1 (*<*) |
|
2 theory Slides3 |
|
3 imports "~~/src/HOL/Library/LaTeXsugar" "Nominal" |
|
4 begin |
|
5 |
|
6 declare [[show_question_marks = false]] |
|
7 |
|
8 notation (latex output) |
|
9 set ("_") and |
|
10 Cons ("_::/_" [66,65] 65) |
|
11 |
|
12 (*>*) |
|
13 |
|
14 text_raw {* |
|
15 \renewcommand{\slidecaption}{UNIF, Edinburgh, 14.~July 2010} |
|
16 |
|
17 \newcommand{\abst}[2]{#1.#2}% atom-abstraction |
|
18 \newcommand{\pair}[2]{\langle #1,#2\rangle} % pairing |
|
19 \newcommand{\susp}{{\boldsymbol{\cdot}}}% for suspensions |
|
20 \newcommand{\unit}{\langle\rangle}% unit |
|
21 \newcommand{\app}[2]{#1\,#2}% application |
|
22 \newcommand{\eqprob}{\mathrel{{\approx}?}} |
|
23 \newcommand{\freshprob}{\mathrel{\#?}} |
|
24 \newcommand{\redu}[1]{\stackrel{#1}{\Longrightarrow}}% reduction |
|
25 \newcommand{\id}{\varepsilon}% identity substitution |
|
26 |
|
27 \pgfdeclareradialshading{smallbluesphere}{\pgfpoint{0.5mm}{0.5mm}}% |
|
28 {rgb(0mm)=(0,0,0.9); |
|
29 rgb(0.9mm)=(0,0,0.7); |
|
30 rgb(1.3mm)=(0,0,0.5); |
|
31 rgb(1.4mm)=(1,1,1)} |
|
32 |
|
33 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
34 \mode<presentation>{ |
|
35 \begin{frame}<1>[c] |
|
36 \frametitle{Quiz} |
|
37 |
|
38 Assuming that \smath{a} and \smath{b} are distinct variables,\\ |
|
39 is it possible to find $\lambda$-terms \smath{M_1} to \smath{M_7} |
|
40 that make the following pairs \alert{$\alpha$-equivalent}? |
|
41 |
|
42 \begin{tabular}{@ {\hspace{14mm}}p{12cm}} |
|
43 \begin{itemize} |
|
44 \item \smath{\lambda a.\lambda b. (M_1\,b)\;} and |
|
45 \smath{\lambda b.\lambda a. (a\,M_1)\;} |
|
46 |
|
47 \item \smath{\lambda a.\lambda b. (M_2\,b)\;} and |
|
48 \smath{\lambda b.\lambda a. (a\,M_3)\;} |
|
49 |
|
50 \item \smath{\lambda a.\lambda b. (b\,M_4)\;} and |
|
51 \smath{\lambda b.\lambda a. (a\,M_5)\;} |
|
52 |
|
53 \item \smath{\lambda a.\lambda b. (b\,M_6)\;} and |
|
54 \smath{\lambda a.\lambda a. (a\,M_7)\;} |
|
55 \end{itemize} |
|
56 \end{tabular} |
|
57 |
|
58 If there is one solution for a pair, can you describe all its solutions? |
|
59 |
|
60 \end{frame}} |
|
61 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
62 *} |
|
63 |
|
64 text_raw {* |
|
65 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
66 \mode<presentation>{ |
|
67 \begin{frame}<1>[t] |
|
68 \frametitle{% |
|
69 \begin{tabular}{@ {\hspace{-3mm}}c@ {}} |
|
70 \\ |
|
71 \huge Nominal Unification\\[-2mm] |
|
72 \Large Hitting a Sweet Spot\\[5mm] |
|
73 \end{tabular}} |
|
74 \begin{center} |
|
75 Christian Urban |
|
76 \end{center} |
|
77 \begin{center} |
|
78 \small initial spark from Roy Dyckhoff in November 2001\\[0mm] |
|
79 \small joint work with Andy Pitts and Jamie Gabbay\\[0mm] |
|
80 \end{center} |
|
81 \end{frame}} |
|
82 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
83 |
|
84 *} |
|
85 text_raw {* |
|
86 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
87 \mode<presentation>{ |
|
88 \begin{frame}<1-4>[c] |
|
89 \frametitle{One Motivation} |
|
90 |
|
91 \onslide<2->{Typing implemented in Prolog \textcolor{darkgray}{(from a textbook)}}\bigskip\\ |
|
92 |
|
93 \onslide<3->{\color{darkgray} |
|
94 \begin{tabular}{l} |
|
95 type (Gamma, var(X), T) :- member (X,T) Gamma.\smallskip\medskip\\ |
|
96 |
|
97 type (Gamma, app(M, N), T') :-\\ |
|
98 \hspace{3cm}type (Gamma, M, arrow(T, T')),\\ |
|
99 \hspace{3cm}type (Gamma, N, T).\smallskip\medskip\\ |
|
100 |
|
101 type (Gamma, lam(X, M), arrow(T, T')) :-\\ |
|
102 \hspace{3cm}type ((X, T)::Gamma, M, T').\smallskip\medskip\\ |
|
103 |
|
104 member X X::Tail.\\ |
|
105 member X Y::Tail :- member X Tail.\\ |
|
106 \end{tabular}} |
|
107 |
|
108 \only<4>{ |
|
109 \begin{textblock}{6}(2.5,2) |
|
110 \begin{tikzpicture} |
|
111 \draw (0,0) node[inner sep=3mm,fill=cream, ultra thick, draw=red, rounded corners=2mm] |
|
112 {\color{darkgray} |
|
113 \begin{minipage}{8cm}\raggedright |
|
114 The problem is that \smath{\lambda x.\lambda x. (x\;x)} |
|
115 will have the types |
|
116 \begin{center} |
|
117 \begin{tabular}{l} |
|
118 \smath{T\rightarrow (T\rightarrow S) \rightarrow S} and\\ |
|
119 \smath{(T\rightarrow S)\rightarrow T \rightarrow S}\\ |
|
120 \end{tabular} |
|
121 \end{center} |
|
122 \end{minipage}}; |
|
123 \end{tikzpicture} |
|
124 \end{textblock}} |
|
125 |
|
126 \end{frame}} |
|
127 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
128 *} |
|
129 |
|
130 text_raw {* |
|
131 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
132 \mode<presentation>{ |
|
133 \begin{frame}<1>[c] |
|
134 \frametitle{Higher-Order Unification} |
|
135 |
|
136 State of the art at the time: |
|
137 |
|
138 \begin{itemize} |
|
139 \item Lambda Prolog with full Higher-Order Unification\\ |
|
140 \textcolor{darkgray}{(no mgus, undecidable, modulo $\alpha\beta$)}\bigskip |
|
141 \item Higher-Order Pattern Unification\\ |
|
142 \textcolor{darkgray}{(has mgus, decidable, some restrictions, modulo $\alpha\beta_0$)} |
|
143 \end{itemize} |
|
144 |
|
145 \end{frame}} |
|
146 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
147 *} |
|
148 |
|
149 text_raw {* |
|
150 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
151 \mode<presentation>{ |
|
152 \begin{frame}<1-10>[t] |
|
153 \frametitle{Underlying Ideas} |
|
154 |
|
155 \begin{itemize} |
|
156 \item<1-> Unification (\alert{only}) up to $\alpha$ |
|
157 |
|
158 \item<2-> Swappings / Permutations |
|
159 |
|
160 \only<2-5>{ |
|
161 \begin{center} |
|
162 \begin{tabular}{r@ {\hspace{1mm}}l@ {\hspace{12mm}}r@ {\hspace{1mm}}l} |
|
163 \\ |
|
164 \only<2>{\smath{\textcolor{white}{[b\!:=\!a]}}}% |
|
165 \only<3>{\smath{[b\!:=\!a]}}% |
|
166 \only<4-5>{\smath{\alert{\swap{a}{b}\,\act}}} & |
|
167 \onslide<2-5>{\smath{\lambda a.b}} & |
|
168 |
|
169 \only<2>{\smath{\textcolor{white}{[b\!:=\!a]}}}% |
|
170 \only<3>{\smath{[b\!:=\!a]}}% |
|
171 \only<4-5>{\smath{\alert{\swap{a}{b}\,\act}}} & |
|
172 \onslide<2-5>{\smath{\lambda c.b}}\\ |
|
173 |
|
174 \onslide<3-5>{\smath{=}} & \only<3>{\smath{\lambda a.a}}\only<4-5>{\smath{\lambda b.a}} & |
|
175 \onslide<3-5>{\smath{=}} & \only<3>{\smath{\lambda c.a}}\only<4-5>{\smath{\lambda c.a}}\\ |
|
176 \end{tabular} |
|
177 \end{center}\bigskip |
|
178 |
|
179 \onslide<4-5>{ |
|
180 \begin{center} |
|
181 \begin{tikzpicture} |
|
182 \draw (0,0) node[inner sep=0mm,fill=cream, ultra thick, draw=cream] |
|
183 {\begin{minipage}{8cm} |
|
184 \begin{tabular}{r@ {\hspace{3mm}}l} |
|
185 \smath{\swap{a}{b}\act t} $\;\dn$ & \alert{swap} {\bf all} occurrences of\\ |
|
186 & \smath{b} and \smath{a} in \smath{t} |
|
187 \end{tabular} |
|
188 \end{minipage}}; |
|
189 \end{tikzpicture} |
|
190 \end{center}}\bigskip |
|
191 |
|
192 \onslide<5>{ |
|
193 Unlike for \smath{[b\!:=\!a]\act(-)}, for \smath{\swap{a}{b}\act (-)} we do |
|
194 have if \smath{t =_\alpha t'} then \smath{\pi \act t =_\alpha \pi \act t'.}}} |
|
195 |
|
196 \item<6-> Variables (or holes)\bigskip |
|
197 |
|
198 \begin{center} |
|
199 \onslide<7->{\mbox{}\hspace{-25mm}\smath{\lambda x\hspace{-0.5mm}s .}} |
|
200 \onslide<8-9>{\raisebox{-1.7mm}{\huge\smath{(}}}\raisebox{-4mm}{\begin{tikzpicture} |
|
201 \fill[blue] (0, 0) circle (5mm); |
|
202 \end{tikzpicture}} |
|
203 \onslide<8-9>{\smath{y\hspace{-0.5mm}s}{\raisebox{-1.7mm}{\huge\smath{)}}}}\bigskip |
|
204 \end{center} |
|
205 |
|
206 \only<8-9>{\smath{y\hspace{-0.5mm}s} are the parameters the hole can depend on\onslide<9->{, but |
|
207 then you need $\beta_0$-reduction\medskip |
|
208 \begin{center} |
|
209 \smath{(\lambda x. t) y \longrightarrow_{\beta_0} t[x:=y]} |
|
210 \end{center}}} |
|
211 |
|
212 \only<10>{we will record the information about which parameters a hole |
|
213 \alert{\bf cannot} depend on} |
|
214 |
|
215 \end{itemize} |
|
216 |
|
217 \end{frame}} |
|
218 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
219 *} |
|
220 |
|
221 text_raw {* |
|
222 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
223 \mode<presentation>{ |
|
224 \begin{frame}<1-4>[c] |
|
225 \frametitle{Terms} |
|
226 |
|
227 \begin{tabular}{lll @ {\hspace{10mm}}lll} |
|
228 |
|
229 \onslide<1->{\pgfuseshading{smallbluesphere}} & |
|
230 \onslide<1->{\colorbox{cream}{\smath{\unit}}} & |
|
231 \onslide<1->{Units} & |
|
232 |
|
233 \onslide<2->{\pgfuseshading{smallbluesphere}} & |
|
234 \onslide<2->{\colorbox{cream}{\smath{a}}} & |
|
235 \onslide<2->{Atoms} \\[5mm] |
|
236 |
|
237 \onslide<1->{\pgfuseshading{smallbluesphere}} & |
|
238 \onslide<1->{\colorbox{cream}{\smath{\pair{t}{t'}}}} & |
|
239 \onslide<1->{Pairs} & |
|
240 |
|
241 \onslide<3->{\pgfuseshading{smallbluesphere}} & |
|
242 \onslide<3->{\colorbox{cream}{\smath{\abst{a}{t}}}} & |
|
243 \onslide<3->{Abstractions}\\[5mm] |
|
244 |
|
245 \onslide<1->{\pgfuseshading{smallbluesphere}} & |
|
246 \onslide<1->{\colorbox{cream}{\smath{\app{F}{t}}}} & |
|
247 \onslide<1->{Funct.} & |
|
248 |
|
249 \onslide<4->{\pgfuseshading{smallbluesphere}} & |
|
250 \onslide<4->{\colorbox{cream}{\smath{\pi\susp X}}} & |
|
251 \onslide<4->{Suspensions} |
|
252 \end{tabular} |
|
253 |
|
254 \only<2>{ |
|
255 \begin{textblock}{13}(1.5,12) |
|
256 \small Atoms are constants \textcolor{darkgray}{(infinitely many of them)} |
|
257 \end{textblock}} |
|
258 |
|
259 \only<3>{ |
|
260 \begin{textblock}{13}(1.5,12) |
|
261 \small \smath{\ulcorner \lambda\abst{a}{a}\urcorner \mapsto \text{fn\ }\abst{a}{a}}\\ |
|
262 \small constructions like \smath{\text{fn\ }\abst{X}{X}} are not allowed |
|
263 \end{textblock}} |
|
264 |
|
265 \only<4>{ |
|
266 \begin{textblock}{13}(1.5,12) |
|
267 \small \smath{X} is a variable standing for a term\\ |
|
268 \small \smath{\pi} is an explicit permutation \smath{\swap{a_1}{b_1}\ldots\swap{a_n}{b_n}}, |
|
269 waiting to be applied to the term that is substituted for \smath{X} |
|
270 \end{textblock}} |
|
271 |
|
272 \end{frame}} |
|
273 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
274 *} |
|
275 |
|
276 text_raw {* |
|
277 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
278 \mode<presentation>{ |
|
279 \begin{frame}<1-3>[c] |
|
280 \frametitle{Permutations} |
|
281 |
|
282 a permutation applied to a term |
|
283 |
|
284 \begin{center} |
|
285 \begin{tabular}{lrcl} |
|
286 \pgfuseshading{smallbluesphere} & |
|
287 \smath{[]\act c} & \smath{\dn} & \smath{c} \\ |
|
288 |
|
289 \pgfuseshading{smallbluesphere} & |
|
290 \smath{\swap{a}{b}\!::\!\pi\act c} & \smath{\dn} & |
|
291 \smath{\begin{cases} |
|
292 a & \text{if}\;\pi\act c = b\\ |
|
293 b & \text{if}\;\pi\act c = a\\ |
|
294 \pi\act c & \text{otherwise} |
|
295 \end{cases}}\\ |
|
296 |
|
297 \onslide<2->{\pgfuseshading{smallbluesphere}} & |
|
298 \onslide<2->{\smath{\pi\act\abst{a}{t}}} & \onslide<2->{\smath{\dn}} & |
|
299 \onslide<2->{\smath{\abst{\pi\act a}{\pi\act t}}}\\ |
|
300 |
|
301 \onslide<3->{\pgfuseshading{smallbluesphere}} & |
|
302 \onslide<3->{\smath{\pi\act\pi'\act X}} & \onslide<3->{\smath{\dn}} & |
|
303 \onslide<3->{\smath{(\pi @ \pi')\act X}}\\ |
|
304 \end{tabular} |
|
305 \end{center} |
|
306 |
|
307 \end{frame}} |
|
308 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
309 *} |
|
310 |
|
311 text_raw {* |
|
312 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
313 \mode<presentation>{ |
|
314 \begin{frame}<1-3>[c] |
|
315 \frametitle{Freshness Constraints} |
|
316 |
|
317 Recall \smath{\lambda a. \raisebox{-0.7mm}{\tikz \fill[blue] (0, 0) circle (2.5mm);}} |
|
318 \bigskip\pause |
|
319 |
|
320 We therefore will identify |
|
321 |
|
322 \begin{center} |
|
323 \smath{\text{fn\ } a. X \;\approx\; \text{fn\ } b. \alert<3->{\swap{a}{b}}\act X} |
|
324 \end{center} |
|
325 |
|
326 provided that `\smath{b} is fresh for \smath{X} --- (\smath{b\fresh X})', |
|
327 i.e., does not occur freely in any ground term that might be substituted for |
|
328 \smath{X}.\bigskip\pause |
|
329 |
|
330 If we know more about \smath{X}, e.g., if we knew that \smath{a\fresh X} and |
|
331 \smath{b\fresh X}, then we can replace\\ \smath{\swap{a}{b}\act X} by |
|
332 \smath{X}. |
|
333 |
|
334 \end{frame}} |
|
335 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
336 *} |
|
337 |
|
338 text_raw {* |
|
339 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
340 \mode<presentation>{ |
|
341 \begin{frame}<1-4>[c] |
|
342 \frametitle{Equivalence Judgements} |
|
343 |
|
344 \alt<1>{Our equality is {\bf not} just}{but judgements} |
|
345 |
|
346 \begin{center} |
|
347 \begin{tabular}{rl} |
|
348 \colorbox{cream}{\smath{\onslide<2->{\nabla \vdash} t \approx t'}} & \alert{$\alpha$-equivalence}\\[1mm] |
|
349 \onslide<4->{\colorbox{cream}{\smath{\onslide<2->{\nabla \vdash} a \fresh t}}} & |
|
350 \onslide<4->{\alert{freshness}} |
|
351 \end{tabular} |
|
352 \end{center} |
|
353 |
|
354 \onslide<2->{ |
|
355 where |
|
356 \begin{center} |
|
357 \smath{\nabla = \{a_1\fresh X_1,\ldots, a_n\fresh X_n\}} |
|
358 \end{center} |
|
359 is a finite set of \alert{freshness assumptions}.} |
|
360 |
|
361 \onslide<3->{ |
|
362 \begin{center} |
|
363 \smath{\{a\fresh X,b\fresh X\} \vdash \text{fn\ } a. X \approx \text{fn\ } b. X} |
|
364 \end{center}} |
|
365 |
|
366 \end{frame}} |
|
367 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
368 *} |
|
369 |
|
370 text_raw {* |
|
371 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
372 \mode<presentation>{ |
|
373 \begin{frame}<1>[c] |
|
374 \frametitle{Rules for Equivalence} |
|
375 |
|
376 \begin{center} |
|
377 \begin{tabular}{c} |
|
378 Excerpt\\ |
|
379 (i.e.~only the interesting rules) |
|
380 \end{tabular} |
|
381 \end{center} |
|
382 |
|
383 \end{frame}} |
|
384 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
385 *} |
|
386 |
|
387 text_raw {* |
|
388 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
389 \mode<presentation>{ |
|
390 \begin{frame}<1>[c] |
|
391 \frametitle{Rules for Equivalence} |
|
392 |
|
393 \begin{center} |
|
394 \begin{tabular}{c} |
|
395 \colorbox{cream}{\smath{\infer{\nabla \vdash a \approx a}{}}}\\[8mm] |
|
396 |
|
397 \colorbox{cream}{% |
|
398 \smath{\infer{\nabla \vdash \abst{a}{t} \approx \abst{a}{t'}} |
|
399 {\nabla \vdash t \approx t'}}}\\[8mm] |
|
400 |
|
401 \colorbox{cream}{% |
|
402 \smath{\infer{\nabla \vdash \abst{a}{t} \approx \abst{b}{t'}} |
|
403 {a\not=b\;\; & \nabla \vdash t \approx \swap{a}{b}\act t'\;\;& \nabla \vdash a\fresh t'}}} |
|
404 \end{tabular} |
|
405 \end{center} |
|
406 |
|
407 \end{frame}} |
|
408 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
409 *} |
|
410 |
|
411 text_raw {* |
|
412 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
413 \mode<presentation>{ |
|
414 \begin{frame}<1-3>[c] |
|
415 \frametitle{Rules for Equivalence} |
|
416 |
|
417 \begin{center} |
|
418 \colorbox{cream}{% |
|
419 \smath{% |
|
420 \infer{\nabla \vdash \pi\act X \approx \pi'\act X} |
|
421 {\begin{array}{c} |
|
422 (a\fresh X)\in\nabla\\ |
|
423 \text{for all}\; a \;\text{with}\;\pi\act a \not= \pi'\act a |
|
424 \end{array} |
|
425 }}} |
|
426 \end{center} |
|
427 |
|
428 \onslide<2->{ |
|
429 for example\\[4mm] |
|
430 |
|
431 \alt<2>{% |
|
432 \begin{center} |
|
433 \smath{\{a\fresh\!X, b\fresh\!X\} \vdash X \approx \swap{a}{b}\act X} |
|
434 \end{center}} |
|
435 {% |
|
436 \begin{center} |
|
437 \smath{\{a\fresh\!X, c\fresh\!X\} \vdash \swap{a}{c}\swap{a}{b}\act X \approx \swap{b}{c}\act X} |
|
438 \end{center}} |
|
439 |
|
440 \onslide<3->{ |
|
441 \begin{tabular}{@ {}lllll@ {}} |
|
442 because & |
|
443 \smath{\swap{a}{c}\swap{a}{b}}: & |
|
444 \smath{a\mapsto b} & |
|
445 \smath{\swap{b}{c}}: & |
|
446 \smath{a\mapsto a}\\ |
|
447 & & \smath{b\mapsto c} & & \smath{b\mapsto c}\\ |
|
448 & & \smath{c\mapsto a} & & \smath{c\mapsto b}\\ |
|
449 \end{tabular} |
|
450 disagree at \smath{a} and \smath{c}.} |
|
451 } |
|
452 |
|
453 \end{frame}} |
|
454 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
455 *} |
|
456 |
|
457 text_raw {* |
|
458 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
459 \mode<presentation>{ |
|
460 \begin{frame}<1>[c] |
|
461 \frametitle{Rules for Freshness} |
|
462 |
|
463 \begin{center} |
|
464 \begin{tabular}{c} |
|
465 Excerpt\\ |
|
466 (i.e.~only the interesting rules) |
|
467 \end{tabular} |
|
468 \end{center} |
|
469 |
|
470 \end{frame}} |
|
471 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
472 *} |
|
473 |
|
474 text_raw {* |
|
475 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
