyright notice and disclaimer of warranty; keep intact all the 83 | notices that refer to this License and to the absence of any warranty; 84 | and give any other recipients of the Program a copy of this License 85 | along with the Program. 86 | 87 | You may charge a fee for the physical act of transferring a copy, and 88 | you may at your option offer warranty protection in exchange for a fee. 89 | 90 | 2. You may modify your copy or copies of the Program or any portion 91 | of it, thus forming a work based on the Program, and copy and 92 | distribute such modifications or work under the terms of Section 1 93 | above, provided that you also meet all of these conditions: 94 | 95 | a) You must cause the modified files to carry prominent notices 96 | stating that you changed the files and the date of any change. 97 | 98 | b) You must cause any work that you distribute or publish, that in 99 | whole or in part contains or is derived from the Program or any 100 | part thereof, to be licensed as a whole at no charge to all third 101 | parties under the terms of this License. 102 | 103 | c) If the modified program normally reads commands interactively 104 | when run, you must cause it, when started running for such 105 | interactive use in the most ordinary way, to print or display an 106 | announcement including an appropriate copyright notice and a 107 | notice that there is no warranty (or else, saying that you provide 108 | a warranty) and that users may redistribute the program under 109 | these conditions, and telling the user how to view a copy of this 110 | License. (Exception: if the Program itself is interactive but 111 | does not normally print such an announcement, your work based on 112 | the Program is not required to print an announcement.) 113 | 114 | These requirements apply to the modified work as a whole. If 115 | identifiable sections of that work are not derived from the Program, 116 | and can be reasonably considered independent and separate works in 117 | themselves, then this License, and its terms, do not apply to those 118 | sections when you distribute them as separate works. But when you 119 | distribute the same sections as part of a whole which is a work based 120 | on the Program, the distribution of the whole must be on the terms of 121 | this License, whose permissions for other licensees extend to the 122 | entire whole, and thus to each and every part regardless of who wrote it. 123 | 124 | Thus, it is not the intent of this section to claim rights or contest 125 | your rights to work written entirely by you; rather, the intent is to 126 | exercise the right to control the distribution of derivative or 127 | collective works based on the Program. 128 | 129 | In addition, mere aggregation of another work not based on the Program 130 | with the Program (or with a work based on the Program) on a volume of 131 | a storage or distribution medium does not bring the other work under 132 | the scope of this License. 133 | 134 | 3. You may copy and distribute the Program (or a work based on it, 135 | under Section 2) in object code or executable form under the terms of 136 | Sections 1 and 2 above provided that you also do one of the following: 137 | 138 | a) Accompany it with the complete corresponding machine-readable 139 | source code, which must be distributed under the terms of Sections 140 | 1 and 2 above on a medium customarily used for software interchange; or, 141 | 142 | b) Accompany it with a written offer, valid for at least three 143 | years, to give any third party, for a charge no more than your 144 | cost of physically performing source distribution, a complete 145 | machine-readable copy of the corresponding source code, to be 146 | distributed under the terms of Sections 1 and 2 above on a medium 147 | customarily used for software interchange; or, 148 | 149 | c) Accompany it with the information you received as to the offer 150 | to distribute corresponding source code. (This alternative is 151 | allowed only for noncommercial distribution and only if you 152 | received the program in object code or executable form with such 153 | an offer, in accord with Subsection b above.) 154 | 155 | The source code for a work means the preferred form of the work for 156 | making modifications to it. For an executable work, complete source 157 | code means all the source code for all modules it contains, plus any 158 | associated interface definition files, plus the scripts used to 159 | control compilation and installation of the executable. However, as a 160 | special exception, the source code distributed need not include 161 | anything that is normally distributed (in either source or binary 162 | form) with the major components (compiler, kernel, and so on) of the 163 | operating system on which the executable runs, unless that component 164 | itself accompanies the executable. 165 | 166 | If distribution of executable or object code is made by offering 167 | access to copy from a designated place, then offering equivalent 168 | access to copy the source code from the same place counts as 169 | distribution of the source code, even though third parties are not 170 | compelled to copy the source along with the object code. 171 | 172 | 4. You may not copy, modify, sublicense, or distribute the Program 173 | except as expressly provided under this License. Any attempt 174 | otherwise to copy, modify, sublicense or distribute the Program is 175 | void, and will automatically terminate your rights under this License. 176 | However, parties who have received copies, or rights, from you under 177 | this License will not have their licenses terminated so long as such 178 | parties remain in full compliance. 