Something Claude and I have run into over these months is that certain non-Latin scripts don’t cleanly get extracted from the sources and end up in the resulting post. Claude omits these, and I have chosen not to go in and try to manually add the omitted names. I assume that you, dear readers, are reading more for the higher-level detail and not every smidgen of data (at least I hope not). In that regard, what’s being left out are non-Latin versions of party names. For example, in my most recent OFAC designations post, it could not extract the Chinese names clearly, and the Arabic came out reversed.

If I try to cut and paste from the HTML or a PDF, I get:

ARABIANPRO MATERIALS FZCO (Arabic: ارابيانبرو ماتيريالز ش م ح), Dubai, United Arab Emirates; Executive Order 13846 information: BLOCKING PROPERTY AND INTERESTS IN PROPERTY. Sec. 5(a)(iv); Organization Established Date 2024; License DV1082 (United Arab Emirates); Business Registration Number 49879 (United Arab Emirates) [IRAN-EO13846].

HECHUANG INTERNATIONAL GROUP LIMITED (Chinese Traditional: 和創國際集團有限公司), Room 3707, Tower B, China Resources Building, 6, Shandong Lu, Shinan Qu, Qingdao, Shandong, China; Hong Kong, China; Organization Established Date 08 Aug 2024; Identification Number IMO 0032920; Business Registration Number 76912518 (Hong Kong) [IRAN-EO13902].

which does actually come out properly… I will work on it with Claude…

[Update] we’re now working off the HTML instead of the PDF, and we’ve updated the master prompts. So, now, hopefully, you’ll get it all.