476 \mode<presentation>{ |
|
477 \begin{frame}<1>[c] |
|
478 \frametitle{Rules for Freshness} |
|
479 |
|
480 \begin{center} |
|
481 \begin{tabular}{c} |
|
482 \colorbox{cream}{% |
|
483 \smath{\infer{\nabla \vdash a\fresh b}{a\not= b}}}\\[5mm] |
|
484 |
|
485 \colorbox{cream}{% |
|
486 \smath{\infer{\nabla \vdash a\fresh\abst{a}{t}}{}}}\hspace{7mm} |
|
487 \colorbox{cream}{% |
|
488 \smath{\infer{\nabla \vdash a\fresh\abst{b}{t}} |
|
489 {a\not= b\;\; & \nabla \vdash a\fresh t}}}\\[5mm] |
|
490 |
|
491 \colorbox{cream}{% |
|
492 \smath{\infer{\nabla \vdash a\fresh \pi\act X} |
|
493 {(\pi^{-1}\act a\fresh X)\in\nabla}}} |
|
494 \end{tabular} |
|
495 \end{center} |
|
496 |
|
497 \end{frame}} |
|
498 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
499 *} |
|
500 |
|
501 text_raw {* |
|
502 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
503 \mode<presentation>{ |
|
504 \begin{frame}<1-4>[t] |
|
505 \frametitle{$\approx$ is an Equivalence} |
|
506 \mbox{}\\[5mm] |
|
507 |
|
508 \begin{center} |
|
509 \colorbox{cream}{\alert{Theorem:} |
|
510 $\approx$ is an equivalence relation.} |
|
511 \end{center}\bigskip |
|
512 |
|
513 \only<1>{% |
|
514 \begin{tabular}{ll} |
|
515 (Reflexivity) & $\smath{\nabla\vdash t\approx t}$\\[2mm] |
|
516 (Symmetry) & if $\smath{\nabla\vdash t_1\approx t_2}\;$ |
|
517 then $\;\smath{\nabla\vdash t_2\approx t_1}$\\[2mm] |
|
518 (Transitivity) & if $\smath{\nabla\vdash t_1\approx t_2}\;$ and |
|
519 $\;\smath{\nabla\vdash t_2\approx t_3}$\\ |
|
520 & then $\smath{\nabla\vdash t_1\approx t_3}$\\ |
|
521 \end{tabular}} |
|
522 |
|
523 \only<2->{% |
|
524 \begin{itemize} |
|
525 \item<2-> \smath{\nabla \vdash t\approx t'} then \smath{\nabla \vdash \pi\act t\approx \pi\act t'} |
|
526 |
|
527 \item<2-> \smath{\nabla \vdash a\fresh t} then |
|
528 \smath{\nabla \vdash \pi\act a\fresh \pi\act t} |
|
529 |
|
530 \item<3-> \smath{\nabla \vdash t\approx \pi\act t'} then |
|
531 \smath{\nabla \vdash (\pi^{-1})\act t\approx t'} |
|
532 |
|
533 \item<3-> \smath{\nabla \vdash a\fresh \pi\act t} then |
|
534 \smath{\nabla \vdash (\pi^{-1})\act a\fresh t} |
|
535 |
|
536 \item<4-> \smath{\nabla \vdash a\fresh t} and \smath{\nabla \vdash t\approx t'} then |
|
537 \smath{\nabla \vdash a\fresh t'} |
|
538 \end{itemize} |
|
539 } |
|
540 |
|
541 \end{frame}} |
|
542 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
543 *} |
|
544 |
|
545 text_raw {* |
|
546 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
547 \mode<presentation>{ |
|
548 \begin{frame}<1-4> |
|
549 \frametitle{Comparison $=_\alpha$} |
|
550 |
|
551 Traditionally \smath{=_\alpha} is defined as |
|
552 |
|
553 \begin{center} |
|
554 \colorbox{cream}{% |
|
555 \begin{minipage}{9cm} |
|
556 \raggedright least congruence which identifies \smath{\abst{a}{t}} |
|
557 with \smath{\abst{b}{[a:=b]t}} provided \smath{b} is not free |
|
558 in \smath{t} |
|
559 \end{minipage}} |
|
560 \end{center} |
|
561 |
|
562 where \smath{[a:=b]t} replaces all free occurrences of\\ |
|
563 \smath{a} by \smath{b} in \smath{t}. |
|
564 \bigskip |
|
565 |
|
566 \only<2>{% |
|
567 \begin{textblock}{13}(1.2,10) |
|
568 For \alert{ground} terms: |
|
569 |
|
570 \begin{center} |
|
571 \colorbox{cream}{% |
|
572 \begin{minipage}{9.0cm} |
|
573 \begin{tabular}{@ {}rl} |
|
574 \underline{Theorem:} |
|
575 & \smath{t=_\alpha t'\;\;} if\hspace{-0.5mm}f~\smath{\;\;\emptyset \vdash t\approx t'}\\[2mm] |
|
576 & \smath{a\not\in F\hspace{-0.9mm}A(t)\;\;} if\hspace{-0.5mm}f~\smath{\;\;\emptyset\vdash a\fresh t} |
|
577 \end{tabular} |
|
578 \end{minipage}} |
|
579 \end{center} |
|
580 \end{textblock}} |
|
581 |
|
582 \only<3>{% |
|
583 \begin{textblock}{13}(1.2,10) |
|
584 In general \smath{=_\alpha} and \smath{\approx} are distinct! |
|
585 \begin{center} |
|
586 \colorbox{cream}{% |
|
587 \begin{minipage}{6.0cm} |
|
588 \smath{\abst{a}{X}=_\alpha \abst{b}{X}\;} but not\\[2mm] |
|
589 \smath{\emptyset \vdash \abst{a}{X} \approx \abst{b}{X}\;} (\smath{a\not=b}) |
|
590 \end{minipage}} |
|
591 \end{center} |
|
592 \end{textblock}} |
|
593 |
|
594 \only<4>{ |
|
595 \begin{textblock}{6}(1,2) |
|
596 \begin{tikzpicture} |
|
597 \draw (0,0) node[inner sep=3mm,fill=cream, ultra thick, draw=red, rounded corners=2mm] |
|
598 {\color{darkgray} |
|
599 \begin{minipage}{10cm}\raggedright |
|
600 That is a crucial point: if we had\\[-2mm] |
|
601 \[\smath{\emptyset \vdash \abst{a}{X}\approx \abst{b}{X}}\mbox{,}\] |
|
602 then applying $\smath{[X:=a]}$, $\smath{[X:=b]}$, $\ldots$\\ |
|
603 give two terms that are {\bf not} $\alpha$-equivalent.\\[3mm] |
|
604 The freshness constraints $\smath{a\fresh X}$ and $\smath{b\fresh X}$ |
|
605 rule out the problematic substitutions. Therefore |
|
606 |
|
607 \[\smath{\{a\fresh X,b\fresh X\} \vdash \abst{a}{X}\approx \abst{b}{X}}\] |
|
608 |
|
609 does hold. |
|
610 \end{minipage}}; |
|
611 \end{tikzpicture} |
|
612 \end{textblock}} |
|
613 |
|
614 \end{frame}} |
|
615 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
616 *} |
|
617 |
|
618 text_raw {* |
|
619 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
620 \mode<presentation>{ |
|
621 \begin{frame}<1-9> |
|
622 \frametitle{Substitution} |
|
623 |
|
624 \begin{tabular}{l@ {\hspace{8mm}}r@ {\hspace{1.5mm}}c@ {\hspace{1.5mm}}l@ {}} |
|
625 \pgfuseshading{smallbluesphere} & |
|
626 \smath{\sigma(\abst{a}{t})} & \smath{\dn} & \smath{\abst{a}{\sigma(t)}}\\[2mm] |
|
627 |
|