179 | 180 | 5. You are not required to accept this License, since you have not 181 | signed it. However, nothing else grants you permission to modify or 182 | distribute the Program or its derivative works. These actions are 183 | prohibited by law if you do not accept this License. Therefore, by 184 | modifying or distributing the Program (or any work based on the 185 | Program), you indicate your acceptance of this License to do so, and 186 | all its terms and conditions for copying, distributing or modifying 187 | the Program or works based on it. 188 | 189 | 6. Each time you redistribute the Program (or any work based on the 190 | Program), the recipient automatically receives a license from the 191 | original licensor to copy, distribute or modify the Program subject to 192 | these terms and conditions. You may not impose any further 193 | restrictions on the recipients' exercise of the rights granted herein. 194 | You are not responsible for enforcing compliance by third parties to 195 | this License. 196 | 197 | 7. If, as a consequence of a court judgment or allegation of patent 198 | infringement or for any other reason (not limited to patent issues), 199 | conditions are imposed on you (whether by court order, agreement or 200 | otherwise) that contradict the conditions of this License, they do not 201 | excuse you from the conditions of this License. If you cannot 202 | distribute so as to satisfy simultaneously your obligations under this 203 | License and any other pertinent obligations, then as a consequence you 204 | may not distribute the Program at all. For example, if a patent 205 | license would not permit royalty-free redistribution of the Program by 206 | all those who receive copies directly or indirectly through you, then 207 | the only way you could satisfy both it and this License would be to 208 | refrain entirely from distribution of the Program. 209 | 210 | If any portion of this section is held invalid or unenforceable under 211 | any particular circumstance, the balance of the section is intended to 212 | apply and the section as a whole is intended to apply in other 213 | circumstances. 214 | 215 | It is not the purpose of this section to induce you to infringe any 216 | patents or other property right claims or to contest validity of any 217 | such claims; this section has the sole purpose of protecting the 218 | integrity of the free software distribution system, which is 219 | implemented by public license practices. Many people have made 220 | generous contributions to the wide range of software distributed 221 | through that system in reliance on consistent application of that 222 | system; it is up to the author/donor to decide if he or she is willing 223 | to distribute software through any other system and a licensee cannot 224 | impose that choice. 225 | 226 | This section is intended to make thoroughly clear what is believed to 227 | be a consequence of the rest of this License. 228 | 229 | 8. If the distribution and/or use of the Program is restricted in 230 | certain countries either by patents or by copyrighted interfaces, the 231 | original copyright holder who places the Program under this License 232 | may add an explicit geographical distribution limitation excluding 233 | those countries, so that distribution is permitted only in or among 234 | countries not thus excluded. In such case, this License incorporates 235 | the limitation as if written in the body of this License. 236 | 237 | 9. The Free Software Foundation may publish revised and/or new versions 238 | of the General Public License from time to time. Such new versions will 239 | be similar in spirit to the present version, but may differ in detail to 240 | address new problems or concerns. 241 | 242 | Each version is given a distinguishing version number. If the Program 243 | specifies a version number of this License which applies to it and "any 244 | later version", you have the option of following the terms and conditions 245 | either of that version or of any later version published by the Free 246 | Software Foundation. If the Program does not specify a version number of 247 | this License, you may choose any version ever published by the Free Software 248 | Foundation. 249 | 250 | 10. If you wish to incorporate parts of the Program into other free 251 | programs whose distribution conditions are different, write to the author 252 | to ask for permission. For software which is copyrighted by the Free 253 | Software Foundation, write to the Free Software Foundation; we sometimes 254 | make exceptions for this. Our decision will be guided by the two goals 255 | of preserving the free status of all derivatives of our free software and 256 | of promoting the sharing and reuse of software generally. 257 | 258 | NO WARRANTY 259 | 260 | 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY 261 | FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN 262 | OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES 263 | PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED 264 | OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 265 | MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS 266 | TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE 267 | PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, 268 | REPAIR OR CORRECTION. 269 | 270 | 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 271 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR 272 | REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, 273 | INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING 274 | OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED 275 | TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY 276 | YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER 277 | PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE 278 | POSSIBILITY OF SUCH DAMAGES. 