Well, you're right of course. I saw the original question, thought "I think 1a is a SUB" and was going to look it up to be sure. Then someone else responded that it was a SUB and I just went with that. My bad.
So the sequence may have been something like Input perhaps in UTF-8 translated to ASCII. Untranslatable character converted correctly to 1A. ASCII translated to EBCDIC and the 1A incorrectly preserved as 1A. 3270 emulator barfs on receipt of a 1A Will be interesting to hear back from the OP about my theory of a cut-and-paste of a curved quote. Hey, it is wrong to dis curved quotes. They are proper quotes. All quotes used to be curved quotes. Then the typewriter came along and they economized on keys by going to a single straight quote. ASCII and EBCDIC followed suit. Yes, automatic conversion of "typewriter quotes" to proper quotes can cause problems, but don't blame the quotes. They're right, at least historically speaking. Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf Of Paul Gilmartin Sent: Thursday, October 28, 2021 6:10 PM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: Invalid character in file On Thu, 28 Oct 2021 14:54:56 -0700, Charles Mills wrote: >1A is the EBCDIC sub character as @Mike says. Something is taking input in >character set A and translating it to character set B, finding a character >that is not in character set B, and properly substituting a 1A. > I believe 1A is the ASCII SUB; EBCDIC SUB is 3F. $ printf '\x1A' | iconv -f ISO8859-1 -t IBM-1047 | od -tx1 0000000 3f >Then TN3270 is saying "I don't know how to display a 1A" and barfing. > >Apostrophe, huh? A wild guess would be that the character that is not liked >might be a "smart quote": ’ > ITYM "smartass quote". But good guess. >I don't think they could truly key that in but they might count cut-and-paste >as "keying in." >-----Original Message----- >On Thu, Oct 28, 2021 at 8:24 PM Tony Thigpen wrote: >> >> This is a new one for me. >> >> I have a customer that has sites in both the US and Canada. In the last >> week, something strange has happened in their files TWICE. >> >> They are using CICS to add a new vendor to an IMS database. The next >> time they try to access the vendor record, the terminal session gets a >> prog-something because the customer name now has a x'1A' in the name >> where it should have been an apostrophe. TN3270 just does not like a >> x'1A' in a data field. >> >> For example: >> GROUT.S WELDING LIMITED >> CDDEE1E4ECDCCDC4DCDCECC44 >> 79643A2065349570394935400 >> >> I asked if they are creating new vendor records via an off-platform >> feed, but they say the are keying them manually in a CICS terminal session. >> >> I do know that both times this has happened, the vendor was a Canadian >> supplier so maybe this has something to do with French keyboards? >> (Grasping for straws, maybe?) >> >> Anybody seen this or think they may know what is happening? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN