CS 3723 Programming Languages  Spring 2014
Questions and Corrections,      with Responses

Please direct questions or corrections or comments to the following email address, where δοτ is a "dot" and ατ is an "at-sign":

<nealδοτwagnerατgmailδοτcom>

I will often post technical questions and my answers to this page, after stripping off any personal matters or identifying portions. The questions may be edited to improve them. Please do NOT use the email address reserved for submissions (the one with "extra" embedded inside) in order to ask questions.

---

1. Question (Sun Feb 16 20:48:00 CST 2014): I have a question regarding the function G() which is supposed to generate the MIPS instructions to get the Input . The recitation has the following sample for the input:

   p = 3 + (1/7);
   > r;
   v = (4/3)*p*r*r*r;
   < p; < N;
   < r; < N;
   < v; < N;
   $

   Could you please let me know what would be the MIPS instructions for reading an input at Runtime and assigning it to a variable. These would be the instructions that the G() function of the compiler should generate.

   Answer: The "Euler series" example, with a link on the calendar page, shows how to write MIPS code to read a double: Euler Series.

   Here, the Tiny statement: "> m; " is translated to the following MIPS:

           # Read M[22] as double
           li      $v0, 7
           syscall
           s.d     $f0, 176($s1)

   This is the same for all reads, except for the "176", which is specific for the "m" in the tiny statement. This should be directly translated to the address "176($S1)" where the variable m should reside.

   ---

2. Question (Mon Feb 17 08:53:37 CST 2014): Recitation 5 refers once to "Recitation 6". Also there is a reference to a non-terminal U and to a parsing function U().

   Answer: Sorry, there were several misprints in Recitation 5. I hope they're all gone now.

   ---

3. Correction (Thu Feb 20 15:49:13 CST 2014): This is based on questions from two students in the recitation session: they were having trouble with the grammar for Recitation 6:

   Old Grammar for Tiny® language
   M −−> L '$' L −−> S { S } S −−> I | W | A | P | G I −−> '[' E '?' L ':' L ']' | '[' E '?' L ']' W −−> '{' E '?' L '}' A −−> lower-case '=' E ';' P −−> '<' E ';' | '<' ('B' | 'N' | 'T') ';' G −−> '>' lower-case ';' E −−> T {('+' | '-') T} T −−> F {('*' | '/') F} F −−> '(' E ')' | lower-case | digit

   Their problem was with the grammar rule:

   L −−>  S { S }

   This works for a short way to describe the language, but using it makes the R-D parser somewhat more difficult. For simplicity, I would get rid of L, and put in S { S } wherever L appeared, to get the following grammar:

   New Grammar for Tiny® language
   M −−> S { S } '$' S −−> I | W | A | P | G I −−> '[' E '?' S { S } ':' S { S } ']' | '[' E '?' S { S } ']' W −−> '{' E '?' S { S } '}' A −−> lower-case '=' E ';' P −−> '<' E ';' | '<' ('B' | 'N' | 'T') ';' G −−> '>' lower-case ';' E −−> T {('+' | '-') T} T −−> F {('*' | '/') F} F −−> '(' E ')' | lower-case | digit

   To write the parser, wherever S { S } appears (which means there is a sequence of one or more statements), keep calling S( ) until next has the proper value to terminate the sequence of calls:

   • Inside M( ) call S( ) until next == '$'
   • Inside W( ) call S( ) until next == '}'
   • Inside I( ) call S( ) until next == ':' or next == ']'

   Notice that each call to S( ) should end with finding one of ';', ']', or '}' in the next variable. After scanning past that character, return from S( ). Then call S( ) again unless you find the proper one of the four characters above.

   (It is possible to use the rule L above: Inside L keep calling S as long as the character in next is the start of a statement, namely, a lower-case letter, or one of '<'. '>', '[', or '{'.)