279 | 280 | END OF TERMS AND CONDITIONS 281 | 282 | How to Apply These Terms to Your New Programs 283 | 284 | If you develop a new program, and you want it to be of the greatest 285 | possible use to the public, the best way to achieve this is to make it 286 | free software which everyone can redistribute and change under these terms. 287 | 288 | To do so, attach the following notices to the program. It is safest 289 | to attach them to the start of each source file to most effectively 290 | convey the exclusion of warranty; and each file should have at least 291 | the "copyright" line and a pointer to where the full notice is found. 292 | 293 | 294 | Copyright (C) 295 | 296 | This program is free software; you can redistribute it and/or modify 297 | it under the terms of the GNU General Public License as published by 298 | the Free Software Foundation; either version 2 of the License, or 299 | (at your option) any later version. 300 | 301 | This program is distributed in the hope that it will be useful, 302 | but WITHOUT ANY WARRANTY; without even the implied warranty of 303 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 304 | GNU General Public License for more details. 305 | 306 | You should have received a copy of the GNU General Public License along 307 | with this program; if not, write to the Free Software Foundation, Inc., 308 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. 309 | 310 | Also add information on how to contact you by electronic and paper mail. 311 | 312 | If the program is interactive, make it output a short notice like this 313 | when it starts in an interactive mode: 314 | 315 | Gnomovision version 69, Copyright (C) year name of author 316 | Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. 317 | This is free software, and you are welcome to redistribute it 318 | under certain conditions; type `show c' for details. 319 | 320 | The hypothetical commands `show w' and `show c' should show the appropriate 321 | parts of the General Public License. Of course, the commands you use may 322 | be called something other than `show w' and `show c'; they could even be 323 | mouse-clicks or menu items--whatever suits your program. 324 | 325 | You should also get your employer (if you work as a programmer) or your 326 | school, if any, to sign a "copyright disclaimer" for the program, if 327 | necessary. Here is a sample; alter the names: 328 | 329 | Yoyodyne, Inc., hereby disclaims all copyright interest in the program 330 | `Gnomovision' (which makes passes at compilers) written by James Hacker. 331 | 332 | , 1 April 1989 333 | Ty Coon, President of Vice 334 | 335 | This General Public License does not permit incorporating your program into 336 | proprietary programs. If your program is a subroutine library, you may 337 | consider it more useful to permit linking proprietary applications with the 338 | library. If this is what you want to do, use the GNU Lesser General 339 | Public License instead of this License. 340 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # DatasetScraper 2 | Tool to create image datasets for machine learning problems by scraping search engines like Google, Bing and Baidu. 3 | 4 | # Features: 5 | - **Search engine support**: Google, Bing, Baidu. (in-production): Yahoo, Yandex, Duckduckgo 6 | - **Image format support**: jpg, png, svg, gif, jpeg 7 | - Fast multiprocessing enabled scraper 8 | - Very fast multithreaded downloader 9 | - Data verification after download for assertion of image files 10 | 11 | # Installation 12 | `pip install datasetscraper` 13 | Alternatively, you can clone this repository: 14 | 15 | 16 | # About 17 | - `Scraper.Scraper()` 18 | - creates a scraper object which can be used to scrape URLs from search engines and later download them. 19 | - `Scraper.Scraper(logLevel = logging.WARNING, headless=True)` are the defaults. 20 | - headless: If False then you can see the actual pypeeter browser 21 | - logLevel: Simple python logger level 22 | 23 | - `fetch_urls()` 24 | - To fetch image URLs 25 | - `fetch_urls(query, engine='google', maxlist=[200])` 26 | - query: Search query 27 | - engine: search engine or list of search engines 28 | - maxlist: Maximum number of images to download. If multiple search engines exist then this needs to be a list specifying maximum number of images to fetch from each engine. 29 | 30 | - download 31 | - To download the fetched URLs into a directory 32 | - download(urls, directory='images/', 33 | formats=['jpg', 'png', 'jpeg'], default='jpg', 34 | nworkers=10, timeout=30) 35 | - urls: List of urls 36 | - directory: Place on disk where downloaded images are stored 37 | - formats: List of file formats to dwonload 38 | - nworkers: Number of threads for multithreaded download 39 | - timeout: maximum time before which the download is cancelled 40 | 41 | # Usage: 42 | - Import 43 | `from datasetscraper import Scraper` 44 | 45 | - Defaults 46 | ```python 47 | obj = Scraper.Scraper() 48 | urls = obj.fetch_urls('kiniro mosaic') 49 | obj.download(urls, directory='kiniro_mosaic/') 50 | ``` 51 | 52 | - Specify a search engine 53 | ```python 54 | obj = Scraper.Scraper() 55 | urls = obj.fetch_urls('kiniro mosaic', engine=['google']) 56 | obj.download(urls, directory='kiniro_mosaic/') 57 | ``` 58 | 59 | - Specify a list of search engines 60 | ```python 61 | obj = Scraper.Scraper() 62 | urls = obj.fetch_urls('kiniro mosaic', engine=['google', 'bing']) 63 | obj.download(urls, directory='kiniro_mosaic/') 64 | ``` 65 | 66 | - Specify max images (default was 200) 67 | ```python 68 | obj = Scraper.Scraper() 69 | urls = obj.fetch_urls('kiniro mosaic', engine=['google', 'bing'], maxlist=[500, 300]) 70 | obj.download(urls, directory='kiniro_mosaic/') 71 | ``` 72 | 73 | # FAQs 74 | - Why aren't yandex, yahoo, duckduckgo and other search engines supported? 75 | They are hard to scrape, I am working on them and will update as soon as I can. 76 | 77 | - I set maxlist=[500] why are only (x<500) images downloaded? 78 | There can be several reasons for this: 79 | - Search ran out: This happens very often, google/bing might not have enough images for your query 80 | - Slow internet: Increase the timeout (default is 60 seconds) as follows: ```obj.download(urls, directory='kiniro_mosaic/', timeout=100)``` 81 | 82 | - How to debug? 83 | You can change the logging level while making the scraper object : `obj = Scraper(logger.INFO)` 84 | 85 | # TODO: 86 | - More search engines 87 | - Better debug 88 | - Write documentation 89 | - Text data? Audio data? 90 | 91 | # Black hole chan kawaii 92 | ![](https://i.imgur.com/6OIxoup.jpg) 93 | -------------------------------------------------------------------------------- /build/lib/datasetscraper/Scraper.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | from .downloader import Downloader 12 | from .pageagent import PageAgent 13 | 14 | class Scraper(): 15 | def __init__(self, logLevel = logging.WARNING, headless=True): 16 | self.configure_logging(logLevel) 17 | self.patch_pyppeteer() 18 | self.headless = headless 19 | 20 | def patch_pyppeteer(self): 21 | import pyppeteer.connection 22 | original_method = pyppeteer.connection.websockets.client.connect 23 | def new_method(*args, **kwargs): 24 | kwargs['ping_interval'] = None 25 | kwargs['ping_timeout'] = None 26 | return original_method(*args, **kwargs) 27 | 28 | pyppeteer.connection.websockets.client.connect = new_method 29 | 30 | def configure_logging(self, logLevel): 31 | logger = logging.getLogger() 32 | pyppeteer_logger = logging.getLogger('pyppeteer') 33 | logger.setLevel(logLevel) 34 | handler = logging.StreamHandler() 35 | handler.setFormatter( 36 | logging.Formatter('[%(asctime)s %(levelname)s \ 37 | %(module)s]: %(message)s')) 38 | logger.addHandler(handler) 39 | return logger 40 | 41 | def fetch_urls(self, query, engine='google', maxlist=[200], format='jpg'): 42 | logger = logging.getLogger() 43 | print("Fetching URLs..") 44 | if type(engine) is not list: engine = [ engine ] 45 | if type(format) is not list: format = [ format ] 46 | if type(maxlist) is not list: maxlist = [ maxlist for e in engine ] 47 | assert len(maxlist) == len(engine), f"Length of max ({len(maxlist)}) not same as engine ({len(engine)})" 48 | self.engine, self.format = engine, format 49 | urlDict, totalLen = {}, 0 50 | for e, max in zip(engine, maxlist): 51 | logger.info(f"Fetching URLs from {e}") 52 | urlDict[e] = asyncio.get_event_loop().run_until_complete( 53 | self.launch_engine(e, query, max)) 54 | totalLen += len(urlDict[e]) 55 | logger.info(f"Fetched {len(urlDict[e])} URLs from {e}") 56 | print(f"Fetched {len(urlDict[e])} URLs from {e}") 57 | logger = logging.getLogger() 58 | logger.info(f"Total Number of URLs fetched: {totalLen}") 59 | urlSaved = self.mixUrls(urlDict, maxlist) 60 | logger.info(f"Number of URLs saving: {len(urlSaved)}") 61 | print(f"{len(urlSaved)} URLs fetched") 62 | return urlSaved 63 | 64 | def mixUrls(self, urlDict, maxlist): 65 | urls = [] 66 | for key, max in zip(urlDict, maxlist): 67 | urlDict[key] = urlDict[key][:max] 68 | for url in urlDict[key]: 69 | if url not in urls: 70 | urls.append(url) 71 | return urls 72 | 73 | def download(self, urls, directory='images/', 74 | formats=['jpg', 'png', 'jpeg'], default='jpg', 75 | nworkers=10, timeout=30): 76 | downloader = Downloader(directory, formats, default, 77 | nworkers, timeout) 78 | downloader.download(urls) 79 | 80 | async def launch_engine(self, engine, query, max=200): 81 | agent = PageAgent(engine, query, max, self.headless) 82 | return await agent.get_list() 83 | -------------------------------------------------------------------------------- /build/lib/datasetscraper/__init__.py: -------------------------------------------------------------------------------- 1 | name = "datasetscraper" 2 | -------------------------------------------------------------------------------- /build/lib/datasetscraper/downloader.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | import requests as req 12 | import traceback 13 | from multiprocessing.pool import ThreadPool 14 | from fastprogress import progress_bar 15 | import imghdr 16 | 17 | class Downloader(): 18 | def __init__(self, directory='images/', formats=['jpg', 'png', 'jpeg'], 19 | default='jpg', nworkers=10, timeout=60): 20 | self.formats, self.default = formats, default 21 | self.directory = Path(directory) 22 | self.MBFACTOR = float(1 << 20) 23 | self.nworkers, self.timeout = nworkers, timeout 24 | if not os.path.exists(directory): 25 | os.makedirs(directory) 26 | 27 | def download(self, urls, timeout=200): 28 | self.timeout = timeout 29 | logger = logging.getLogger() 30 | urls = self.clean_urls(urls) 31 | with concurrent.futures.ThreadPoolExecutor( 32 | max_workers = self.nworkers) as ex: 33 | futures = [ex.submit(self.save_image, url['url'], 34 | self.directory/f"{str(i) + url['format']}") for i,\ 35 | url in enumerate(urls)] 36 | for f in progress_bar(concurrent.futures.as_completed(futures), 37 | total=len(urls)): pass 38 | self.verify() 39 | 40 | def save_file(self, file_name, response): 41 | with open(file_name, 'wb') as fh: 42 | for chunk in response.iter_content(1024 * 1024): 43 | fh.write(chunk) 44 | 45 | def save_image(self, url, file_name): 46 | logger = logging.getLogger() 47 | opener = request.build_opener() 48 | user_agent = 'Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; \ 49 | rv:1.9.0.7) Gecko/2009021910 Firefox/3.0.7' 50 | try: 51 | opener.addheaders = [('User-agent', user_agent)] 52 | request.install_opener(opener) 53 | logger.info(f"[Downloading] {url} [AS FILE] {file_name}") 54 | response = req.get(url, timeout=self.timeout) 55 | self.save_file(file_name, response) 56 | try: 57 | size = int(response.headers['Content-Length'])/self.MBFACTOR 58 | except: 59 | size = 0 60 | except Exception as e: 61 | # logger.info(f"[FAILED] {file_name} - {url} because: \n{e}") 62 | # logger.info(f"[EXCEPTION]:{traceback.print_tb(e.