--- title: Eboard 06 Pattern matching with regular expressions number: 6 section: eboards held: 2019-02-06 link: true --- CSC 151 2019S, Class 06: Pattern matching with regular expressions =================================================================== _Overview_ * Preliminaries * Notes and news * Upcoming work * Extra credit * Questions * Lab! * Time to meet with HW partners. * Debrief (if time) Preliminaries ------------- ### News / Etc. * Don't forget that we have mentor sessions tonight 8-9pm and Thursday 8-9 in the CS commons. * Mentors should not be asked for help outside of class time and mentor sessions. Feel free to use evening tutors and to send me email. * As we'll discuss on Friday, we'd really like you to work together on problems. (If you'd like to think about problems in advance, that's okay, but there's good benefit to working together.) * Even after I talked about pulling the keyboard away from your partner, some of you did that. * The [current flash cards page](../flashcards/flashcards03) now contains one set of questions I've received. (If all goes well, it will soon contain all of them.) (At some point, I'll create a better UI.) * I'll try to remember to set the "switch person at keyboard" alarm. * Do any statisticians want to compute the odds that people will get the same partner a second time? (Assume we have 34 students and this is the fifth time we've had random partners.) EC for first two or three answers. ### Upcoming work * Readings due before class Friday * Pair programming (ready) * How Scheme works (not ready) * [Assignment 4](../assignments/assignment04) due Tuesday night. * Partners to be assigned via email. * [Flash cards](../flashcards/flashcards03) due TONIGHT at 8:00 p.m. * No lap writeups. * Quiz Friday * Basic types * Writing procedures ### Extra Credit #### Extra credit (Academic/Artistic) * John Garrison reads from _Shakespeare and the Afterlife_, Thursday, 7 February 2019 in the Faculconer gallery. * _Once Upon a Time Wolf_ (tickets required), Bucksbaum. Friday, 8 February, 7pm. * _Once Upon a Time Wolf_ (tickets required), Bucksbaum. Saturday, 9 February, 7pm * Any Data Week activity next week. * HackGC weekend of 15-17 February 2019. (I'm still looking for links.) * Friday night Gardner concert, 8:30 p.m. #### Extra credit (Peer) * Home track meet, Saturday, 9 Feb 2019, all-day and beyond. (30 min suffices) * Conference Swim and Dive meet, 15-17 February 2019. Dive times to be announced later. #### Extra credit (Wellness) * HIIT training, 4:30 pm, Tuesday, Dance Studio, Bear. (Cap of two EC units.) * Hatha Yoga, 7:00 pm, Tuesday, Dance Studo, Bear. (Cap of two EC units.) * Brazilian Jiu-Itsu, Wednesday and Friday, 6:30, Dance Studio (cap of two EC units.) * Any Sex Week activity next week. #### Extra credit (Misc) ### Other good things ### Questions How do we get the loudhum libraries on our personal computers? > Start DrRacket > **File** > **Install Package** > Enter "https://github.com/grinnell-cs/loudhum.git" (without the quotation marks). > Click **Install* or **Update** (whichever appears). > When the **Close** button becomes available, click it. Could you explain `#px"a[a-z]*a"` and how it differs from `#px"a[a-z]+a"`? > Sure. (or at least I can try) > Three parts: `a`, `[a-z]*`, `a` > The `a`'s match the letter a (and nothing else). So we are looking for strings that start with a and end with a. > We'll pull apart the `[a-z]*`. `[a-z]` is shorthand for a or b or c or d or e or ... or z. "Lowercase letters". > An expression followed by a star means "zero or more copies" `[a-z]*` means "0 or more lowercase letters". > `#px"a[a-z]*a"` means "sequences of characters that start with a, end with a, and have only lowercase letters in between." > (regexp-match* #px"a[a-z]*a" "alphabet aardvark aardwolf samr") '("alpha" "aardva" "aa") > An expression followed by a plus means "one or more copies" `[a-z]+` means "1 or more lowercase letters". > `#px"a[a-z]+a"` means "sequences of characters that start with a, end with a, and have at least one lowercase letter in between." Why did we get "aardva" from "aardvark", rather than "aa"? > In general, Racket looks for the *longest* string that matches the pattern. Do we need the `#px`? > Often, but not always. It's safer to include it. Could you explain the `"\\1\\2"` replacement? > Once again, I can try. Apologies for limited creativity. > We can parenthesize parts of an expression. Sometimes for clarity. Sometimes to deal with precedence issues. `#px"ab+"` means "a followed by at least one b." `#px"(ab)+"` means "sequences of repeating abababab" > (regexp-match* #px"ab+" "abbbba ababab") '("abbbb" "ab" "ab" "ab") > (regexp-match* #px"(ab)+" "abbbba ababab") '("ab" "ababab") > We may want to refer to things from the pattern when we do a replacement. For example, I may want to replace "X and Y" with "Y and X". > The pattern is `#px"(\\S+) and (\\S+)` > (regexp-match* #px"(\\S+) and (\\S+)" "pb and j, rock and roll, foo and bar") '("pb and j," "rock and roll," "foo and bar") > The replacement is "\\2 and \\1" > (regexp-replace* #px"(\\S+) and (\\S+)" "pb and j, rock and roll, foo and bar" "\\2 and \\1") "j, and pb roll, and rock bar and foo" Did you really want the comma? > No. Sometimes I write bad regular expressions. What if I want aa or ee or ii or oo or uu > `#px"(aa|ee|ii|oo|uu)"` Lab --- How do I get the first twenty characters? > `(take book-letters 20)` How do I figure out how many letters? > `(length book-letters)` or `(string-length book-contents)` How do get rid of letters or words or lines from a list? > `(drop lst num)` Meet with HW partners ---------------------