That led me in the same direction as you, and I checked out the live, Java versions of the puzzles on The Herald's site. Unfortunately, the solution grids are never arrays or tables, but images, and I am neither a graphics nor OCR programmer. In particular, I want to get the puzzles from the Herald (Glasgow, Scotland) into. PUZ files for cryptic crosswords, I recently thought about seeing if I could decode the "printer-safe" puzzles that appear on many sites. Having long been frustrated by the lack of. I'll let you have it, but you'll have to run it yourself to get the puzzles, and I can't guarantee it will work. I have a similar script to retrieve New York Times puzzles, but it won't work unless you're a subscriber. PUZ puzzles up to the current day), and the LA Times (performs archive verification only, and only back to the date when I started retrieving those puzzles from the link at the Cruciverbalist site). I can also share archives and scripts for Newsday (retrieves separate PDF files for puzzles and solutions), the New York Sun (retrieves unlocked. If there's a reasonable level of interest, I will put it up on my web site. PUZ files from November 16, 1997, through today, along with the Perl source code that will help you keep your archive current (run it with -t to retrieve today's puzzle and with -v to verify that you have all of the puzzles), please let me know. The Post, unlike the Times, makes its puzzles available free to anyone with a web browser, so if you would like a. It's running now, sucking down all the puzzles from Novem(the first day that's available, I found by trial and error) to the present date. So I took my Perl script that retrieves the New York Times puzzle and modified it to retrieve the Washington Post puzzle, strip off the first 13 characters, and save it with a. It opened up in Across Lite just fine and had a revealable solution (no unscrambling necessary) even though it's the current day's puzzle. I deleted that using the hex editor and saved the file. As it turns out, the only difference is that very first thing in the Washington Post puzzle is a 12-character string in the format csyymmdd.puz followed by a line feed (ASCII 10) character. puz file from the New York Times archive. I printed the hex dump (just two pages), then opened (also in the hex editor) a known-good. It looked suspiciously like an Across Lite puzzle. Imagine my disappointment when it came up corrupt.Īny other day, I might have given up at this point, but I'm extremely bored today, so I opened up the file in a hex editor. puz extension and tried to open it in Across Lite. It immediately prompted me to save the file. I opened another copy of my browser and pasted the URL into the address bar. I viewed the page source, and there was a URL. I know that's the company behind Across Lite, so I thought maybe the puzzles were stored somewhere in Across Lite format and converted to Java on the fly. While looking at the Java version of today's puzzle, I saw a little "Literate Software" tag in the lower right corner. I prefer Across Lite puzzles because I can save them to my hard drive, solve them offline with the Across Lite software, or print them from the Across Lite software for lunchtime or airplane solving. Unfortunately, it seemed like the choices were (a) a Java-based puzzle you had to solve online, or (b) a printable copy that you had to print off and solve on paper. I recently discovered the Washington Post crossword puzzle archive.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |