path: root/check_issn_urls.py

#!/usr/bin/env python3
"""
Check journal homepage status (live web and wayback)

Takes a tsv filepath or lines on stdin and dumps to stdout. The stdin thing means you can:

    # be sure to randomize order if you are going to use high parallelism so no
    # individual domain gets swamped. also remember this hits CDX API multiple
    # times.
    parallel -j10 --bar --pipepart -a urls_to_crawl.shuf.tsv ./check_issn_urls.py > url_status.json

Input columns (no header):

    ISSN-L, URL

For each URL, do a request and record, as JSON:

    issnl: passthrough
    url: passthrough
    status_code: initial HTTP crawl status
    terminal_url: final URL (or original if no redirects)
    terminal_status_code: final URL (or original if no redirects)
    terminal_content_type: content type (mimetype)
    platform_software: slug of hosting platform, if detected
    issnl_in_body: whether raw issnl appears in body text
    blocked: whether we think crawler was "blocked"
    gwb_url_success_dt: latest wayback datetime that an HTTP 200 exists
    gwb_terminal_url_success_dt: latest wayback datetime that an HTTP 200 exists

HTTP status will be -1 if domain does not even resolve.
"""

import sys
import json
import time
import requests


def sniff_platform(resp):
    """
    This function would try to figure out what software platform (eg, OJS) the
    site is running.
    TODO: unimplemented
    """
    # these are mostly here to filter out huge platforms and stop sniffing
    domain_map = {
        "jstor.org/": "jstor",
        "springer.com/": "springer",
        "springerlink.com/": "springer",
        "tandfonline.com/": "t_and_f",
        "elsevier.com/": "elsevier",
        "wiley.com/": "wiley",
        "sciencedirect.com/": "elsevier",
        "sagepub.com/": "sage",
        "hypotheses.org/": "hypothesis",
        "tandf.co.uk/": "t_and_f",
        "scielo": "scielo",
    }
    for domain, platform in domain_map.items():
        if domain in resp.url:
            return platform
    if '<meta name="generator" content="Open Journal Systems' in resp.text:
        return "ojs"
    return None


def sniff_blocked(resp):
    """
    This function would try to figure out if we got blocked: soft-block, hard
    block, etc.
    TODO: unimplemented
    """
    if resp.status_code in (403, 420, 429):
        return True
    # JSTOR does this
    if (
        "Our systems have detected unusual traffic activity from your network. Please complete this reCAPTCHA"
        in resp.text
    ):
        return True
    if (
        resp.status_code == 416
        and "something about your browser made us think you were a bot" in resp.text
    ):
        return True
    return None


def check_gwb(url, match_type="exact"):
    if "//web.archive.org/" in url:
        return None
    # crude/bad retry loop to work around CDX API throttling
    for i in range(2):
        try:
            resp = requests.get(
                "https://web.archive.org/cdx/search/cdx",
                params={
                    "url": url,
                    "matchType": match_type,
                    "limit": -1,
                    "filter": "statuscode:200",
                },
            )
        except Exception:
            # nasty blanket catch
            return None
        if resp.status_code not in [200, 404]:
            break
        time.sleep(0.1)
    if not resp.status_code == 200:
        sys.stderr.write("CDX ERR {}: {}\n".format(resp.status_code, url))
        # TODO: this isn't really correct, but not sure what to return/record
        # if we failed through all timeouts
        return None
    line = resp.text.strip().split("\n")[0]
    if line:
        dt = line.split()[1]
        int(dt)
        return dt
    else:
        return None

HOST_SKIP_LIST = [
    "www.jstor.org",
    "www.tandfonline.com",
    "www.sciencedirect.com",
    "link.springer.com",
    "onlinelibrary.wiley.com",
    "dialnet.unirioja.es",
    "www.springer.com",
    "www.journals.elsevier.com",
    "web.archive.org",
    "catalog.hathitrust.org",
]

def check_url(issnl, url):
    # print("Fetching: %s" % url)
    info = dict(issnl=issnl, url=url)
    if "://" not in url:
        info["error"] = "bad-url"
        info["terminal_status_code"] = -1
        return info
    if not url.startswith('http'):
        info["error"] = "url-not-http"
        info["terminal_status_code"] = -1
        return info
    for host in HOST_SKIP_LIST:
        if f"://{host}/" in url:
            info["error"] = "skip-host"
            info["terminal_status_code"] = -1
            return info

    try:
        resp = requests.get(
            url,
            timeout=(5.0, 5.0),
            headers={
                "User-Agent": "ia_bot/0.0 (python requests) journal-live-check; contact:info@archive.org"
            },
        )
    except requests.exceptions.TooManyRedirects:
        info["error"] = "TooManyRedirects"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.SSLError:
        info["error"] = "SSLError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.ReadTimeout:
        info["error"] = "ReadTimeout"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.ConnectionError:
        info["error"] = "ConnectionError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.ChunkedEncodingError:
        info["error"] = "ChunkedEncodingError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.ContentDecodingError:
        info["error"] = "ContentDecodingError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.InvalidSchema:
        info["error"] = "InvalidSchema"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except ConnectionResetError:
        info["error"] = "ConnectionResetError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except requests.exceptions.InvalidURL:
        info["error"] = "ProtocolError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except UnicodeDecodeError:
        info["error"] = "UnicodeDecodeError"
        info["terminal_status_code"] = info["status_code"] = -1
        return info
    except Exception as e:
        # nasty blanket catch
        print(e, file=sys.stderr)
        info["error"] = "other"
        info["terminal_status_code"] = info["status_code"] = -1
        return info

    if resp.history:
        info["status_code"] = resp.history[0].status_code
    else:
        info["status_code"] = resp.status_code

    info["terminal_status_code"] = resp.status_code
    info["terminal_url"] = resp.url
    content_type = resp.headers.get("Content-Type")
    if content_type:
        info["terminal_content_type"] = content_type.split(";")[0]
    info["issnl_in_body"] = bool(issnl in resp.text)
    info["gwb_url_success_dt"] = check_gwb(url, match_type="exact")
    info["gwb_terminal_url_success_dt"] = check_gwb(
        info["terminal_url"], match_type="exact"
    )
    info["blocked"] = sniff_blocked(resp)
    info["software_platform"] = sniff_platform(resp)
    # info['gwb_host_success_dt'] = check_gwb(url, match_type='host')
    return info


def run(tsvfile):
    for line in tsvfile:
        records = line.split("\t")
        issnl = records[0]
        url = records[1].strip()
        print(json.dumps(check_url(issnl, url)))


if __name__ == "__main__":
    if len(sys.argv) != 2:
        f = sys.stdin
    else:
        f = open(sys.argv[1], "r")
    run(f)