LBAFRIN%clemson.csnet@csnet-relay.arpa (06/18/86)
Unix-Wizards and Info-Unix were so helpful on my last query, I thought I'd throw this one out, too. (I'm still taking responses to that query regarding the purchase of a Gould, Pyramid, Celerity, or VAX!) My organization has need of a program which will scan a text file (presumably a reference work, a journal paper, a manual, documents, stuff like that) and produce an output listing giving, for each word in the input file, the location (line number and input file name) and surrounding context of the word (like the preceding and succeeding 1-10 words). Location should include input file name so that I could invoke the program with a command like "invert *.txt" and identify in which file each listed reference is found. The "ptx" program (comes with System V and other Unices, I think) comes close, but no cigar. It quite often picks up multi-word phrases as keywords. It doesn't give the user the degree of control over keyword and context selection that we need. So "ptx" is out the window. We can of course write the software ourselves, but (you know the story) if it's already out there somewhere... Thanks in advance for any leads you can provide. -- Larry Afrin Dept. of Computer Science Clemson University ================================ Please send replies, if any, to: lbafrin@clemson.csnet or lbafrin%eureka@clemson.csnet or, as a last resort, any reasonable-looking string with "lbafrin", "eureka", and "clemson" in it (And I'm told that Usenet fans can try ihnp4!seismo!clemson.CSNET!lbafrin) I disclaim everything anybody ever said about anything.