__traceback__)}") 63 | if os.path.isfile(file_name): #Delete if urlretrieve fails 64 | os.remove(file_name) 65 | return #exit function 66 | # logger.info(f"[Downloading] {file_name}") 67 | logger.info(f"[Done] {file_name} with size: {round(size, 3)} MB") 68 | 69 | def clean_urls(self, urls): 70 | formats = [f'.{format}' for format in self.formats] 71 | default = f".{self.default}" 72 | furls = [] 73 | for url in urls: 74 | found = False 75 | for format in formats: 76 | if format in url: 77 | found = True 78 | break 79 | if not found: 80 | continue 81 | furls.append({ 82 | 'url':url, 83 | 'format':format, 84 | }) 85 | return furls 86 | 87 | def verify(self): 88 | print("Verifying download...") 89 | logger = logging.getLogger("Verifying download...") 90 | deleted = 0 91 | for file in os.listdir(self.directory): 92 | if not imghdr.what(self.directory/file): 93 | os.remove(self.directory/file) 94 | deleted += 1 95 | print(f"Deleted {deleted} corrupt/non-image files") 96 | logger = logging.getLogger(f"Deleted {deleted} corrupt/non-image files") 97 | -------------------------------------------------------------------------------- /build/lib/datasetscraper/pageagent.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | 12 | SANDBOX_ERROR = """No usable sandbox for chrome.\nFor a quick fix try \ 13 | running: `sudo sysctl -w kernel.unprivileged_userns_clone=1` on your linux \ 14 | shell. You need to set up chrome sandbox, for more info check out: \nhttps\ 15 | ://github.com/GoogleChrome/\ 16 | puppeteer/blob/master/docs/troubleshooting.md""" 17 | 18 | class PageAgent(): 19 | def __init__(self, engine, query, max=200, headless=True): 20 | self.engine = engine 21 | self.query = query 22 | self.max = max 23 | self.headless = headless 24 | # More search engine support coming soon, please update the INFO 25 | # dict below if you can. 26 | self.INFO = { 27 | "google":{ 28 | "nclick": 400, 29 | "nscroll": 5, 30 | "urlStart": "https://www.google.com/search?q=", 31 | "urlEnd": "&source=lnms&tbm=isch", 32 | "jClick": "document.getElementById(\"smb\").click();", 33 | "jsFunc": "Array.from(document.querySelectorAll('.rg_di .rg_meta')).\ 34 | map(el=>JSON.parse(el.textContent).ou);" 35 | }, 36 | "yahoo":{ 37 | "nclick": 500, 38 | "nscroll": 10, 39 | "urlStart": "https://images.search.yahoo.com/search/images;?&p=", 40 | "urlEnd": "&ei=UTF-8&iscqry=&fr=sfp", 41 | "jClick": "document.getElementsByClassName(\"ygbt more-res\")[0].click();", 42 | #jsFunc is pretty hard to write here 43 | "jsFunc": "Array.from(document.querySelectorAll('a.iusc')).map(x=>x.attributes.m).map(x=>JSON.parse(x.nodeValue)[\"murl\"]);", 44 | }, 45 | "bing":{ 46 | "nclick": 400, 47 | "nscroll": 10, 48 | "urlStart": "https://www.bing.com/images/search?q=", 49 | "urlEnd": "&source=lnms&tbm=isch", 50 | "jClick": "document.getElementsByClassName(\"btn_seemore cbtn mBtn\")[0].click();", 51 | "jsFunc": "Array.from(document.querySelectorAll('a.iusc')).map(x=>x.attributes.m).map(x=>JSON.parse(x.nodeValue)[\"murl\"]);", 52 | }, 53 | "duckduckgo":{ 54 | "nclick": 500, 55 | "nscroll": 10, 56 | "urlStart": "https://duckduckgo.com/?q=", 57 | "urlEnd": "&t=h_&ia=images&iax=images", 58 | "jClick": "document.getElementsByClassName(\"btn_seemore cbtn mBtn\")[0].click();", 59 | "jsFunc": "Array.from(document.querySelectorAll('img.tile--img__img')).map(x=>x.src).map(x=>JSON.parse(x.nodeValue)[\"murl\"]);", 60 | }, 61 | "baidu":{ 62 | "nclick": 500, 63 | "nscroll": 25, 64 | "urlStart": "https://image.baidu.com/search/index?tn=baiduimage&word=", 65 | "urlEnd": "", 66 | "jClick": "document.getElementsByClassName(\"btn_seemore cbtn mBtn\")[0].click();", 67 | "jsFunc": "Array.from(document.querySelectorAll('li.imgitem')).map(x=>x.dataset.objurl);", 68 | }, 69 | } 70 | 71 | async def scroll(self, page, nscroll): 72 | logger = logging.getLogger() 73 | jscroll = "window.scrollBy(0, document.body.scrollHeight);" 74 | logger.info(f"[SCROLLING PAGE..]") 75 | for x in range(nscroll): 76 | try: 77 | await page.evaluate(jscroll) 78 | await page.waitFor(2000) 79 | except: 80 | logger.info(f"Can't scroll") 81 | return 82 | 83 | async def click_more(self, page, nscroll, jClick, nclicks = 1): 84 | logger = logging.getLogger() 85 | logger.info(f"[NClicks: ] {nclicks}") 86 | if nclicks == 0: 87 | await self.scroll(page, nscroll) 88 | return 89 | 90 | for _ in range(nclicks): 91 | try: 92 | await self.scroll(page, nscroll) 93 | await page.evaluate(jClick, force_expr=True), 94 | await self.scroll(page, nscroll) 95 | except: 96 | return 97 | 98 | async def get_list(self): 99 | engine, query, max = self.engine, self.query, self.max 100 | info = self.INFO[engine] 101 | nclick = max // info['nclick'] + 1 102 | if nclick < 2: 103 | nclick = 0 104 | try: 105 | browser = await launch(headless=self.headless) 106 | except Exception as exception: 107 | print(SANDBOX_ERROR) 108 | sys.exit(0) 109 | if self.engine == 'baidu': 110 | nclick = 0 111 | page = await browser.newPage() 112 | await page.setUserAgent("Mozilla/5.0 (Windows NT 6.1; Win64; x64)\ 113 | AppleWebKit/537.36 (KHTML, like Gecko) \ 114 | Chrome/66.0.3359.181 Safari/537.36") 115 | url = info['urlStart'] + str(query) + info['urlEnd'] 116 | await page.goto(url, timeout=0) 117 | await self.click_more(page, info['nscroll'], info['jClick'], nclick) 118 | urlList = await page.evaluate(info["jsFunc"], force_expr=True) 119 | await browser.close() 120 | return urlList 121 | -------------------------------------------------------------------------------- /datasetscraper.egg-info/PKG-INFO: -------------------------------------------------------------------------------- 1 | Metadata-Version: 2.1 2 | Name: datasetscraper 3 | Version: 0.0.4 4 | Summary: Tool to create image datasets for machine learning problemsby scraping search engines like Google, Bing and Baidu. 5 | Home-page: https://github.com/TimeTraveller-San/datasetscraper 6 | Author: Time Traveller 7 | Author-email: time.traveller.san@gmail.com 8 | License: GPLv2 9 | Description: # DatasetScraper 10 | Tool to create image datasets for machine learning problems by scraping search engines like Google, Bing and Baidu. 11 | 12 | # Features: 13 | - **Search engine support**: Google, Bing, Baidu. (in-production): Yahoo, Yandex, Duckduckgo 14 | - **Image format support**: jpg, png, svg, gif, jpeg 15 | - Fast multiprocessing enabled scraper 16 | - Very fast multithreaded downloader 17 | - Data verification after download for assertion of image files 18 | 19 | # Installation 20 | - COMING SOON on pypi 21 | 22 | # Usage: 23 | - Import 24 | `from datasetscraper import Scraper` 25 | 26 | - Defaults 27 | ```python 28 | obj = Scraper() 29 | urls = obj.fetch_urls('kiniro mosaic') 30 | obj.download(urls, directory='kiniro_mosaic/') 31 | ``` 32 | 33 | - Specify a search engine 34 | ```python 35 | obj = Scraper() 36 | urls = obj.fetch_urls('kiniro mosaic', engine=['google']) 37 | obj.download(urls, directory='kiniro_mosaic/') 38 | ``` 39 | 40 | - Specify a list of search engines 41 | ```python 42 | obj = Scraper() 43 | urls = obj.fetch_urls('kiniro mosaic', engine=['google', 'bing']) 44 | obj.download(urls, directory='kiniro_mosaic/') 45 | ``` 46 | 47 | - Specify max images (default was 200) 48 | ```python 49 | obj = Scraper() 50 | urls = obj.fetch_urls('kiniro mosaic', engine=['google', 'bing'], maxlist=[500, 300]) 51 | obj.download(urls, directory='kiniro_mosaic/') 52 | ``` 53 | 54 | # FAQs 55 | - Why aren't yandex, yahoo, duckduckgo and other search engines supported? 56 | They are hard to scrape, I am working on them and will update as soon as I can. 57 | 58 | - I set maxlist=[500] why are only (x<500) images downloaded? 59 | There can be several reasons for this: 60 | - Search ran out: This happens very often, google/bing might not have enough images for your query 61 | - Slow internet: Increase the timeout (default is 60 seconds) as follows: ```obj.download(urls, directory='kiniro_mosaic/', timeout=100)``` 62 | 63 | - How to debug? 64 | You can change the logging level while making the scraper object : `obj = Scraper(logger.INFO)` 65 | 66 | # TODO: 67 | - More search engines 68 | - Better debug 69 | - Write documentation 70 | - Text data? Audio data? 71 | 72 | Keywords: dataset_scraper,machine learning,dataset,images,scrape,yandex,google,bing,baidu 73 | Platform: UNKNOWN 74 | Classifier: Programming Language :: Python :: 3 75 | Classifier: Operating System :: POSIX :: Linux 76 | Requires-Python: >=3 77 | Description-Content-Type: text/markdown 78 | -------------------------------------------------------------------------------- /datasetscraper.egg-info/SOURCES.txt: -------------------------------------------------------------------------------- 1 | README.md 2 | setup.py 3 | datasetscraper/Scraper.py 4 | datasetscraper/__init__.py 5 | datasetscraper/downloader.py 6 | datasetscraper/pageagent.py 7 | datasetscraper.egg-info/PKG-INFO 8 | datasetscraper.egg-info/SOURCES.txt 9 | datasetscraper.egg-info/dependency_links.txt 10 | datasetscraper.egg-info/requires.txt 11 | datasetscraper.egg-info/top_level.txt -------------------------------------------------------------------------------- /datasetscraper.egg-info/dependency_links.txt: -------------------------------------------------------------------------------- 1 | 2 | -------------------------------------------------------------------------------- /datasetscraper.egg-info/requires.txt: -------------------------------------------------------------------------------- 1 | pyppeteer>=0.0.25 2 | fastprogress>=0.1.21 3 | requests>=2.19.1 4 | -------------------------------------------------------------------------------- /datasetscraper.egg-info/top_level.txt: -------------------------------------------------------------------------------- 1 | datasetscraper 2 | -------------------------------------------------------------------------------- /datasetscraper/Scraper.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | from .downloader import Downloader 12 | from .pageagent import PageAgent 13 | 14 | class Scraper(): 15 | def __init__(self, logLevel = logging.WARNING, headless=True): 16 | self.configure_logging(logLevel) 17 | self.patch_pyppeteer() 18 | self.headless = headless 19 | 20 | def patch_pyppeteer(self): 21 | import pyppeteer.connection 22 | original_method = pyppeteer.connection.websockets.client.connect 23 | def new_method(*args, **kwargs): 24 | kwargs['ping_interval'] = None 25 | kwargs['ping_timeout'] = None 26 | return original_method(*args, **kwargs) 27 | 28 | pyppeteer.connection.websockets.client.connect = new_method 29 | 30 | def configure_logging(self, logLevel): 31 | logger = logging.getLogger() 32 | pyppeteer_logger = logging.getLogger('pyppeteer') 33 | logger.setLevel(logLevel) 34 | handler = logging.StreamHandler() 35 | handler.setFormatter( 36 | logging.Formatter('[%(asctime)s %(levelname)s \ 37 | %(module)s]: %(message)s')) 38 | logger.addHandler(handler) 39 | return logger 40 | 41 | def fetch_urls(self, query, engine='google', maxlist=[200], format='jpg'): 42 | logger = logging.getLogger() 43 | print("Fetching URLs..") 44 | if type(engine) is not list: engine = [ engine ] 45 | if type(format) is not list: format = [ format ] 46 | if type(maxlist) is not list: maxlist = [ maxlist for e in engine ] 47 | assert len(maxlist) == len(engine), f"Length of max ({len(maxlist)}) not same as engine ({len(engine)})" 48 | self.engine, self.format = engine, format 49 | urlDict, totalLen = {}, 0 50 | for e, max in zip(engine, maxlist): 51 | logger.info(f"Fetching URLs from {e}") 52 | urlDict[e] = asyncio.get_event_loop().run_until_complete( 53 | self.launch_engine(e, query, max)) 54 | totalLen += len(urlDict[e]) 55 | logger.info(f"Fetched {len(urlDict[e])} URLs