> ## Content Index > Fetch the complete content index at: https://serpapi.com/blog/llms.txt > Use this file to discover other available public pages before exploring further. # Scrape ResearchGate all institution members in Python - URL: https://serpapi.com/blog/scrape-researchgate-all-institution-members-in-python/ - Published: 2022-05-06T10:34:59.000Z - Updated: 2023-06-08T16:10:31.000Z - Description: A line-by-line tutorial on scraping ResearchGate institution members with Python and playwright/parsel libraries. - Author: Dmitriy Zub - Tags: ResearchGate, Python, Web Scraping ## ****Intro** Currently, we don't have an API that supports extracting data from ResearchGate institution members page. This blog post is to show you way how you can do it yourself with the provided DIY solution below while we're working on releasing our proper API. The solution can be used for personal use as it doesn't include the [Legal US Shield](https://serpapi.com/#features) that we offer for our paid [production and above plans](https://serpapi.com/pricing) and has its limitations such as the need to bypass blocks, for example, CAPTCHA. You can check our public roadmap to track the progress for this API: πŸ—ΊοΈ [\[New API\] ResearchGate - Institution Members](https://github.com/serpapi/public-roadmap/issues/315) ## What will be scraped ![image](https://user-images.githubusercontent.com/78694043/167081287-3cbb8a5b-f500-4e94-a0e6-b2ab41ef2bf1.png) πŸ“ŒNote: For now, we don't have an API that supports extracting ResearchGate data. This blog post is to show you way how you can do it yourself. [Let us know](https://serpapi.com/#contact) if you want to have this API made by us. ## Prerequisites **Basic knowledge scraping with CSS selectors** CSS selectors declare which part of the markup a style applies to, thus allowing you to extract data from matching tags and attributes. If you haven't scraped with CSS selectors, there's a dedicated blog post of mine about [how to use CSS selectors when web-scraping](https://serpapi.com/blog/web-scraping-with-css-selectors-using-python/) that covers what it is, pros and cons, and why they matter from a web-scraping perspective and show the most common approaches of using CSS selectors when web scraping. **Separate virtual environment** In short, it's a thing that creates an independent set of installed libraries including different Python versions that can coexist with each other in the same system, thus preventing libraries or Python version conflicts. If you haven't worked with a virtual environment before, have a look at the dedicated [Python virtual environments tutorial using Virtualenv and Poetry](https://serpapi.com/blog/python-virtual-environments-using-virtualenv-and-poetry/) blog post of mine to get familiar. πŸ“ŒNote: This is not a strict requirement for this blog post. **Install libraries**: ```lang-none pip install parsel playwright ``` **Reduce the chance of being blocked** There's a chance that a request might be blocked. Have a look at [how to reduce the chance of being blocked while web-scraping](https://serpapi.com/blog/how-to-reduce-chance-of-being-blocked-while-web/), there are eleven methods to bypass blocks from most websites. ## Full Code ```python from parsel import Selector from playwright.sync_api import sync_playwright import re, json, time def scrape_institution_members(institution: str): with sync_playwright() as p: institution_memebers = [] page_num = 1 members_is_present = True while members_is_present: browser = p.chromium.launch(headless=True, slow_mo=50) page = browser.new_page() page.goto(f"https://www.researchgate.net/institution/{institution}/members/{page_num}") selector = Selector(text=page.content()) print(f"page number: {page_num}") for member in selector.css(".nova-legacy-v-person-list-item"): name = member.css(".nova-legacy-v-person-list-item__align-content a::text").get() link = f'https://www.researchgate.net{member.css(".nova-legacy-v-person-list-item__align-content a::attr(href)").get()}' profile_photo = member.css(".nova-legacy-l-flex__item img::attr(src)").get() department = member.css(".nova-legacy-v-person-list-item__stack-item:nth-child(2) span::text").get() desciplines = member.css("span .nova-legacy-e-link::text").getall() institution_memebers.append({ "name": name, "link": link, "profile_photo": profile_photo, "department": department, "descipline": desciplines }) # check for Page not found selector if selector.css(".headline::text").get(): members_is_present = False else: time.sleep(2) # use proxies and captcha solver instead of this page_num += 1 # increment a one. Pagination print(json.dumps(institution_memebers, indent=2, ensure_ascii=False)) print(len(institution_memebers)) # 624 from a EM-Normandie-Business-School browser.close() """ you can also render the page and extract data from the inline JSON string, however, it's messy and from my perspective, it is easier to scrape the page directly. """ # https://regex101.com/r/8qjfnH/1 # extracted_data = re.findall(r"\s+RGCommons\.react\.mountWidgetTree\(({\"data\":{\"menu\".*:true})\);;", # str(page.content()))[0] # json_data = json.loads(extracted_data) # print(json_data) scrape_institution_members(institution="EM-Normandie-Business-School") ``` ### Code explanation Import libraries: ```python from parsel import Selector from playwright.sync_api import sync_playwright import re, json, time ``` | Code | Explanation | | ------------------------------------------------------------------- | ------------------------------------------------------------------------------------ | | [parsel](https://parsel.readthedocs.io/) | to parse HTML/XML documents. Supports XPath. | | [playwright](https://playwright.dev/python/docs/intro#first-script) | to render the page with a browser instance. | | re | to match parts of the data with regular expression. | | json | to convert Python dictionary to JSON string. | | time | is not a practical way to bypass request blocks. Use proxies/captcha solver instead. | Define a function: ```python def scrape_institution_members(institution: str): # ... ``` | Code | Explanation | | ---------------- | ------------------------------------------------- | | institution: str | to tell Python that institution should be an str. | Open a `playwright` with a [context manager](https://book.pythontips.com/en/latest/context%5Fmanagers.html): ```python with sync_playwright() as p: # ... ``` Launch a browser instance, open and `goto` the page and pass response to HTML/XML parser: ```python browser = p.chromium.launch(headless=True, slow_mo=50) page = browser.new_page() page.goto(f"https://www.researchgate.net/institution/{institution}/members/{page_num}") selector = Selector(text=page.content()) ``` | Code | Explanation | | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------- | | [p.chromium.launch()](https://github.com/microsoft/playwright-python/blob/3b7968fb2ea4238a89447b3c7766f9f1f9c9c0e3/playwright/sync%5Fapi/%5Fgenerated.py#L11423) | to launch Chromium browser instance. | | [headless](https://github.com/microsoft/playwright-python/blob/3b7968fb2ea4238a89447b3c7766f9f1f9c9c0e3/playwright/sync%5Fapi/%5Fgenerated.py#L11500-L11504) | to explicitly tell playwright to run in headless mode even though it's a defaut value. | | [slow\_mo](https://github.com/microsoft/playwright-python/blob/3b7968fb2ea4238a89447b3c7766f9f1f9c9c0e3/playwright/sync%5Fapi/%5Fgenerated.py#L11514-L11515) | to tell playwright to slow down execution. | | [browser.new\_page()](https://playwright.dev/python/docs/api/class-browser#browser-new-page) | to open new page. | Add a temporary list, set up a page number, while loop, and check for an exception to exit the loop: ```python institution_memebers = [] page_num = 1 members_is_present = True while members_is_present: # extraction code # check for Page not found selector if selector.css(".headline::text").get(): members_is_present = False else: time.sleep(2) # use proxies and captcha solver instead of this page_num += 1 # increment a one. Pagination ``` Iterate over member results on each page, extract the data and `append` to a temporary `list`: ```python for member in selector.css(".nova-legacy-v-person-list-item"): name = member.css(".nova-legacy-v-person-list-item__align-content a::text").get() link = f'https://www.researchgate.net{member.css(".nova-legacy-v-person-list-item__align-content a::attr(href)").get()}' profile_photo = member.css(".nova-legacy-l-flex__item img::attr(src)").get() department = member.css(".nova-legacy-v-person-list-item__stack-item:nth-child(2) span::text").get() desciplines = member.css("span .nova-legacy-e-link::text").getall() institution_memebers.append({ "name": name, "link": link, "profile_photo": profile_photo, "department": department, "descipline": desciplines }) ``` | Code | Explanation | | ------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | css() | [to parse data from the passed CSS selector(s)](https://github.com/scrapy/parsel/blob/90397dcd0b2c1cbb91e44f65c50f9e11628ba028/parsel/selector.py#L351-L362). Every [CSS query traslates to XPath using csselect package](https://github.com/scrapy/parsel/blob/90397dcd0b2c1cbb91e44f65c50f9e11628ba028/parsel/selector.py#L357-L358) under the hood. | | ::text/::attr(attribute) | [to extract textual or attribute data](https://github.com/scrapy/parsel/blob/90397dcd0b2c1cbb91e44f65c50f9e11628ba028/parsel/csstranslator.py#L48-L51) from the node. | | get()/getall() | [to get actual data from a matched node](https://github.com/scrapy/parsel/blob/90397dcd0b2c1cbb91e44f65c50f9e11628ba028/parsel/selector.py#L197-L204), or to [get a list of matched data from nodes](https://github.com/scrapy/parsel/blob/90397dcd0b2c1cbb91e44f65c50f9e11628ba028/parsel/selector.py#L447-L451). | Print extracted data, `length` of extracted members, and `close` browser instance: ```python print(json.dumps(institution_memebers, indent=2, ensure_ascii=False)) print(len(institution_memebers)) # 624 from a EM-Normandie-Business-School browser.close() ``` Part of the JSON output (first result is a first member, last is the latest member): ```json [ { "name": "Sylvaine Castellano", "link": "https://www.researchgate.netprofile/Sylvaine-Castellano", "profile_photo": "https://i1.rgstatic.net/ii/profile.image/341867548954625-1458518983237_Q64/Sylvaine-Castellano.jpg", "department": "EM Normandie Business School", "descipline": [ "Sustainable Development", "Sustainability", "Innovation" ] }, ... other results { "name": "Constance Biron", "link": "https://www.researchgate.netprofile/Constance-Biron-3", "profile_photo": "https://c5.rgstatic.net/m/4671872220764/images/template/default/profile/profile_default_m.jpg", "department": "Marketing", "descipline": [] } ] ``` ### Extracting data from the JSON string You can scrape the data without using `parsel` by printing page [content()](https://playwright.dev/python/docs/api/class-page#page-content) data which will get the full HTML contents of the page, including the doctype, and parsing the data using regular expression. I'm showing this option as well, just in case some of you prefer this approach over directly parsing the page. ```python # https://regex101.com/r/8qjfnH/1 extracted_data = re.findall(r"\s+RGCommons\.react\.mountWidgetTree\(({\"data\":{\"menu\".*:true})\);;", str(page.content()))[0] json_data = json.loads(extracted_data) print(json_data) ``` Outputs: ![image](https://user-images.githubusercontent.com/78694043/167090719-4c829164-96f8-4142-bcc4-dc96b4fefd72.png) Here's the thing, extracting data seems to be practically more convincing from the JSON string but let's look at the example of accessing the `fullName` key: ``` πŸ‘‡πŸ‘‡πŸ‘‡πŸ‘‡πŸ‘‡ initState.rigel.store.account(id:\\\"AC:2176142\\\").fullName ``` This way, we also got two additional steps: *find* and *compare* user ID to make sure that ID would match the user ID. --- ## Links - [GitHub file in the repository](https://github.com/dimitryzub/serpapi-blog-posts-archive/tree/main/ResearchGate/python/institution-members/researchgate-scrape-all-institution-members-parsel-playwright.py) --- Join us on [Twitter](https://twitter.com/serp%5Fapi) | [YouTube](https://www.youtube.com/channel/UCUgIHlYBOD3yA3yDIRhg%5Fmg) Add a [Feature Request](https://github.com/serpapi/public-roadmap/issues)πŸ’« or a [Bug](https://github.com/serpapi/public-roadmap/issues)🐞