diff options
| author | Bryan Newbold <bnewbold@archive.org> | 2021-09-08 19:36:43 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@archive.org> | 2021-09-08 19:36:43 -0700 |
| commit | cc240ab5d5c3a9970333abdc0b16867a0dc5c418 (patch) | |
| tree | b310513ef58f02580889b11950ac43f87f213160 | |
| parent | 047cfcebd8ff792d6da9c10b0bee27ec689a4e9f (diff) | |
| download | sandcrawler-cc240ab5d5c3a9970333abdc0b16867a0dc5c418.tar.gz sandcrawler-cc240ab5d5c3a9970333abdc0b16867a0dc5c418.zip | |
old HTML extractors: handle null tag
| -rw-r--r-- | python/sandcrawler/html.py | 17 |
1 files changed, 9 insertions, 8 deletions
diff --git a/python/sandcrawler/html.py b/python/sandcrawler/html.py index e3d95bc..cd0a8e8 100644 --- a/python/sandcrawler/html.py +++ b/python/sandcrawler/html.py @@ -99,14 +99,15 @@ def extract_fulltext_url(html_url, html_body): if 'sciencedirect.com/science/article/pii/' in html_url and not html_url.endswith(".pdf"): json_tag = soup.find("script", attrs={"type": "application/json", "data-iso-key": "_0"}) url = None - try: - json_text = json_tag.string - json_meta = json.loads(json_text) - pdf_meta = json_meta['article']['pdfDownload']['urlMetadata'] - # https://www.sciencedirect.com/science/article/pii/S0169204621000670/pdfft?md5=c4a83d06b334b627ded74cf9423bfa56&pid=1-s2.0-S0169204621000670-main.pdf - url = html_url + pdf_meta['pdfExtension'] + "?md5=" + pdf_meta['queryParams']['md5'] + "&pid=" + pdf_meta['queryParams']['pid'] - except (KeyError, TypeError, json.JSONDecodeError): - pass + if json_tag: + try: + json_text = json_tag.string + json_meta = json.loads(json_text) + pdf_meta = json_meta['article']['pdfDownload']['urlMetadata'] + # https://www.sciencedirect.com/science/article/pii/S0169204621000670/pdfft?md5=c4a83d06b334b627ded74cf9423bfa56&pid=1-s2.0-S0169204621000670-main.pdf + url = html_url + pdf_meta['pdfExtension'] + "?md5=" + pdf_meta['queryParams']['md5'] + "&pid=" + pdf_meta['queryParams']['pid'] + except (KeyError, TypeError, json.JSONDecodeError): + pass if url: return dict(pdf_url=url, technique="sciencedirect-munge-json") |