628 \pgfuseshading{smallbluesphere} & |
|
629 \smath{\sigma(\pi\act X)} & \smath{\dn} & |
|
630 \smath{\begin{cases}% |
|
631 \pi\;\act\;\sigma(X) & \!\!\text{if\ } \sigma(X)\not=X\\ |
|
632 \pi\act X & \!\!\text{otherwise}% |
|
633 \end{cases}}\\[6mm] |
|
634 \end{tabular}\bigskip\bigskip |
|
635 |
|
636 \pause |
|
637 \only<2-5>{ |
|
638 \only<2->{for example} |
|
639 \def\arraystretch{1.3} |
|
640 \begin{tabular}{@ {\hspace{14mm}}l@ {\hspace{3mm}}l} |
|
641 \onslide<2->{\textcolor{white}{$\Rightarrow$}} & |
|
642 \onslide<2->{\alt<3>{\smath{\underline{\abst{a}{\swap{a}{b}\act X}\;\,[X:=\pair{b}{Y}]}}} |
|
643 {\smath{\abst{a}{\swap{a}{b}\act X}\;\,[X:=\pair{b}{Y}]}}}\\ |
|
644 \onslide<3->{\smath{\Rightarrow}} & |
|
645 \onslide<3->{\alt<3,4>{\smath{\abst{a}{\underline{\swap{a}{b}\act X[X:=\pair{b}{Y}]}}}} |
|
646 {\smath{\abst{a}{\swap{a}{b}\act X}[X:=\pair{b}{Y}]}}}\\ |
|
647 \onslide<4->{\smath{\Rightarrow}} & |
|
648 \onslide<4->{\alt<4>{\smath{\abst{a}{\swap{a}{b}\act \underline{\pair{b}{Y}}}}} |
|
649 {\smath{\abst{a}{\underline{\swap{a}{b}}\act \pair{b}{Y}}}}}\\ |
|
650 \onslide<5->{\smath{\Rightarrow}} & |
|
651 \onslide<5->{\smath{\abst{a}{\pair{a}{\swap{a}{b}\act Y}}}} |
|
652 \end{tabular}} |
|
653 |
|
654 \only<6-> |
|
655 {\begin{tabular}{l@ {\hspace{8mm}}l@ {}} |
|
656 \pgfuseshading{smallbluesphere} & |
|
657 if \smath{\nabla\vdash t\approx t'} and\hspace{-2mm}\mbox{} |
|
658 \raisebox{-2.7mm}{ |
|
659 \alt<7>{\begin{tikzpicture} |
|
660 \draw (0,0) node[inner sep=1mm,fill=cream, very thick, draw=red, rounded corners=3mm] |
|
661 {\smath{\;\nabla'\vdash\sigma(\nabla)\;}}; |
|
662 \end{tikzpicture}} |
|
663 {\begin{tikzpicture} |
|
664 \draw (0,0) node[inner sep=1mm,fill=white, very thick, draw=white, rounded corners=3mm] |
|
665 {\smath{\;\nabla'\vdash\sigma(\nabla)\;}}; |
|
666 \end{tikzpicture}}}\\ |
|
667 & then \smath{\nabla'\vdash\sigma(t)\approx\sigma(t')} |
|
668 \end{tabular}} |
|
669 |
|
670 \only<9> |
|
671 {\begin{tabular}{l@ {\hspace{8mm}}l@ {}} |
|
672 \\[-4mm] |
|
673 \pgfuseshading{smallbluesphere} & |
|
674 \smath{\sigma(\pi\act t)=\pi\act\sigma(t)} |
|
675 \end{tabular}} |
|
676 |
|
677 |
|
678 \only<7>{ |
|
679 \begin{textblock}{6}(10,10.5) |
|
680 \begin{tikzpicture} |
|
681 \draw (0,0) node[inner sep=1mm,fill=cream, very thick, draw=red, rounded corners=2mm] |
|
682 {\color{darkgray} |
|
683 \begin{minipage}{3.8cm}\raggedright |
|
684 this means\\[1mm] |
|
685 \smath{\nabla'\vdash a\fresh\sigma(X)}\\[1mm] |
|
686 holds for all\\[1mm] |
|
687 \smath{(a\fresh X)\in\nabla} |
|
688 \end{minipage}}; |
|
689 \end{tikzpicture} |
|
690 \end{textblock}} |
|
691 |
|
692 \end{frame}} |
|
693 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
694 *} |
|
695 |
|
696 text_raw {* |
|
697 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
698 \mode<presentation>{ |
|
699 \begin{frame}<1-> |
|
700 \frametitle{Equational Problems} |
|
701 |
|
702 An equational problem |
|
703 \[ |
|
704 \colorbox{cream}{\smath{t \eqprob t'}} |
|
705 \] |
|
706 is \alert{solved} by |
|
707 |
|
708 \begin{center} |
|
709 \begin{tabular}{ll} |
|
710 \pgfuseshading{smallbluesphere} & a substitution \smath{\sigma} (terms for variables)\\[3mm] |
|
711 \pgfuseshading{smallbluesphere} & {\bf and} a set of freshness assumptions \smath{\nabla} |
|
712 \end{tabular} |
|
713 \end{center} |
|
714 |
|
715 so that \smath{\nabla\vdash \sigma(t)\approx \sigma(t')}. |
|
716 |
|
717 |
|
718 \end{frame}} |
|
719 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
720 *} |
|
721 |
|
722 text_raw {* |
|
723 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
724 \mode<presentation>{ |
|
725 \begin{frame}<1-> |
|
726 |
|
727 Unifying equations may entail solving |
|
728 \alert{freshness problems}. |
|
729 |
|
730 \bigskip |
|
731 |
|
732 E.g.~assuming that \smath{a\not=a'}, then |
|
733 \[ |
|
734 \smath{\abst{a}{t}\eqprob \abst{a'}{t'}} |
|
735 \] |
|
736 can only be solved if |
|
737 \[ |
|
738 \smath{t\eqprob \swap{a}{a'}\act t'} \quad\text{\emph{and}}\quad |
|
739 \smath{a\freshprob t'} |
|
740 \] |
|
741 can be solved. |
|
742 |
|
743 \end{frame}} |
|
744 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
745 *} |
|
746 |
|
747 text_raw {* |
|
748 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
749 \mode<presentation>{ |
|
750 \begin{frame}<1-> |
|
751 \frametitle{Freshness Problems} |
|
752 |
|
753 A freshness problem |
|
754 \[ |
|
755 \colorbox{cream}{\smath{a \freshprob t}} |
|
756 \] |
|
757 is \alert{solved} by |
|
758 |
|
759 \begin{center} |
|
760 \begin{tabular}{ll} |
|
761 \pgfuseshading{smallbluesphere} & a substitution \smath{\sigma}\\[3mm] |
|
762 \pgfuseshading{smallbluesphere} & and a set of freshness assumptions \smath{\nabla} |
|
763 \end{tabular} |
|
764 \end{center} |
|
765 |
|
766 so that \smath{\nabla\vdash a \fresh \sigma(t)}. |
|
767 |
|
768 \end{frame}} |
|
769 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
770 *} |
|
771 |
|
772 text_raw {* |
|
773 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
774 \mode<presentation>{ |
|
775 \begin{frame}<1-3> |
|
776 \frametitle{Existence of MGUs} |
|
777 |
|
778 \underline{Theorem}: There is an algorithm which, given a nominal |
|
779 unification problem \smath{P}, decides whether\\ |
|
780 or not it has a solution \smath{(\sigma,\nabla)}, and returns a \\ |
|
781 \alert{most general} one if it does.\bigskip\bigskip |
|
782 |
|
783 \only<3>{ |
|