from {e}") 56 | print(f"Fetched {len(urlDict[e])} URLs from {e}") 57 | logger = logging.getLogger() 58 | logger.info(f"Total Number of URLs fetched: {totalLen}") 59 | urlSaved = self.mixUrls(urlDict, maxlist) 60 | logger.info(f"Number of URLs saving: {len(urlSaved)}") 61 | print(f"{len(urlSaved)} URLs fetched") 62 | return urlSaved 63 | 64 | def mixUrls(self, urlDict, maxlist): 65 | urls = [] 66 | for key, max in zip(urlDict, maxlist): 67 | urlDict[key] = urlDict[key][:max] 68 | for url in urlDict[key]: 69 | if url not in urls: 70 | urls.append(url) 71 | return urls 72 | 73 | def download(self, urls, directory='images/', 74 | formats=['jpg', 'png', 'jpeg'], default='jpg', 75 | nworkers=10, timeout=30): 76 | downloader = Downloader(directory, formats, default, 77 | nworkers, timeout) 78 | downloader.download(urls) 79 | 80 | async def launch_engine(self, engine, query, max=200): 81 | agent = PageAgent(engine, query, max, self.headless) 82 | return await agent.get_list() 83 | -------------------------------------------------------------------------------- /datasetscraper/__init__.py: -------------------------------------------------------------------------------- 1 | name = "datasetscraper" 2 | -------------------------------------------------------------------------------- /datasetscraper/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vaibkumr/DatasetScraper/099420240309fb29b40623a2400b3dd769bf49e0/datasetscraper/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /datasetscraper/__pycache__/downloader.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vaibkumr/DatasetScraper/099420240309fb29b40623a2400b3dd769bf49e0/datasetscraper/__pycache__/downloader.cpython-36.pyc -------------------------------------------------------------------------------- /datasetscraper/__pycache__/pageagent.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vaibkumr/DatasetScraper/099420240309fb29b40623a2400b3dd769bf49e0/datasetscraper/__pycache__/pageagent.cpython-36.pyc -------------------------------------------------------------------------------- /datasetscraper/__pycache__/scraper.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vaibkumr/DatasetScraper/099420240309fb29b40623a2400b3dd769bf49e0/datasetscraper/__pycache__/scraper.cpython-36.pyc -------------------------------------------------------------------------------- /datasetscraper/downloader.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | import requests as req 12 | import traceback 13 | from multiprocessing.pool import ThreadPool 14 | from fastprogress import progress_bar 15 | import imghdr 16 | 17 | class Downloader(): 18 | def __init__(self, directory='images/', formats=['jpg', 'png', 'jpeg'], 19 | default='jpg', nworkers=10, timeout=60): 20 | self.formats, self.default = formats, default 21 | self.directory = Path(directory) 22 | self.MBFACTOR = float(1 << 20) 23 | self.nworkers, self.timeout = nworkers, timeout 24 | if not os.path.exists(directory): 25 | os.makedirs(directory) 26 | 27 | def download(self, urls, timeout=200): 28 | self.timeout = timeout 29 | logger = logging.getLogger() 30 | urls = self.clean_urls(urls) 31 | with concurrent.futures.ThreadPoolExecutor( 32 | max_workers = self.nworkers) as ex: 33 | futures = [ex.submit(self.save_image, url['url'], 34 | self.directory/f"{str(i) + url['format']}") for i,\ 35 | url in enumerate(urls)] 36 | for f in progress_bar(concurrent.futures.as_completed(futures), 37 | total=len(urls)): pass 38 | self.verify() 39 | 40 | def save_file(self, file_name, response): 41 | with open(file_name, 'wb') as fh: 42 | for chunk in response.iter_content(1024 * 1024): 43 | fh.write(chunk) 44 | 45 | def save_image(self, url, file_name): 46 | logger = logging.getLogger() 47 | opener = request.build_opener() 48 | user_agent = 'Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; \ 49 | rv:1.9.0.7) Gecko/2009021910 Firefox/3.0.7' 50 | try: 51 | opener.addheaders = [('User-agent', user_agent)] 52 | request.install_opener(opener) 53 | logger.info(f"[Downloading] {url} [AS FILE] {file_name}") 54 | response = req.get(url, timeout=self.timeout) 55 | self.save_file(file_name, response) 56 | try: 57 | size = int(response.headers['Content-Length'])/self.MBFACTOR 58 | except: 59 | size = 0 60 | except Exception as e: 61 | # logger.info(f"[FAILED] {file_name} - {url} because: \n{e}") 62 | # logger.info(f"[EXCEPTION]:{traceback.print_tb(e.__traceback__)}") 63 | if os.path.isfile(file_name): #Delete if urlretrieve fails 64 | os.remove(file_name) 65 | return #exit function 66 | # logger.info(f"[Downloading] {file_name}") 67 | logger.info(f"[Done] {file_name} with size: {round(size, 3)} MB") 68 | 69 | def clean_urls(self, urls): 70 | formats = [f'.{format}' for format in self.formats] 71 | default = f".{self.default}" 72 | furls = [] 73 | for url in urls: 74 | found = False 75 | for format in formats: 76 | if format in url: 77 | found = True 78 | break 79 | if not found: 80 | continue 81 | furls.append({ 82 | 'url':url, 83 | 'format':format, 84 | }) 85 | return furls 86 | 87 | def verify(self): 88 | print("Verifying download...") 89 | logger = logging.getLogger("Verifying download...") 90 | deleted = 0 91 | for file in os.listdir(self.directory): 92 | if not imghdr.what(self.directory/file): 93 | os.remove(self.directory/file) 94 | deleted += 1 95 | print(f"Deleted {deleted} corrupt/non-image files") 96 | logger = logging.getLogger(f"Deleted {deleted} corrupt/non-image files") 97 | -------------------------------------------------------------------------------- /datasetscraper/pageagent.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | from pyppeteer import launch 3 | import time 4 | from pathlib import Path 5 | import urllib 6 | import urllib.request as request 7 | import os 8 | import logging 9 | import concurrent.futures 10 | import sys 11 | 12 | SANDBOX_ERROR = """No usable sandbox for chrome.\nFor a quick fix try \ 13 | running: `sudo sysctl -w kernel.unprivileged_userns_clone=1` on your linux \ 14 | shell. You need to set up chrome sandbox, for more info check out: \nhttps\ 15 | ://github.com/GoogleChrome/\ 16 | puppeteer/blob/master/docs/troubleshooting.md""" 17 | 18 | class PageAgent(): 19 | def __init__(self, engine, query, max=200, headless=True): 20 | self.engine = engine 21 | self.query = query 22 | self.max = max 23 | self.headless = headless 24 | # More search engine support coming soon, please update the INFO 25 | # dict below if you can. 26 | self.INFO = { 27 | "google":{ 28 | "nclick": 400, 29 | "nscroll": 5, 30 | "urlStart": "https://www.google.com/search?q=", 31 | "urlEnd": "&source=lnms&tbm=isch", 32 | "jClick": "document.getElementById(\"smb\").click();", 33 | "jsFunc": "Array.from(document.querySelectorAll('.rg_di .rg_meta')).\ 34 | map(el=>JSON.parse(el.textContent).ou);" 35 | }, 36 | "yahoo":{ 37 | "nclick": 500, 38 | "nscroll": 10, 39 | "urlStart": "https://images.search.yahoo.com/search/images;?&p=", 40 | "urlEnd": "&ei=UTF-8&iscqry=&fr=sfp", 41 | "jClick": "document.getElementsByClassName(\"ygbt more-res\")[0].click();", 42 | #jsFunc is pretty hard to write here 43 | "jsFunc": "Array.from(document.querySelectorAll('a.iusc')).map(x=>x.attributes.m).map(x=>JSON.parse(x.nodeValue)[\"murl\"]);", 44 | }, 45 | "bing":{ 46 | "nclick": 400, 47 | "nscroll": 10, 48 | "urlStart": "https://www.bing.com/images/search?q=", 49 | "urlEnd": "&source=lnms&tbm=isch", 50 | "jClick": "document.getElementsByClassName(\"btn_seemore cbtn mBtn\")[0].click();", 51 | "jsFunc": "Array.from(document.querySelectorAll('a.iusc')).map(x=>x.attributes.m).map(x=>JSON.parse(x.nodeValue)[\"murl\"]);", 52 | }, 53 | "duckduckgo":{ 54 | "nclick": 500, 55 | "nscroll": 10, 56 | "urlStart": "https://duckduckgo.com/?q=", 57 | "urlEnd": "&t=h_&ia=images&iax=images", 58 | "jClick": "document.getElementsByClassName(\"btn_seemore cbtn mBtn\")[0].click();", 59 | "jsFunc": "Array.from(document.querySelectorAll('img.tile--img__img')).map(x=>x.src).map(x=>JSON.parse(x.nodeValue)[\"murl\"]);", 60 | }, 61 | "baidu":{ 62 | "nclick": 500, 63 | "nscroll": 25, 64 | "urlStart": "https://image.baidu.com/search/index?tn=baiduimage&word=", 65 | "urlEnd": "", 66 | "jClick": "document.getElementsByClassName(\"btn_seemore cbtn mBtn\")[0].click();", 67 | "jsFunc": "Array.from(document.querySelectorAll('li.imgitem')).map(x=>x.dataset.objurl);", 68 | }, 69 | } 70 | 71 | async def scroll(self, page, nscroll): 72 | logger = logging.getLogger() 73 | jscroll = "window.scrollBy(0, document.body.scrollHeight);" 74 | logger.info(f"[SCROLLING PAGE..]") 75 | for x in range(nscroll): 76 | try: 77 | await page.evaluate(jscroll) 78 | await page.waitFor(2000) 79 | except: 80 | logger.info(f"Can't scroll") 81 | return 82 | 83 | async def click_more(self, page, nscroll, jClick, nclicks = 1): 84 | logger = logging.getLogger() 85 | logger.info(f"[NClicks: ] {nclicks}") 86 | if nclicks == 0: 87 | await self.scroll(page, nscroll) 88 | return 89 | 90 | for _ in range(nclicks): 91 | try: 92 | await self.scroll(page, nscroll) 93 | await page.evaluate(jClick, force_expr=True), 94 | await self.scroll(page, nscroll) 95 | except: 96 | return 97 | 98 | async def get_list(self): 99 | engine, query, max = self.engine, self.query, self.max 100 | info = self.INFO[engine] 101 | nclick = max // info['nclick'] + 1 102 | if nclick < 2: 103 | nclick = 0 104 | try: 105 | browser = await launch(headless=self.headless) 106 | except Exception as exception: 107 | print(SANDBOX_ERROR) 108 | sys.exit(0) 109 | if self.engine == 'baidu': 110 | nclick = 0 111 | page = await browser.newPage() 112 | await page.setUserAgent("Mozilla/5.0 (Windows NT 6.1; Win64; x64)\ 113 | AppleWebKit/537.36 (KHTML, like Gecko) \ 114 | Chrome/66.0.3359.181 Safari/537.36") 115 | url = info['urlStart'] + str(query) + info['urlEnd'] 116 | await page.goto(url, timeout=0) 117 | await self.click_more(page, info['nscroll'], info['jClick'], nclick) 118 | urlList = await page.evaluate(info["jsFunc"], force_expr=True) 119 | await browser.close() 120 | return urlList 121 | -------------------------------------------------------------------------------- /dist/datasetscraper-0.0.4-py3-none-any.whl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vaibkumr/DatasetScraper/099420240309fb29b40623a2400b3dd769bf49e0/dist/datasetscraper-0.0.4-py3-none-any.whl -------------------------------------------------------------------------------- /dist/datasetscraper-0.0.4.tar.gz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/vaibkumr/DatasetScraper/099420240309fb29b40623a2400b3dd769bf49e0/dist/datasetscraper-0.0.4.tar.gz -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- 1 | import setuptools 2 | 3 | # reqs = (line.strip() for line in open("requirements.txt")) 4 | LONG_DESC = open('README.md').read() 5 | setuptools.setup( 6 | name="datasetscraper", 7 | version="0.0.4", 8 | author="Time Traveller", 9 | author_email="time.traveller.san@gmail.com", 10 | keywords = ['dataset_scraper', 'machine learning', 'dataset', 'images', 'scrape', 11 | 'yandex', 'google', 'bing', 'baidu'], 12 | description="Tool to create image datasets for machine learning problems\ 13 | by scraping search engines like Google, Bing and Baidu. ", 14 | long_description=LONG_DESC, 15 | long_description_content_type="text/markdown", 16 | url="https://github.com/TimeTraveller-San/datasetscraper", 17 | license="GPLv2", 18 | packages=setuptools.find_packages(), 19 | install_requires=[ 20 | 'pyppeteer>=0.0.25', 21 | 'fastprogress>=0.1.21', 22 | 'requests>=2.19.1', 23 | ], 24 | python_requires='>=3', 25 | classifiers=[ 26 | "Programming Language :: Python :: 3", 27 | "Operating System :: POSIX :: Linux", 28 | ], 29 | ) 30 | --------------------------------------------------------------------------------