784 Proof: one can reduce all the equations to `solved form' |
|
785 first (creating a substitution), and then solve the freshness |
|
786 problems (easy).} |
|
787 |
|
788 \only<2>{ |
|
789 \begin{textblock}{6}(2.5,9.5) |
|
790 \begin{tikzpicture} |
|
791 \draw (0,0) node[inner sep=3mm,fill=cream, ultra thick, draw=red, rounded corners=2mm] |
|
792 {\color{darkgray} |
|
793 \begin{minipage}{8cm}\raggedright |
|
794 \alert{most general:}\\ |
|
795 straightforward definition\\ |
|
796 ``if\hspace{-0.5mm}f there exists a \smath{\tau} such that \ldots'' |
|
797 \end{minipage}}; |
|
798 \end{tikzpicture} |
|
799 \end{textblock}} |
|
800 |
|
801 \end{frame}} |
|
802 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
803 *} |
|
804 |
|
805 text_raw {* |
|
806 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
807 \mode<presentation>{ |
|
808 \begin{frame}<1> |
|
809 \frametitle{Remember the Quiz?} |
|
810 |
|
811 \textcolor{gray}{Assuming that $a$ and $b$ are distinct variables,\\ |
|
812 is it possible to find $\lambda$-terms $M_1$ to $M_7$ |
|
813 that make the following pairs $\alpha$-equivalent?} |
|
814 |
|
815 \begin{tabular}{@ {\hspace{14mm}}p{12cm}} |
|
816 \begin{itemize} |
|
817 \item \smath{\lambda a.\lambda b. (M_1\,b)\;} and |
|
818 \smath{\lambda b.\lambda a. (a\,M_1)\;} |
|
819 |
|
820 \item \textcolor{gray}{$\lambda a.\lambda b. (M_2\,b)\;$ and |
|
821 $\lambda b.\lambda a. (a\,M_3)\;$} |
|
822 |
|
823 \item \textcolor{gray}{$\lambda a.\lambda b. (b\,M_4)\;$ and |
|
824 $\lambda b.\lambda a. (a\,M_5)\;$} |
|
825 |
|
826 \item \smath{\lambda a.\lambda b. (b\,M_6)\;} and |
|
827 \smath{\lambda a.\lambda a. (a\,M_7)\;} |
|
828 \end{itemize} |
|
829 \end{tabular} |
|
830 |
|
831 \textcolor{gray}{If there is one solution for a pair, can you |
|
832 describe all its solutions?} |
|
833 |
|
834 |
|
835 \end{frame}} |
|
836 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
837 *} |
|
838 |
|
839 text_raw {* |
|
840 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
841 \mode<presentation>{ |
|
842 \begin{frame}<1-> |
|
843 \frametitle{Answers to the Quiz} |
|
844 \small |
|
845 \def\arraystretch{1.6} |
|
846 \begin{tabular}{c@ {\hspace{2mm}}l} |
|
847 & \only<1>{\smath{\lambda a.\lambda b. (M_1\,b)\;} and \smath{\;\lambda b.\lambda a. (a\,M_1)}}% |
|
848 \only<2->{\smath{\abst{a}{\abst{b}{\pair{M_1}{b}}} \;\eqprob\; \abst{b}{\abst{a}{\pair{a}{M_1}}}}}\\ |
|
849 |
|
850 \onslide<3->{\smath{\redu{\id}}} & |
|
851 \only<3>{\smath{\abst{b}{\pair{M_1}{b}} \eqprob |
|
852 \alert{\swap{a}{b}} \act \abst{a}{\pair{a}{M_1}}\;,\;a\freshprob \abst{a}{\pair{a}{M_1}}}}% |
|
853 \only<4->{\smath{\abst{b}{\pair{M_1}{b}} \eqprob \abst{b}{\pair{b}{\swap{a}{b}\act M_1}}\;,\ |
|
854 a\freshprob \abst{a}{\pair{a}{M_1}}}}\\ |
|
855 |
|
856 \onslide<5->{\smath{\redu{\id}}} & |
|
857 \only<5->{\smath{\pair{M_1}{b} \eqprob \pair{b}{\swap{a}{b}\act M_1}\;,\;% |
|
858 a\freshprob \abst{a}{\pair{a}{M_1}}}}\\ |
|
859 |
|
860 \onslide<6->{\smath{\redu{\id}}} & |
|
861 \only<6->{\smath{M_1 \eqprob b \;,\; b \eqprob \swap{a}{b}\act M_1\;,\;% |
|
862 a\freshprob \abst{a}{\pair{a}{M_1}}}}\\ |
|
863 |
|
864 \onslide<7->{\smath{\redu{[M_1:=b]}}} & |
|
865 \only<7>{\smath{b \eqprob \swap{a}{b}\act \alert{b}\;,\;% |
|
866 a\freshprob \abst{a}{\pair{a}{\alert{b}}}}}% |
|
867 \only<8->{\smath{b \eqprob a\;,\; a\freshprob \abst{a}{\pair{a}{b}}}}\\ |
|
868 |
|
869 \onslide<9->{\smath{\redu{}}} & |
|
870 \only<9->{\smath{F\hspace{-0.5mm}AIL}} |
|
871 \end{tabular} |
|
872 |
|
873 \only<10>{ |
|
874 \begin{textblock}{6}(2,11) |
|
875 \begin{tikzpicture} |
|
876 \draw (0,0) node[inner sep=3mm,fill=cream, ultra thick, draw=red, rounded corners=2mm] |
|
877 {\color{darkgray} |
|
878 \begin{minipage}{9cm}\raggedright |
|
879 \smath{\lambda a.\lambda b. (M_1\,b)} \smath{=_\alpha} |
|
880 \smath{\lambda b.\lambda a. (a\,M_1)} has no solution |
|
881 \end{minipage}}; |
|
882 \end{tikzpicture} |
|
883 \end{textblock}} |
|
884 |
|
885 \end{frame}} |
|
886 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
887 *} |
|
888 |
|
889 text_raw {* |
|
890 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
891 \mode<presentation>{ |
|
892 \begin{frame}<1-> |
|
893 \frametitle{Answers to the Quiz} |
|
894 \small |
|
895 \def\arraystretch{1.6} |
|
896 \begin{tabular}{c@ {\hspace{2mm}}l} |
|
897 & \only<1>{\smath{\lambda a.\lambda b. (b\,M_6)\;} and \smath{\;\lambda a.\lambda a. (a\,M_7)}}% |
|
898 \only<2->{\smath{\abst{a}{\abst{b}{\pair{b}{M_6}}} \;\eqprob\; \abst{a}{\abst{a}{\pair{a}{M_7}}}}}\\ |
|
899 |
|
900 \onslide<3->{\smath{\redu{\id}}} & |
|
901 \only<3->{\smath{\abst{b}{\pair{b}{M_6}} \eqprob \abst{a}{\pair{a}{M_7}}}}\\ |
|
902 |
|
903 \onslide<4->{\smath{\redu{\id}}} & |
|
904 \only<4->{\smath{\pair{b}{M_6} \eqprob \pair{b}{\swap{b}{a}\act M_7}\;,\;b\freshprob\pair{a}{M_7}}}\\ |
|
905 |
|
906 \onslide<5->{\smath{\redu{\id}}} & |
|
907 \only<5->{\smath{b\eqprob b\;,\; M_6 \eqprob \swap{b}{a}\act M_7\;,\;% |
|
908 b\freshprob \pair{a}{M_7}}}\\ |
|
909 |
|
910 \onslide<6->{\smath{\redu{\id}}} & |
|
911 \only<6->{\smath{M_6 \eqprob \swap{b}{a}\act M_7\;,\;% |
|
912 b\freshprob \pair{a}{M_7}}}\\ |
|
913 |
|
914 \onslide<7->{\makebox[0mm]{\smath{\redu{[M_6:=\swap{b}{a}\act M_7]}}}} & |
|
915 \only<7->{\smath{\qquad b\freshprob \pair{a}{M_7}}}\\ |
|
916 |
|
917 \onslide<8->{\smath{\redu{\varnothing}}} & |
|
918 \only<8->{\smath{b\freshprob a\;,\;b\freshprob M_7}}\\ |
|
919 |
|
920 \onslide<9->{\smath{\redu{\varnothing}}} & |
|
921 \only<9->{\smath{b\freshprob M_7}}\\ |
|
922 |
|
923 \onslide<10->{\makebox[0mm]{\smath{\redu{\{b\fresh M_7\}}}}} & |
|
924 \only<10->{\smath{\;\;\varnothing}}\\ |
|
925 |
|
926 \end{tabular} |
|
927 |
|
928 \only<10>{ |
|
929 \begin{textblock}{6}(6,9) |
|
930 \begin{tikzpicture} |
|
931 \draw (0,0) node[inner sep=3mm,fill=cream, ultra thick, draw=red, rounded corners=2mm] |
|
932 {\color{darkgray} |
|
933 \begin{minipage}{7cm}\raggedright |
|
934 \smath{\lambda a.\lambda b. (b\,M_6)\;} \smath{=_\alpha} |
|
935 \smath{\;\lambda a.\lambda a. (a\,M_7)}\\[2mm] |
|
936 we can take \smath{M_7} to be any $\lambda$-term that does not |
|
937 contain free occurrences of \smath{b}, so long as we take \smath{M_6} to |
|
938 be the result of swapping all occurrences of \smath{b} and \smath{a} |
|
939 throughout \smath{M_7} |
|
940 \end{minipage}}; |
|
941 \end{tikzpicture} |
|
942 \end{textblock}} |
|
943 |
|
944 \end{frame}} |
|
945 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
946 *} |
|
947 |
|
948 text_raw {* |
|
949 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
950 \mode<presentation>{ |
|
951 \begin{frame}<1-> |
|
952 \frametitle{Properties} |
|
953 |
|
954 \begin{itemize} |
|
955 \item An interesting feature of nominal unification is that it |
|
956 does not need to create new atoms.\bigskip |
|
957 |
|
958 \begin{center}\small |
|
959 \colorbox{cream}{ |
|
960 \smath{\{a.t \eqprob b.t'\}\cup P \redu{\id} \{t \eqprob \swap{a}{b}\act t', a \freshprob t'\} \cup P}} |
|
961 \end{center}\bigskip\bigskip |
|
962 \pause |
|
963 |
|
964 \item The alternative rule |
|
965 |
|
966 \begin{center}\small |
|
967 \colorbox{cream}{ |
|
968 \begin{tabular}{@ {}l@ {}} |
|
969 \smath{\{a.t \eqprob b.t'\}\cup P \redu{\id}}\\ |
|
970 \mbox{}\hspace{2cm}\smath{\{\swap{a}{c}\act t \eqprob |
|
971 \swap{b}{c}\act t', c \freshprob t, c \freshprob t'\} \cup P} |
|
972 \end{tabular}} |
|
973 \end{center} |
|
974 |
|
975 leads to a more complicated notion of mgu.\medskip\pause |
|
976 |
|
977 \footnotesize |
|
978 \smath{\{a.X \eqprob b.Y\} \redu{} (\{a\fresh Y, c\fresh Y\}, [X:=\swap{a}{c}\swap{b}{c}\act Y])} |
|
979 \end{itemize} |
|
980 |
|
981 \end{frame}} |
|
982 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
983 *} |
|
984 |
|
985 text_raw {* |
|
986 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
987 \mode<presentation>{ |
|
988 \begin{frame}<1-3> |
|
989 \frametitle{Is it Useful?} |
|
990 |
|
991 Yes. $\alpha$Prolog by James Cheney (main developer)\bigskip\bigskip |
|
992 |
|
993 \color{darkgray} |
|
994 \begin{tabular}{@ {}l} |
|
995 type (Gamma, var(X), T) :- member (X,T) Gamma.\smallskip\medskip\\ |
|
996 |
|
997 type (Gamma, app(M, N), T') :-\\ |
|
998 \hspace{3cm}type (Gamma, M, arrow(T, T')),\\ |
|
999 \hspace{3cm}type (Gamma, N, T).\smallskip\medskip\\ |
|
1000 |
|
1001 type (Gamma, lam(\alert{x.M}), arrow(T, T')) / \alert{x \# Gamma} :-\\ |
|
1002 \hspace{3cm}type ((x, T)::Gamma, M, T').\smallskip\medskip\\ |
|
1003 |
|
1004 member X X::Tail.\\ |
|
1005 member X Y::Tail :- member X Tail.\\ |
|
1006 \end{tabular} |
|
1007 |
|
1008 \only<2->{ |
|
1009 \begin{textblock}{6}(1.5,0.5) |
|
1010 \begin{tikzpicture} |
|
1011 \draw (0,0) node[inner sep=3mm,fill=cream, ultra thick, draw=red, rounded corners=2mm] |
|
1012 {\color{darkgray} |
|
1013 \begin{minipage}{9cm}\raggedright |
|
1014 {\bf One problem:} If we ask whether |
|
1015 |
|
1016 \begin{center} |
|
1017 ?- type ([(x, T')], lam(x.Var(x)), T) |
|
1018 \end{center} |
|
1019 |
|
1020 is typable, we expect an answer for T.\bigskip |
|
1021 |
|
1022 \onslide<3>{Solution: Before back-chaining freshen all variables and atoms |
|
1023 in a program (clause).} |
|
1024 \end{minipage}}; |
|
1025 \end{tikzpicture} |
|
1026 \end{textblock}} |
|
1027 |
|
1028 \end{frame}} |
|
1029 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1030 *} |
|
1031 |
|
1032 text_raw {* |
|
1033 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1034 \mode<presentation>{ |
|
1035 \begin{frame}<1-> |
|
1036 \frametitle{Equivariant Unification} |
|
1037 |
|
1038 James Cheney proposed |
|
1039 |
|
1040 \begin{center} |
|
1041 \colorbox{cream}{ |
|
1042 \smath{t \eqprob t' \redu{\nabla, \sigma, \pi} |
|
1043 \nabla \vdash \sigma(t) \approx \pi \act \sigma(t')}} |
|
1044 \end{center}\bigskip\bigskip |
|
1045 \pause |
|
1046 |
|
1047 But he also showed this problem is undecidable\\ in general. :( |
|
1048 |
|
1049 \end{frame}} |
|
1050 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1051 *} |
|
1052 |
|
1053 text_raw {* |
|
1054 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1055 \mode<presentation>{ |
|
1056 \begin{frame}<1-> |
|
1057 \frametitle{Taking Atoms as Variables} |
|
1058 |
|
1059 Instead of \smath{a.X}, have \smath{A.X}.\bigskip |
|
1060 \pause |
|
1061 |
|
1062 Unfortunately this breaks the mgu-property: |
|
1063 |
|
1064 \begin{center} |
|
1065 \smath{a.Z \eqprob X.Y.v(a)} |
|
1066 \end{center} |
|
1067 |
|
1068 can be solved by |
|
1069 |
|
1070 \begin{center} |
|
1071 \smath{[X:=a, Z:=Y.v(a)]} and |
|
1072 \smath{[Y:=a, Z:=Y.v(Y)]} |
|
1073 \end{center} |
|
1074 |
|
1075 \end{frame}} |
|
1076 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1077 *} |
|
1078 |
|
1079 text_raw {* |
|
1080 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1081 \mode<presentation>{ |
|
1082 \begin{frame}<1>[c] |
|
1083 \frametitle{HOPU vs. NOMU} |
|
1084 |
|
1085 \begin{itemize} |
|
1086 \item James Cheney showed\bigskip |
|
1087 \begin{center} |
|
1088 \colorbox{cream}{\smath{HOPU \Rightarrow NOMU}} |
|
1089 \end{center}\bigskip |
|
1090 |
|
1091 \item Jordi Levy and Mateu Villaret established\bigskip |
|
1092 \begin{center} |
|
1093 \colorbox{cream}{\smath{HOPU \Leftarrow NOMU}} |
|
1094 \end{center}\bigskip |
|
1095 \end{itemize} |
|
1096 |
|
1097 The translations `explode' the problems quadratically. |
|
1098 |
|
1099 \end{frame}} |
|
1100 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1101 *} |
|
1102 |
|
1103 text_raw {* |
|
1104 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1105 \mode<presentation>{ |
|
1106 \begin{frame}<1> |
|
1107 \small\tt |
|
1108 |
|
1109 \begin{minipage}{13cm} |
|
1110 \begin{tabular}{@ {\hspace{-2mm}}p{11.5cm}} |
|
1111 \\ |
|
1112 From: Zhenyu Qian <zhqian@microsoft.com>\\ |
|
1113 To: Christian Urban <urbanc@in.tum.de>\\ |
|
1114 Subject: RE: Linear Higher-Order Pattern Unification\\ |
|
1115 Date: Mon, 14 Apr 2008 09:56:47 +0800\\ |
|
1116 \\ |
|
1117 Hi Christian,\\ |
|
1118 \\ |
|
1119 Thanks for your interests and asking. I know that that paper is complex. As |
|
1120 I told Tobias when we met last time, I have raised the question to myself |
|
1121 many times whether the proof could have some flaws, and so making it through |
|
1122 a theorem prover would definitely bring piece to my mind (no matter what |
|
1123 the result would be). The only problem for me is the time.\\ |
|
1124 \ldots\\ |
|
1125 |
|
1126 Thanks/Zhenyu |
|
1127 \end{tabular} |
|
1128 \end{minipage} |
|
1129 |
|
1130 \end{frame}} |
|
1131 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1132 *} |
|
1133 |
|
1134 text_raw {* |
|
1135 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1136 \mode<presentation>{ |
|
1137 \begin{frame}<1> |
|
1138 \frametitle{Complexity} |
|
1139 |
|
1140 \begin{itemize} |
|
1141 \item Christiopher Calves and Maribel Fernandez showed first that |
|
1142 it is polynomial and then also quadratic |
|
1143 |
|
1144 \item Jordi Levy and Mateu Villaret showed that it is quadratic |
|
1145 by a translation into a subset of NOMU and using ideas from |
|
1146 Martelli/Montenari. |
|
1147 |
|
1148 \end{itemize} |
|
1149 |
|
1150 \end{frame}} |
|
1151 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1152 *} |
|
1153 |
|
1154 |
|
1155 text_raw {* |
|
1156 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1157 \mode<presentation>{ |
|
1158 \begin{frame}<1->[c] |
|
1159 \frametitle{Conclusion} |
|
1160 |
|
1161 \begin{itemize} |
|
1162 \item Nominal Unification is a completely first-order |
|
1163 language, but implements unification modulo $\alpha$. |
|
1164 \textcolor{gray}{(verification\ldots Ramana Kumar and Michael Norrish)} |
|
1165 \medskip\pause |
|
1166 |
|
1167 \item NOMU has been applied in term-rewriting and |
|
1168 logic programming. \textcolor{gray}{(Maribel Fernandez et |
|
1169 al has a KB-completion procedure.)} |
|
1170 I hope it will also be used in typing |
|
1171 systems.\medskip\pause |
|
1172 |
|
1173 \item NOMU and HOPU are `equivalent' (it took a long time |
|
1174 and considerable research to find this out).\medskip\pause |
|
1175 |
|
1176 \item The question about complexity is still an ongoing |
|
1177 story.\medskip |
|
1178 \end{itemize} |
|
1179 |
|
1180 |
|
1181 \end{frame}} |
|
1182 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1183 *} |
|
1184 |
|
1185 text_raw {* |
|
1186 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1187 \mode<presentation>{ |
|
1188 \begin{frame}<1>[c] |
|
1189 \frametitle{ |
|
1190 \begin{tabular}{c} |
|
1191 \mbox{}\\[23mm] |
|
1192 \alert{\LARGE Thank you very much!}\\ |
|
1193 \alert{\Large Questions?} |
|
1194 \end{tabular}} |
|
1195 |
|
1196 \end{frame}} |
|
1197 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1198 *} |
|
1199 |
|
1200 text_raw {* |
|
1201 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1202 \mode<presentation>{ |
|
1203 \begin{frame}<1-3> |
|
1204 \frametitle{Most General Unifiers} |
|
1205 |
|
1206 \underline{Definition}: For a unification problem |
|
1207 \smath{P}, a solution \smath{(\sigma_1,\nabla_1)} is |
|
1208 \alert{more general} than another solution |
|
1209 \smath{(\sigma_2,\nabla_2)}, iff~there exists a substitution |
|
1210 \smath{\tau} with |
|
1211 |
|
1212 \begin{center} |
|
1213 \begin{tabular}{ll} |
|
1214 \pgfuseshading{smallbluesphere} & |
|
1215 \alt<2>{\smath{\alert{\nabla_2\vdash\tau(\nabla_1)}}} |
|
1216 {\smath{\nabla_2\vdash\tau(\nabla_1)}}\\ |
|
1217 \pgfuseshading{smallbluesphere} & |
|
1218 \alt<3>{\smath{\alert{\nabla_2\vdash\sigma_2\approx \tau\circ\sigma_1}}} |
|
1219 {\smath{\nabla_2\vdash\sigma_2\approx \tau\circ\sigma_1}} |
|
1220 \end{tabular} |
|
1221 \end{center} |
|
1222 |
|
1223 \only<2>{ |
|
1224 \begin{textblock}{13}(1.5,10.5) |
|
1225 \smath{\nabla_2\vdash a\fresh \sigma(X)} holds for all |
|
1226 \smath{(a\fresh X)\in\nabla_1} |
|
1227 \end{textblock}} |
|
1228 |
|
1229 \only<3>{ |
|
1230 \begin{textblock}{11}(1.5,10.5) |
|
1231 \smath{\nabla_2\vdash \sigma_2(X)\approx |
|
1232 \sigma(\sigma_1(X))} |
|
1233 holds for all |
|
1234 \smath{X\in\text{dom}(\sigma_2)\cup\text{dom}(\sigma\circ\sigma_1)} |
|
1235 \end{textblock}} |
|
1236 |
|
1237 \end{frame}} |
|
1238 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
|
1239 *} |
|
1240 |
|
1241 (*<*) |
|
1242 end |
|
1243 (*>*) |
|